Table of Contents
- 5.1. The MySQL Server
- 5.1.1. Server Option and Variable Reference
- 5.1.2. Server Configuration Defaults
- 5.1.3. Server Command Options
- 5.1.4. Server System Variables
- 5.1.5. Using System Variables
- 5.1.6. Server Status Variables
- 5.1.7. Server SQL Modes
- 5.1.8. Server Plugins
- 5.1.9. IPv6 Support
- 5.1.10. Server-Side Help
- 5.1.11. Server Response to Signals
- 5.1.12. The Shutdown Process
- 5.2. MySQL Server Logs
- 5.3. Managing Disk I/O and File Space for
InnoDBTables - 5.4. Creating and Using
InnoDBTables and Indexes - 5.4.1. Managing InnoDB Tablespaces
- 5.4.2. Grouping DML Operations with Transactions
- 5.4.3. Converting Tables from
MyISAMtoInnoDB - 5.4.4.
AUTO_INCREMENTHandling inInnoDB - 5.4.5.
InnoDBandFOREIGN KEYConstraints - 5.4.6. Working with
InnoDBCompressed Tables - 5.4.7.
InnoDBFile-Format Management - 5.4.8. How
InnoDBStores Variable-Length Columns
- 5.5. Online DDL for
InnoDBTables - 5.5.1. Overview of Online DDL
- 5.5.2. Performance and Concurrency Considerations for Online DDL
- 5.5.3. SQL Syntax for Online DDL
- 5.5.4. Combining or Separating DDL Statements
- 5.5.5. Examples of Online DDL
- 5.5.6. Implementation Details of Online DDL
- 5.5.7. How Crash Recovery Works with Online DDL
- 5.5.8. Online DDL for Partitioned
InnoDBTables - 5.5.9. Limitations of Online DDL
- 5.6. Running Multiple MySQL Instances on One Machine
- 5.7. Tracing mysqld Using DTrace
MySQL Server (mysqld) is the main program that does most of the work in a MySQL installation. This chapter provides an overview of MySQL Server and covers general server administration:
Server configuration.
The server log files.
Management of multiple servers on a single machine.
For additional information on administrative topics, see also:
- 5.1.1. Server Option and Variable Reference
- 5.1.2. Server Configuration Defaults
- 5.1.3. Server Command Options
- 5.1.4. Server System Variables
- 5.1.5. Using System Variables
- 5.1.6. Server Status Variables
- 5.1.7. Server SQL Modes
- 5.1.8. Server Plugins
- 5.1.9. IPv6 Support
- 5.1.10. Server-Side Help
- 5.1.11. Server Response to Signals
- 5.1.12. The Shutdown Process
mysqld is the MySQL server. The following discussion covers these MySQL server configuration topics:
Startup options that the server supports. You can specify these options on the command line, through configuration files, or both.
Server system variables. These variables reflect the current state and values of the startup options, some of which can be modified while the server is running.
Server status variables. These variables contain counters and statistics about runtime operation.
How to set the server SQL mode. This setting modifies certain aspects of SQL syntax and semantics, for example for compatibility with code from other database systems, or to control the error handling for particular situations.
The server shutdown process. There are performance and reliability considerations depending on the type of table (transactional or nontransactional) and whether you use replication.
Not all storage engines are supported by all MySQL server binaries
and configurations. To find out how to determine which storage
engines your MySQL server installation supports, see
Section 13.7.5.15, “SHOW ENGINES Syntax”.
The following table provides a list of all the command line
options, server and status variables applicable within
mysqld.
The table lists command-line options (Cmd-line), options valid in configuration files (Option file), server system variables (System Var), and status variables (Status var) in one unified list, with notification of where each option/variable is valid. If a server option set on the command line or in an option file differs from the name of the corresponding server system or status variable, the variable name is noted immediately below the corresponding option. For status variables, the scope of the variable is shown (Scope) as either global, session, or both. Please see the corresponding sections for details on setting and using the options and variables. Where appropriate, a direct link to further information on the item as available.
Table 5.1. Option/Variable Summary
The MySQL server has many operating parameters, which you can change at server startup using command-line options or configuration files (option files). It is also possible to change many parameters at runtime. For general instructions on setting parameters at startup or runtime, see Section 5.1.3, “Server Command Options”, and Section 5.1.4, “Server System Variables”.
On Unix platforms, mysql_install_db creates a
default option file named my.cnf in the base
installation directory. This file is created from a template
included in the distribution package named
my-default.cnf. You can find the template in
or under the base installation directory. When started using
mysqld_safe, the server uses
my.cnf file by default. If
my.cnf already exists,
mysql_install_db assumes it to be in use and
writes a new file named my-new.cnf instead.
With one exception, the settings in the default option file are
commented and have no effect. The exception is that the file
changes the sql_mode system
variable from its default of
NO_ENGINE_SUBSTITUTION to also
include STRICT_TRANS_TABLES:
sql_mode=NO_ENGINE_SUBSTITUTION,STRICT_TRANS_TABLES
This setting produces a server configuration that results in errors rather than warnings for bad data in operations that modify transactional tables. See Section 5.1.7, “Server SQL Modes”.
On Windows, MySQL Installer interacts with the user and creates a
file named my.ini in the base installation
directory as the default option file. If you install on Windows
from a Zip archive, you can copy the
my-default.ini template file in the base
installation directory to my.ini and use the
latter as the default option file.
On Windows, the .ini or
.cnf option file extension might not be
displayed.
On any platform, after completing the installation process, you
can edit the default option file at any time to modify the
parameters used by the server. For example, to use a parameter
setting in the file that is commented with a #
character at the beginning of the line, remove the
#, and modify the parameter value if necessary.
To disable a setting, either add a # to the
beginning of the line or remove it.
For additional information about option file format and syntax, see Section 4.2.3.3, “Using Option Files”.
When you start the mysqld server, you can specify program options using any of the methods described in Section 4.2.3, “Specifying Program Options”. The most common methods are to provide options in an option file or on the command line. However, in most cases it is desirable to make sure that the server uses the same options each time it runs. The best way to ensure this is to list them in an option file. See Section 4.2.3.3, “Using Option Files”.
mysqld reads options from the
[mysqld] and [server]
groups. mysqld_safe reads options from the
[mysqld], [server],
[mysqld_safe], and
[safe_mysqld] groups.
mysql.server reads options from the
[mysqld] and [mysql.server]
groups.
An embedded MySQL server usually reads options from the
[server], [embedded], and
[
groups, where xxxxx_SERVER]xxxxx is the name of the
application into which the server is embedded.
mysqld accepts many command options. For a brief summary, execute mysqld --help. To see the full list, use mysqld --verbose --help.
The following list shows some of the most common server options. Additional options are described in other sections:
Options that affect security: See Section 6.1.4, “Security-Related mysqld Options and Variables”.
SSL-related options: See Section 6.3.9.4, “SSL Command Options”.
Binary log control options: See Section 5.2.4, “The Binary Log”.
Replication-related options: See Section 16.1.4, “Replication and Binary Logging Options and Variables”.
Options for loading plugins such as pluggable storage engines: See Section 5.1.8.1, “Installing and Uninstalling Plugins”.
Options specific to particular storage engines: See Section 14.2.6, “
InnoDBStartup Options and System Variables” and Section 14.3.1, “MyISAMStartup Options”.
You can also set the values of server system variables by using variable names as options, as described at the end of this section.
Some options control the size of buffers or caches. For a given buffer, the server might need to allocate internal data structures. These structures typically are allocated from the total memory allocated to the buffer, and the amount of space required might be platform dependent. This means that when you assign a value to an option that controls a buffer size, the amount of space actually available might differ from the value assigned. In some cases, the amount might be less than the value assigned. It is also possible that the server will adjust a value upward. For example, if you assign a value of 0 to an option for which the minimal value is 1024, the server will set the value to 1024.
Values for buffer sizes, lengths, and stack sizes are given in bytes unless otherwise specified.
Some options take file name values. Unless otherwise specified,
the default file location is the data directory if the value is a
relative path name. To specify the location explicitly, use an
absolute path name. Suppose that the data directory is
/var/mysql/data. If a file-valued option is
given as a relative path name, it will be located under
/var/mysql/data. If the value is an absolute
path name, its location is as given by the path name.
--help,-?Command-Line Format -?--helpOption-File Format helpDisplay a short help message and exit. Use both the
--verboseand--helpoptions to see the full message.-
Command-Line Format --allow-suspicious-udfsOption-File Format allow-suspicious-udfsPermitted Values Type booleanDefault FALSEThis option controls whether user-defined functions that have only an
xxxsymbol for the main function can be loaded. By default, the option is off and only UDFs that have at least one auxiliary symbol can be loaded; this prevents attempts at loading functions from shared object files other than those containing legitimate UDFs. See Section 22.3.2.6, “User-Defined Function Security Precautions”. -
Command-Line Format --ansi-aOption-File Format ansiUse standard (ANSI) SQL syntax instead of MySQL syntax. For more precise control over the server SQL mode, use the
--sql-modeoption instead. See Section 1.8.3, “Running MySQL in ANSI Mode”, and Section 5.1.7, “Server SQL Modes”. -
Command-Line Format --basedir=path-bOption-File Format basedirSystem Variable Name basedirVariable Scope Global Dynamic Variable No Permitted Values Type file nameThe path to the MySQL installation directory. All paths are usually resolved relative to this directory.
-
Command-Line Format --big-tablesOption-File Format big-tablesSystem Variable Name big_tablesVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type booleanEnable large result sets by saving all temporary sets in files. This option prevents most “table full” errors, but also slows down queries for which in-memory tables would suffice. Since MySQL 3.23.2, the server is able to handle large result sets automatically by using memory for small temporary tables and switching to disk tables where necessary.
-
Command-Line Format --bind-address=addrOption-File Format bind-addressSystem Variable Name bind_addressVariable Scope Global Dynamic Variable No Permitted Values Type stringDefault *The MySQL server listens on a single network socket for TCP/IP connections. This socket is bound to a single address, but it is possible for an address to map onto multiple network interfaces. To specify an address, use the
--bind-address=option at server startup, whereaddraddris an IPv4 or IPv6 address or a host name. Ifaddris a host name, the server resolves the name to an IP address and binds to that address.The server treats different types of addresses as follows:
If the address is
*, the server accepts TCP/IP connections on all server host IPv6 and IPv4 interfaces if the server host supports IPv6, or accepts TCP/IP connections on all IPv4 addresses otherwise. Use this address to permit both IPv4 and IPv6 connections on all server interfaces. This value is the default) in MySQL 5.7.If the address is
0.0.0.0, the server accepts TCP/IP connections on all server host IPv4 interfaces.If the address is
::, the server accepts TCP/IP connections on all server host IPv4 and IPv6 interfaces.If the address is an IPv4-mapped address, the server accepts TCP/IP connections for that address, in either IPv4 or IPv6 format. For example, if the server is bound to
::ffff:127.0.0.1, clients can connect using--host=127.0.0.1or--host=::ffff:127.0.0.1.If the address is a “regular” IPv4 or IPv6 address (such as
127.0.0.1or::1), the server accepts TCP/IP connections only for that IPv4 or IPv6 address.
If you intend to bind the server to a specific address, be sure that the
mysql.usergrant table contains an account with administrative privileges that you can use to connect to that address. Otherwise, you will not be able to shut down the server. For example, if you bind the server to*, you can connect to it using all existing accounts. But if you bind the server to::1, it accepts connections only on that address. In that case, first make sure that the'root'@'::1'account is present in themysql.usertable so you can still connect to the server to shut it down. --binlog-format={ROW|STATEMENT|MIXED}Command-Line Format --binlog-format=formatOption-File Format binlog-formatSystem Variable Name binlog_formatVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type enumerationDefault STATEMENTValid Values ROWSTATEMENTMIXEDSpecify whether to use row-based, statement-based, or mixed replication. Statement-based is the default in MySQL 5.7. See Section 16.1.2, “Replication Formats”.
Under some conditions, changing this variable at runtime is not possible, or causes replication to fail. See Section 5.2.4.2, “Setting The Binary Log Format”, for more information.
Setting the binary logging format without enabling binary logging sets the
binlog_formatglobal system variable and logs a warning.-
Command-Line Format --bootstrapOption-File Format bootstrapThis option is used by the mysql_install_db script to create the MySQL privilege tables without having to start a full MySQL server.
Replication and global transaction identifiers are automatically disabled whenever this option is used (Bug #1332602). See Section 16.1.3, “Replication with Global Transaction Identifiers”.
-
Command-Line Format --character-sets-dir=pathOption-File Format character-sets-dirSystem Variable Name character_sets_dirVariable Scope Global Dynamic Variable No Permitted Values Type directory nameThe directory where character sets are installed. See Section 10.5, “Character Set Configuration”.
--character-set-client-handshakeCommand-Line Format --character-set-client-handshakeOption-File Format character-set-client-handshakePermitted Values Type booleanDefault TRUEDo not ignore character set information sent by the client. To ignore client information and use the default server character set, use
--skip-character-set-client-handshake; this makes MySQL behave like MySQL 4.0.--character-set-filesystem=charset_nameCommand-Line Format --character-set-filesystem=nameOption-File Format character-set-filesystemSystem Variable Name character_set_filesystemVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type stringThe file system character set. This option sets the
character_set_filesystemsystem variable.--character-set-server=,charset_name-Ccharset_nameCommand-Line Format --character-set-serverOption-File Format character-set-serverSystem Variable Name character_set_serverVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type stringUse
charset_nameas the default server character set. See Section 10.5, “Character Set Configuration”. If you use this option to specify a nondefault character set, you should also use--collation-serverto specify the collation.--chroot=,path-rpathCommand-Line Format --chroot=name-r nameOption-File Format chrootPermitted Values Type file namePut the mysqld server in a closed environment during startup by using the
chroot()system call. This is a recommended security measure. Note that use of this option somewhat limitsLOAD DATA INFILEandSELECT ... INTO OUTFILE.--collation-server=collation_nameCommand-Line Format --collation-serverOption-File Format collation-serverSystem Variable Name collation_serverVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type stringUse
collation_nameas the default server collation. See Section 10.5, “Character Set Configuration”.-
Command-Line Format --consoleOption-File Format consolePlatform Specific windows (Windows only.) Write error log messages to
stderrandstdouteven if--log-erroris specified. mysqld does not close the console window if this option is used.If both
--log-errorand--consoleare specified,--consoletakes precedence. The server writes to the console, but not to the log file. (In MySQL 5.5 and 5.6, the precedence is reversed:--log-errorcauses--consoleto be ignored.) -
Command-Line Format --core-fileOption-File Format core-filePermitted Values Type booleanDefault OFFWrite a core file if mysqld dies. The name and location of the core file is system dependent. On Linux, a core file named
core.is written to the current working directory of the process, which for mysqld is the data directory.pidpidrepresents the process ID of the server process. On Mac OS X, a core file namedcore.is written to thepid/coresdirectory. On Solaris, use the coreadm command to specify where to write the core file and how to name it.For some systems, to get a core file you must also specify the
--core-file-sizeoption to mysqld_safe. See Section 4.3.2, “mysqld_safe — MySQL Server Startup Script”. On some systems, such as Solaris, you do not get a core file if you are also using the--useroption. There might be additional restrictions or limitations. For example, it might be necessary to execute ulimit -c unlimited before starting the server. Consult your system documentation. --datadir=,path-hpathCommand-Line Format --datadir=path-hOption-File Format datadirSystem Variable Name datadirVariable Scope Global Dynamic Variable No Permitted Values Type file nameThe path to the data directory.
--debug[=,debug_options]-# [debug_options]Command-Line Format --debug[=debug_options]Option-File Format debugSystem Variable Name debugVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type stringDefault 'd:t:o,/tmp/mysqld.trace'If MySQL is configured with
-DWITH_DEBUG=1, you can use this option to get a trace file of what mysqld is doing. A typicaldebug_optionsstring is'd:t:o,. The default isfile_name''d:t:i:o,mysqld.trace'.Using
-DWITH_DEBUG=1to configure MySQL with debugging support enables you to use the--debug="d,parser_debug"option when you start the server. This causes the Bison parser that is used to process SQL statements to dump a parser trace to the server's standard error output. Typically, this output is written to the error log.This option may be given multiple times. Values that begin with
+or-are added to or subtracted from the previous value. For example,--debug=T--debug=+Psets the value toP:T.For more information, see Section 22.4.3, “The DBUG Package”.
-
Command-Line Format --debug-sync-timeout[=#]Option-File Format debug-sync-timeoutPermitted Values Type numericControls whether the Debug Sync facility for testing and debugging is enabled. Use of Debug Sync requires that MySQL be configured with the
-DENABLE_DEBUG_SYNC=1option (see Section 2.9.4, “MySQL Source-Configuration Options”). If Debug Sync is not compiled in, this option is not available. The option value is a timeout in seconds. The default value is 0, which disables Debug Sync. To enable it, specify a value greater than 0; this value also becomes the default timeout for individual synchronization points. If the option is given without a value, the timeout is set to 300 seconds.For a description of the Debug Sync facility and how to use synchronization points, see MySQL Internals: Test Synchronization.
--default-authentication-plugin=plugin_nameCommand-Line Format --default-authentication-plugin=plugin_nameOption-File Format default-authentication-pluginPermitted Values Type enumerationDefault mysql_native_passwordValid Values mysql_native_passwordsha256_passwordThis option sets the default authentication plugin. Permitted values are
mysql_native_password(use MySQL native passwords) andsha256_password(use SHA-256 passwords). For more information about these plugins, see Section 6.3.7.2, “The Native Authentication Plugin”, and Section 6.3.7.4, “The SHA-256 Authentication Plugin”.NoteIf you use this option to change the default authentication plugin to a value other than
mysql_native_password, clients older than MySQL 5.5.6 will no longer be able to connect because they will not understand the resulting change to the authentication protocol.The
--default-authentication-pluginvalue affects these aspects of server operation:It determines which authentication plugin the server assigns to new accounts created by
CREATE USERandGRANTstatements that do not name a plugin explicitly with anIDENTIFIED WITHclause.It sets the
old_passwordssystem variable at startup to the value that is consistent with the password hashing method required by the default plugin. Theold_passwordsvalue affects hashing of passwords specified in theIDENTIFIED BYclause ofCREATE USERandGRANT, and passwords specified as the argument to thePASSWORD()function.For an account created with either of the following statements, the server associates the account with the default authentication plugin and assigns the account the given password, hashed according to the value of
old_passwords.CREATE USER ... IDENTIFIED BY '
cleartext password'; GRANT ... IDENTIFIED BY 'cleartext password';For an account created with either of the following statements, the statement fails if the password hash is not encrypted using the hash format required by the default authentication plugin. Otherwise, the server associates the account with the default authentication plugin and assigns the account the given password hash.
CREATE USER ... IDENTIFIED BY PASSWORD '
encrypted password'; GRANT ... IDENTIFIED BY PASSWORD 'encrypted password';
-
Command-Line Format --default-storage-engine=nameOption-File Format default-storage-engineSystem Variable Name default_storage_engineVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type enumerationDefault InnoDBSet the default storage engine for tables. See Chapter 14, Storage Engines. This option sets the storage engine for permanent tables only. To set the storage engine for
TEMPORARYtables, set thedefault_tmp_storage_enginesystem variable.If you disable the default storage engine at server startup, you must set the default engine for both permanent and
TEMPORARYtables to a different engine or the server will not start. -
Command-Line Format --default-time-zone=nameOption-File Format default-time-zonePermitted Values Type stringSet the default server time zone. This option sets the global
time_zonesystem variable. If this option is not given, the default time zone is the same as the system time zone (given by the value of thesystem_time_zonesystem variable. --delay-key-write[={OFF|ON|ALL}]Command-Line Format --delay-key-write[=name]Option-File Format delay-key-writeSystem Variable Name delay_key_writeVariable Scope Global Dynamic Variable Yes Permitted Values Type enumerationDefault ONValid Values ONOFFALLSpecify how to use delayed key writes. Delayed key writing causes key buffers not to be flushed between writes for
MyISAMtables.OFFdisables delayed key writes.ONenables delayed key writes for those tables that were created with theDELAY_KEY_WRITEoption.ALLdelays key writes for allMyISAMtables. See Section 8.11.2, “Tuning Server Parameters”, and Section 14.3.1, “MyISAMStartup Options”.NoteIf you set this variable to
ALL, you should not useMyISAMtables from within another program (such as another MySQL server or myisamchk) when the tables are in use. Doing so leads to index corruption.-
Command-Line Format --des-key-file=file_nameOption-File Format des-key-fileRead the default DES keys from this file. These keys are used by the
DES_ENCRYPT()andDES_DECRYPT()functions. -
Command-Line Format --enable-named-pipeOption-File Format enable-named-pipePlatform Specific windows Enable support for named pipes. This option applies only on Windows.
-
Command-Line Format --event-scheduler[=value]Option-File Format event-schedulerSystem Variable Name event_schedulerVariable Scope Global Dynamic Variable Yes Permitted Values Type enumerationDefault OFFValid Values ONOFFDISABLEDEnable or disable, and start or stop, the event scheduler.
For detailed information, see The
--event-schedulerOption. --exit-info[=,flags]-T [flags]Command-Line Format --exit-info[=flags]-T [flags]Option-File Format exit-infoPermitted Values Type numericThis is a bit mask of different flags that you can use for debugging the mysqld server. Do not use this option unless you know exactly what it does!
-
Command-Line Format --external-lockingOption-File Format external-lockingPermitted Values Type booleanDefault FALSEEnable external locking (system locking), which is disabled by default as of MySQL 4.0. Note that if you use this option on a system on which
lockddoes not fully work (such as Linux), it is easy for mysqld to deadlock.To disable external locking explicitly, use
--skip-external-locking.External locking affects only
MyISAMtable access. For more information, including conditions under which it can and cannot be used, see Section 8.10.5, “External Locking”. -
Command-Line Format --flushOption-File Format flushSystem Variable Name flushVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanDefault OFFFlush (synchronize) all changes to disk after each SQL statement. Normally, MySQL does a write of all changes to disk only after each SQL statement and lets the operating system handle the synchronizing to disk. See Section C.5.4.2, “What to Do If MySQL Keeps Crashing”.
-
Command-Line Format --gdbOption-File Format gdbPermitted Values Type booleanDefault FALSEInstall an interrupt handler for
SIGINT(needed to stop mysqld with^Cto set breakpoints) and disable stack tracing and core file handling. See Section 22.4, “Debugging and Porting MySQL”. -
Command-Line Format --general-logOption-File Format general-logSystem Variable Name general_logVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanDefault OFFSpecify the initial general query log state. With no argument or an argument of 1, the
--general-logoption enables the log. If omitted or given with an argument of 0, the option disables the log. -
Command-Line Format --ignore-db-dirOption-File Format ignore-db-dirPermitted Values Type directory nameThis option tells the server to ignore the given directory name for purposes of the
SHOW DATABASESstatement orINFORMATION_SCHEMAtables. For example, if a MySQL configuration locates the data directory at the root of a file system on Unix, the system might create alost+founddirectory there that the server should ignore. Starting the server with--ignore-db-dir=lost+foundcauses that name not to be listed as a database.To specify more than one name, use this option multiple times, once for each name. Specifying the option with an empty value (that is, as
--ignore-db-dir=) resets the directory list to the empty list.Instances of this option given at server startup are used to set the
ignore_db_dirssystem variable. -
Command-Line Format --init-file=file_nameOption-File Format init-fileSystem Variable Name init_fileVariable Scope Global Dynamic Variable No Permitted Values Type file nameRead SQL statements from this file at startup. Each statement must be on a single line and should not include comments.
--innodb-xxxSet an option for the
InnoDBstorage engine. TheInnoDBoptions are listed in Section 14.2.6, “InnoDBStartup Options and System Variables”.-
Command-Line Format --install [service_name](Windows only) Install the server as a Windows service that starts automatically during Windows startup. The default service name is
MySQLif noservice_namevalue is given. For more information, see Section 2.3.5.7, “Starting MySQL as a Windows Service”.NoteIf the server is started with the
--defaults-fileand--installoptions,--installmust be first. --install-manual [service_name]Command-Line Format --install-manual [service_name](Windows only) Install the server as a Windows service that must be started manually. It does not start automatically during Windows startup. The default service name is
MySQLif noservice_namevalue is given. For more information, see Section 2.3.5.7, “Starting MySQL as a Windows Service”.NoteIf the server is started with the
--defaults-fileand--install-manualoptions,--install-manualmust be first.--language=lang_name, -Llang_nameDeprecated 5.6.1, by lc-messages-dir Command-Line Format --language=name-LOption-File Format languageSystem Variable Name languageVariable Scope Global Dynamic Variable No Permitted Values Type directory nameDefault /usr/local/mysql/share/mysql/english/The language to use for error messages.
lang_namecan be given as the language name or as the full path name to the directory where the language files are installed. See Section 10.2, “Setting the Error Message Language”.In MySQL 5.7,
--lc-messages-dirand--lc-messagesshould be used rather than--language, which is deprecated (and handled as an alias for--lc-messages-dir). The--languageoption will be removed in a future MySQL release.-
Command-Line Format --large-pagesOption-File Format large-pagesSystem Variable Name large_pagesVariable Scope Global Dynamic Variable No Platform Specific linux Permitted Values Type (linux) booleanDefault FALSESome hardware/operating system architectures support memory pages greater than the default (usually 4KB). The actual implementation of this support depends on the underlying hardware and operating system. Applications that perform a lot of memory accesses may obtain performance improvements by using large pages due to reduced Translation Lookaside Buffer (TLB) misses.
MySQL 5.7 supports the Linux implementation of large page support (which is called HugeTLB in Linux). See Section 8.11.4.2, “Enabling Large Page Support”. For Solaris support of large pages, see the description of the
--super-large-pagesoption.--large-pagesis disabled by default. -
Command-Line Format --lc-messages=nameOption-File Format lc-messagesSystem Variable Name lc_messagesVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type stringThe locale to use for error messages. The server converts the argument to a language name and combines it with the value of the
--lc-messages-dirto produce the location for the error message file. See Section 10.2, “Setting the Error Message Language”. -
Command-Line Format --lc-messages-dir=pathOption-File Format lc-messages-dirSystem Variable Name lc_messages_dirVariable Scope Global Dynamic Variable No Permitted Values Type directory nameThe directory where error messages are located. The value is used together with the value of
--lc-messagesto produce the location for the error message file. See Section 10.2, “Setting the Error Message Language”. -
Command-Line Format --log-error[=name]Option-File Format log-errorSystem Variable Name log_errorVariable Scope Global Dynamic Variable No Permitted Values Type file nameLog errors and startup messages to this file. See Section 5.2.2, “The Error Log”. If you omit the file name, MySQL uses
host_name.err.err. -
Command-Line Format --log-isam[=name]Option-File Format log-isamPermitted Values Type file nameLog all
MyISAMchanges to this file (used only when debuggingMyISAM). -
Command-Line Format --log-output=nameOption-File Format log-outputSystem Variable Name log_outputVariable Scope Global Dynamic Variable Yes Permitted Values Type setDefault FILEValid Values TABLEFILENONEThis option determines the destination for general query log and slow query log output. The option value can be given as one or more of the words
TABLE,FILE, orNONE.TABLEselect logging to thegeneral_logandslow_logtables in themysqldatabase as a destination.FILEselects logging to log files as a destination.NONEdisables logging. IfNONEis present in the option value, it takes precedence over any other words that are present.TABLEandFILEcan both be given to select to both log output destinations.This option selects log output destinations, but does not enable log output. To do that, use the
--general_logand--slow_query_logoptions. ForFILElogging, the--general_log_fileand-slow_query_log_fileoptions determine the log file location. For more information, see Section 5.2.1, “Selecting General Query and Slow Query Log Output Destinations”. --log-queries-not-using-indexesCommand-Line Format --log-queries-not-using-indexesOption-File Format log-queries-not-using-indexesSystem Variable Name log_queries_not_using_indexesVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanDefault OFFIf you are using this option with the slow query log enabled, queries that are expected to retrieve all rows are logged. See Section 5.2.5, “The Slow Query Log”. This option does not necessarily mean that no index is used. For example, a query that uses a full index scan uses an index but would be logged because the index would not limit the number of rows.
-
Command-Line Format --log-raw[=value]Option-File Format log-rawPermitted Values Type booleanDefault OFFIn MySQL 5.7, passwords in certain statements written to the general query log, slow query log, and binary log are rewritten by the server not to occur literally in plain text. Password rewriting can be suppressed for the general query log by starting the server with the
--log-rawoption. This option may be useful for diagnostic purposes, to see the exact text of statements as received by the server, but for security reasons is not recommended for production use.For more information, see Section 6.1.2.3, “Passwords and Logging”.
-
Command-Line Format --log-short-formatOption-File Format log-short-formatPermitted Values Type booleanDefault FALSELog less information to the binary log and slow query log, if they have been activated.
-
Removed 5.7.1 Command-Line Format --log-slow-admin-statementsthrough 5.7.0 Option-File Format log-slow-admin-statementsPermitted Values Type booleanDefault OFFInclude slow administrative statements in the statements written to the slow query log. Administrative statements include
ALTER TABLE,ANALYZE TABLE,CHECK TABLE,CREATE INDEX,DROP INDEX,OPTIMIZE TABLE, andREPAIR TABLE.This command-line option was removed in MySQL 5.7.1 and replaced by the
log_slow_admin_statementssystem variable. The system variable can be set on the command line or in option files the same way as the option, so there is no need for any changes at server startup, but the system variable also makes it possible to examine or set the value at runtime. -
Command-Line Format --log-tc=nameOption-File Format log-tcPermitted Values Type file nameDefault tc.logThe name of the memory-mapped transaction coordinator log file (for XA transactions that affect multiple storage engines when the binary log is disabled). The default name is
tc.log. The file is created under the data directory if not given as a full path name. Currently, this option is unused. -
Command-Line Format --log-tc-size=#Option-File Format log-tc-sizePermitted Values Platform Bit Size 32Type numericDefault 24576Max Value 4294967295Permitted Values Platform Bit Size 64Type numericDefault 24576Max Value 18446744073709547520The size in bytes of the memory-mapped transaction coordinator log. The default size is 24KB.
--log-warnings[=,level]-W [level]Command-Line Format --log-warnings[=#]-W [#]Option-File Format log-warnings[=#]System Variable Name log_warningsVariable Scope Global Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 1Range 0 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 1Range 0 .. 18446744073709547520Print out warnings such as
Aborted connection...to the error log. This option is enabled (1) by default. To disable it, use--log-warnings=0. Specifying the option without alevelvalue increments the current value by 1. Enabling this option by setting it greater than 0 is recommended, for example, if you use replication (you get more information about what is happening, such as messages about network failures and reconnections). If the value is greater than 1, aborted connections are written to the error log, and access-denied errors for new connection attempts are written. See Section C.5.2.11, “Communication Errors and Aborted Connections”.If a slave server was started with
--log-warningsenabled, the slave prints messages to the error log to provide information about its status, such as the binary log and relay log coordinates where it starts its job, when it is switching to another relay log, when it reconnects after a disconnect, and so forth. The server logs messages about statements that are unsafe for statement-based logging if--log-warningsis greater than 0.-
Command-Line Format --low-priority-updatesOption-File Format low-priority-updatesSystem Variable Name low_priority_updatesVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type booleanDefault FALSEGive table-modifying operations (
INSERT,REPLACE,DELETE,UPDATE) lower priority than selects. This can also be done using{INSERT | REPLACE | DELETE | UPDATE} LOW_PRIORITY ...to lower the priority of only one query, or bySET LOW_PRIORITY_UPDATES=1to change the priority in one thread. This affects only storage engines that use only table-level locking (MyISAM,MEMORY,MERGE). See Section 8.10.2, “Table Locking Issues”. --min-examined-row-limit=numberCommand-Line Format --min-examined-row-limit=#Option-File Format min-examined-row-limitSystem Variable Name min_examined_row_limitVariable Scope Global, Session Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 0Range 0 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 0Range 0 .. 18446744073709547520When this option is set, queries which examine fewer than
numberrows are not written to the slow query log. The default is 0.-
Command-Line Format --memlockOption-File Format memlockSystem Variable Name locked_in_memoryVariable Scope Global Dynamic Variable No Permitted Values Type booleanDefault FALSELock the mysqld process in memory. This option might help if you have a problem where the operating system is causing mysqld to swap to disk.
--memlockworks on systems that support themlockall()system call; this includes Solaris, most Linux distributions that use a 2.4 or newer kernel, and perhaps other Unix systems. On Linux systems, you can tell whether or notmlockall()(and thus this option) is supported by checking to see whether or not it is defined in the systemmman.hfile, like this:shell>
grep mlockall /usr/include/sys/mman.hIf
mlockall()is supported, you should see in the output of the previous command something like the following:extern int mlockall (int __flags) __THROW;
ImportantUse of this option may require you to run the server as
root, which, for reasons of security, is normally not a good idea. See Section 6.1.5, “How to Run MySQL as a Normal User”.On Linux and perhaps other systems, you can avoid the need to run the server as
rootby changing thelimits.conffile. See the notes regarding the memlock limit in Section 8.11.4.2, “Enabling Large Page Support”.You must not try to use this option on a system that does not support the
mlockall()system call; if you do so, mysqld will very likely crash as soon as you try to start it. -
Command-Line Format --myisam-block-size=#Option-File Format myisam-block-sizePermitted Values Type numericDefault 1024Range 1024 .. 16384The block size to be used for
MyISAMindex pages. --myisam-recover-options[=option[,option]...]]Command-Line Format --myisam-recover-options[=name]Option-File Format myisam-recover-optionsPermitted Values Type enumerationDefault OFFValid Values OFFDEFAULTBACKUPFORCEQUICKSet the
MyISAMstorage engine recovery mode. The option value is any combination of the values ofOFF,DEFAULT,BACKUP,FORCE, orQUICK. If you specify multiple values, separate them by commas. Specifying the option with no argument is the same as specifyingDEFAULT, and specifying with an explicit value of""disables recovery (same as a value ofOFF). If recovery is enabled, each time mysqld opens aMyISAMtable, it checks whether the table is marked as crashed or was not closed properly. (The last option works only if you are running with external locking disabled.) If this is the case, mysqld runs a check on the table. If the table was corrupted, mysqld attempts to repair it.The following options affect how the repair works.
Option Description OFFNo recovery. DEFAULTRecovery without backup, forcing, or quick checking. BACKUPIf the data file was changed during recovery, save a backup of the tbl_name.MYDtbl_name-datetime.BAKFORCERun recovery even if we would lose more than one row from the .MYDfile.QUICKDo not check the rows in the table if there are not any delete blocks. Before the server automatically repairs a table, it writes a note about the repair to the error log. If you want to be able to recover from most problems without user intervention, you should use the options
BACKUP,FORCE. This forces a repair of a table even if some rows would be deleted, but it keeps the old data file as a backup so that you can later examine what happened.-
Command-Line Format --old-alter-tableOption-File Format old-alter-tableSystem Variable Name old_alter_tableVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type booleanDefault OFFWhen this option is given, the server does not use the optimized method of processing an
ALTER TABLEoperation. It reverts to using a temporary table, copying over the data, and then renaming the temporary table to the original, as used by MySQL 5.0 and earlier. For more information on the operation ofALTER TABLE, see Section 13.1.6, “ALTER TABLESyntax”. -
Command-Line Format --old-style-user-limitsOption-File Format old-style-user-limitsPermitted Values Type booleanDefault FALSEEnable old-style user limits. (Before MySQL 5.0.3, account resource limits were counted separately for each host from which a user connected rather than per account row in the
usertable.) See Section 6.3.4, “Setting Account Resource Limits”. -
Command-Line Format --open-files-limit=#Option-File Format open-files-limitSystem Variable Name open_files_limitVariable Scope Global Dynamic Variable No Permitted Values Type numericDefault -1 (autosized)Range 0 .. 65535Changes the number of file descriptors available to mysqld. You should try increasing the value of this option if mysqld gives you the error
Too many open files. mysqld uses the option value to reserve descriptors withsetrlimit(). If the requested number of file descriptors cannot be allocated, mysqld writes a warning to the error log.mysqld may attempt to allocate more than the requested number of descriptors (if they are available), using the values of
max_connectionsandtable_open_cacheto estimate whether more descriptors will be needed.On Unix, the value cannot be set less than ulimit -n.
-
Command-Line Format --partitionOption-File Format partitionDisabled by skip-partitionPermitted Values Type booleanDefault ONEnables or disables user-defined partitioning support in the MySQL Server.
--performance-schema-xxxConfigure a Performance Schema option. For details, see Section 20.11, “Performance Schema Command Options”.
-
Command-Line Format --pid-file=file_nameOption-File Format pid-fileSystem Variable Name pid_fileVariable Scope Global Dynamic Variable No Permitted Values Type file nameThe path name of the process ID file. The server creates the file in the data directory unless an absolute path name is given to specify a different directory. This file is used by other programs such as mysqld_safe to determine the server's process ID.
Specifies an option that pertains to a server plugin. For example, many storage engines can be built as plugins, and for such engines, options for them can be specified with a
--pluginprefix. Thus, the--innodb_file_per_tableoption forInnoDBcan be specified as--plugin-innodb_file_per_table.For boolean options that can be enabled or disabled, the
--skipprefix and other alternative formats are supported as well (see Section 4.2.3.2, “Program Option Modifiers”). For example,--skip-plugin-innodb_file_per_tabledisablesinnodb_file_per_table.The rationale for the
--pluginprefix is that it enables plugin options to be specified unambiguously if there is a name conflict with a built-in server option. For example, were a plugin writer to name a plugin “sql” and implement a “mode” option, the option name might be--sql-mode, which would conflict with the built-in option of the same name. In such cases, references to the conflicting name are resolved in favor of the built-in option. To avoid the ambiguity, users can specify the plugin option as--plugin-sql-mode. Use of the--pluginprefix for plugin options is recommended to avoid any question of ambiguity.-
Command-Line Format --plugin-load=plugin_listOption-File Format plugin-loadPermitted Values Type stringThis option tells the server to load the named plugins at startup. The option value is a semicolon-separated list of
name=plugin_librarynameis the name of the plugin, andplugin_libraryis the name of the shared library that contains the plugin code. Each library file must be located in the directory named by theplugin_dirsystem variable. For example, if plugins namedmyplug1andmyplug2have library filesmyplug1.soandmyplug2.so, use this option to load them at startup:shell>
mysqld --plugin-load="myplug1=myplug1.so;myplug2=myplug2.so"Quotes are used around the argument value here because semicolon (
;) is interpreted as a special character by some command interpreters. (Unix shells treat it as a command terminator, for example.)If multiple
--plugin-loadoptions are given, only the last one is used. Additional plugins to load may be specified using--plugin-load-addoptions.If a plugin library is named without any preceding plugin name, the server loads all plugins in the library.
Each plugin is loaded for a single invocation of mysqld only. After a restart, the plugin is not loaded unless
--plugin-loadis used again. This is in contrast toINSTALL PLUGIN, which adds an entry to themysql.pluginstable to cause the plugin to be loaded for every normal server startup.Under normal startup, the server determines which plugins to load by reading the
mysql.pluginssystem table. If the server is started with the--skip-grant-tablesoption, it does not consult themysql.pluginstable and does not load plugins listed there.--plugin-loadenables plugins to be loaded even when--skip-grant-tablesis given.--plugin-loadalso enables plugins to be loaded at startup under configurations when plugins cannot be loaded at runtime.For additional information about plugin loading, see Section 5.1.8.1, “Installing and Uninstalling Plugins”.
-
Command-Line Format --plugin-load-add=plugin_listOption-File Format plugin-load-addPermitted Values Type stringThis option complements the
--plugin-loadoption.--plugin-load-addadds a plugin or plugins to the set of plugins to be loaded at startup. The argument format is the same as for--plugin-load.--plugin-load-addcan be used to avoid specifying a large set of plugins as a single long unwieldy--plugin-loadargument.--plugin-load-addcan be given in the absence of--plugin-load, but any instance of--plugin-load-addthat appears before--plugin-load. has no effect because--plugin-loadresets the set of plugins to load. In other words, these options:--plugin-load=x --plugin-load-add=y
are equivalent to this option:
--plugin-load="x;y"
But these options:
--plugin-load-add=y --plugin-load=x
are equivalent to this option:
--plugin-load=x
For additional information about plugin loading, see Section 5.1.8.1, “Installing and Uninstalling Plugins”.
--port=,port_num-Pport_numCommand-Line Format --port=#-POption-File Format portSystem Variable Name portVariable Scope Global Dynamic Variable No Permitted Values Type numericDefault 3306Range 0 .. 65535The port number to use when listening for TCP/IP connections. The port number must be 1024 or higher unless the server is started by the
rootsystem user.-
Command-Line Format --port-open-timeout=#Option-File Format port-open-timeoutPermitted Values Type numericDefault 0On some systems, when the server is stopped, the TCP/IP port might not become available immediately. If the server is restarted quickly afterward, its attempt to reopen the port can fail. This option indicates how many seconds the server should wait for the TCP/IP port to become free if it cannot be opened. The default is not to wait.
-
Command-Line Format --remove [service_name](Windows only) Remove a MySQL Windows service. The default service name is
MySQLif noservice_namevalue is given. For more information, see Section 2.3.5.7, “Starting MySQL as a Windows Service”. -
Command-Line Format --safe-user-createOption-File Format safe-user-createPermitted Values Type booleanDefault FALSEIf this option is enabled, a user cannot create new MySQL users by using the
GRANTstatement unless the user has theINSERTprivilege for themysql.usertable or any column in the table. If you want a user to have the ability to create new users that have those privileges that the user has the right to grant, you should grant the user the following privilege:GRANT INSERT(user) ON mysql.user TO '
user_name'@'host_name';This ensures that the user cannot change any privilege columns directly, but has to use the
GRANTstatement to give privileges to other users. -
Command-Line Format --secure-authOption-File Format secure-authSystem Variable Name secure_authVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanDefault ONThis option causes the server to block connections by clients that attempt to use accounts that have passwords stored in the old (pre-4.1) format. Use it to prevent all use of passwords employing the old format (and hence insecure communication over the network). This option is enabled by default; to disable it, use
--skip-secure-auth.Server startup fails with an error if this option is enabled and the privilege tables are in pre-4.1 format. See Section C.5.2.4, “
Client does not support authentication protocol”.The mysql client also has a
--secure-authoption, which prevents connections to a server if the server requires a password in old format for the client account.NotePasswords that use the pre-4.1 hashing method are less secure than passwords that use the native password hashing method and should be avoided. Pre-4.1 passwords are deprecated and support for them will be removed in a future MySQL release.
-
Command-Line Format --secure-file-priv=pathOption-File Format secure-file-privSystem Variable Name secure_file_privVariable Scope Global Dynamic Variable No Permitted Values Type stringThis option limits the effect of the
LOAD_FILE()function and theLOAD DATAandSELECT ... INTO OUTFILEstatements to work only with files in the specified directory. -
System Variable Name shared_memoryVariable Scope Global Dynamic Variable No Platform Specific windows Enable shared-memory connections by local clients. This option is available only on Windows.
--shared-memory-base-name=nameSystem Variable Name shared_memory_base_nameVariable Scope Global Dynamic Variable No Platform Specific windows The name of shared memory to use for shared-memory connections. This option is available only on Windows. The default name is
MYSQL. The name is case sensitive.Turn off the ability to select and insert at the same time on
MyISAMtables. (This is to be used only if you think you have found a bug in this feature.) See Section 8.10.3, “Concurrent Inserts”.-
Command-Line Format --skip-event-scheduler--disable-event-schedulerOption-File Format skip-event-schedulerTurns the Event Scheduler
OFF. This is not the same as disabling the Event Scheduler, which requires setting--event-scheduler=DISABLED; see The--event-schedulerOption, for more information. This option causes the server to start without using the privilege system at all, which gives anyone with access to the server unrestricted access to all databases. You can cause a running server to start using the grant tables again by executing mysqladmin flush-privileges or mysqladmin reload command from a system shell, or by issuing a MySQL
FLUSH PRIVILEGESstatement after connecting to the server. This option also suppresses loading of plugins that were installed with theINSTALL PLUGINstatement, user-defined functions (UDFs), and scheduled events. To cause plugins to be loaded anyway, use the--plugin-loadoption.Note that
FLUSH PRIVILEGESmight be executed implicitly by other actions performed after startup. For example, mysql_upgrade flushes the privileges during the upgrade procedure.Disable use of the internal host cache for faster name-to-IP resolution. In this case, the server performs a DNS lookup every time a client connects. See Section 8.11.5.2, “DNS Lookup Optimization and the Host Cache”.
Use of
--skip-host-cacheis similar to setting thehost_cache_sizesystem variable to 0, buthost_cache_sizeis more flexible because it can also be used to resize, enable, or disable the host cache at runtime, not just at server startup.If you start the server with
--skip-host-cache, that does not prevent changes to the value ofhost_cache_size, but such changes have no effect and the cache is not re-enabled even ifhost_cache_sizeis set larger than 0.Disable the
InnoDBstorage engine. In this case, because the default storage engine isInnoDB, the server will not start unless you also use--default-storage-engineand--default-tmp-storage-engineto set the default to some other engine for both permanent andTEMPORARYtables.Do not resolve host names when checking client connections. Use only IP addresses. If you use this option, all
Hostcolumn values in the grant tables must be IP addresses orlocalhost. See Section 8.11.5.2, “DNS Lookup Optimization and the Host Cache”.Do not listen for TCP/IP connections at all. All interaction with mysqld must be made using named pipes or shared memory (on Windows) or Unix socket files (on Unix). This option is highly recommended for systems where only local clients are permitted. See Section 8.11.5.2, “DNS Lookup Optimization and the Host Cache”.
-
Command-Line Format --skip-partition--disable-partitionOption-File Format skip-partitionDisables user-defined partitioning. Partitioned tables can be seen using
SHOW TABLESor by querying theINFORMATION_SCHEMA.TABLEStable, but cannot be created or modified, nor can data in such tables be accessed. All partition-specific columns in theINFORMATION_SCHEMA.PARTITIONStable displayNULL.Since
DROP TABLEremoves table definition (.frm) files, this statement works on partitioned tables even when partitioning is disabled using the option. The statement, however, does not remove.parfiles associated with partitioned tables in such cases. For this reason, you should avoid dropping partitioned tables with partitioning disabled, or take action to remove the orphaned.parfiles manually. Options that begin with
--sslspecify whether to permit clients to connect using SSL and indicate where to find SSL keys and certificates. See Section 6.3.9.4, “SSL Command Options”.-
Command-Line Format --standaloneOption-File Format standalonePlatform Specific windows Available on Windows only; instructs the MySQL server not to run as a service.
-
Command-Line Format --super-large-pagesOption-File Format super-large-pagesPlatform Specific solaris Permitted Values Type (solaris) booleanDefault FALSEStandard use of large pages in MySQL attempts to use the largest size supported, up to 4MB. Under Solaris, a “super large pages” feature enables uses of pages up to 256MB. This feature is available for recent SPARC platforms. It can be enabled or disabled by using the
--super-large-pagesor--skip-super-large-pagesoption. --symbolic-links,--skip-symbolic-linksCommand-Line Format --symbolic-linksOption-File Format symbolic-linksEnable or disable symbolic link support. On Unix, enabling symbolic links means that you can link a
MyISAMindex file or data file to another directory with theINDEX DIRECTORYorDATA DIRECTORYoptions of theCREATE TABLEstatement. If you delete or rename the table, the files that its symbolic links point to also are deleted or renamed. See Section 8.11.3.1.2, “Using Symbolic Links forMyISAMTables on Unix”.This option has no meaning on Windows.
-
Command-Line Format --skip-show-databaseOption-File Format skip-show-databaseSystem Variable Name skip_show_databaseVariable Scope Global Dynamic Variable No This option sets the
skip_show_databasesystem variable that controls who is permitted to use theSHOW DATABASESstatement. See Section 5.1.4, “Server System Variables”. -
Command-Line Format --skip-stack-traceOption-File Format skip-stack-traceDo not write stack traces. This option is useful when you are running mysqld under a debugger. On some systems, you also must use this option to get a core file. See Section 22.4, “Debugging and Porting MySQL”.
-
Command-Line Format --slow-query-logOption-File Format slow-query-logSystem Variable Name slow_query_logVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanDefault OFFSpecify the initial slow query log state. With no argument or an argument of 1, the
--slow-query-logoption enables the log. If omitted or given with an argument of 0, the option disables the log. -
Command-Line Format --slow-start-timeout=#Option-File Format slow-start-timeoutPermitted Values Type (windows) numericDefault 15000This option controls the Windows service control manager's service start timeout. The value is the maximum number of milliseconds that the service control manager waits before trying to kill the windows service during startup. The default value is 15000 (15 seconds). If the MySQL service takes too long to start, you may need to increase this value. A value of 0 means there is no timeout.
-
Command-Line Format --socket=nameOption-File Format socketSystem Variable Name socketVariable Scope Global Dynamic Variable No Permitted Values Type file nameDefault /tmp/mysql.sockOn Unix, this option specifies the Unix socket file to use when listening for local connections. The default value is
/tmp/mysql.sock. If this option is given, the server creates the file in the data directory unless an absolute path name is given to specify a different directory. On Windows, the option specifies the pipe name to use when listening for local connections that use a named pipe. The default value isMySQL(not case sensitive). --sql-mode=value[,value[,value...]]Command-Line Format --sql-mode=nameOption-File Format sql-modeSystem Variable Name sql_modeVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type setDefault NO_ENGINE_SUBSTITUTIONValid Values ALLOW_INVALID_DATESANSI_QUOTESERROR_FOR_DIVISION_BY_ZEROHIGH_NOT_PRECEDENCEIGNORE_SPACENO_AUTO_CREATE_USERNO_AUTO_VALUE_ON_ZERONO_BACKSLASH_ESCAPESNO_DIR_IN_CREATENO_ENGINE_SUBSTITUTIONNO_FIELD_OPTIONSNO_KEY_OPTIONSNO_TABLE_OPTIONSNO_UNSIGNED_SUBTRACTIONNO_ZERO_DATENO_ZERO_IN_DATEONLY_FULL_GROUP_BYPAD_CHAR_TO_FULL_LENGTHPIPES_AS_CONCATREAL_AS_FLOATSTRICT_ALL_TABLESSTRICT_TRANS_TABLESSet the SQL mode. See Section 5.1.7, “Server SQL Modes”.
NoteMySQL installation programs may configure the SQL mode during the installation process. For example, mysql_install_db creates a default option file named
my.cnfin the base installation directory. This file contains a line that sets the SQL mode; see Section 4.4.3, “mysql_install_db — Initialize MySQL Data Directory”.If the SQL mode differs from the default or from what you expect, check for a setting in an option file that the server reads at startup.
-
Command-Line Format --sysdate-is-nowOption-File Format sysdate-is-nowPermitted Values Type booleanDefault FALSESYSDATE()by default returns the time at which it executes, not the time at which the statement in which it occurs begins executing. This differs from the behavior ofNOW(). This option causesSYSDATE()to be an alias forNOW(). For information about the implications for binary logging and replication, see the description forSYSDATE()in Section 12.7, “Date and Time Functions” and forSET TIMESTAMPin Section 5.1.4, “Server System Variables”. --tc-heuristic-recover={COMMIT|ROLLBACK}Command-Line Format --tc-heuristic-recover=nameOption-File Format tc-heuristic-recoverPermitted Values Type enumerationValid Values COMMITRECOVERThe type of decision to use in the heuristic recovery process. Currently, this option is unused.
-
Command-Line Format --temp-poolOption-File Format temp-poolPermitted Values Type booleanDefault TRUEThis option causes most temporary files created by the server to use a small set of names, rather than a unique name for each new file. This works around a problem in the Linux kernel dealing with creating many new files with different names. With the old behavior, Linux seems to “leak” memory, because it is being allocated to the directory entry cache rather than to the disk cache. This option is ignored except on Linux.
-
Command-Line Format --transaction-isolation=nameOption-File Format transaction-isolationPermitted Values Type enumerationValid Values READ-UNCOMMITTEDREAD-COMMITTEDREPEATABLE-READSERIALIZABLESets the default transaction isolation level. The
levelvalue can beREAD-UNCOMMITTED,READ-COMMITTED,REPEATABLE-READ, orSERIALIZABLE. See Section 13.3.6, “SET TRANSACTIONSyntax”.The default transaction isolation level can also be set at runtime using the
SET TRANSACTIONstatement or by setting thetx_isolationsystem variable. -
Command-Line Format --transaction-read-onlyOption-File Format transaction-read-onlyPermitted Values Type booleanDefault OFFSets the default transaction access mode. By default, read-only mode is disabled, so the mode is read/write.
To set the default transaction access mode at runtime, use the
SET TRANSACTIONstatement or set thetx_read_onlysystem variable. See Section 13.3.6, “SET TRANSACTIONSyntax”. --tmpdir=,path-tpathCommand-Line Format --tmpdir=path-tOption-File Format tmpdirSystem Variable Name tmpdirVariable Scope Global Dynamic Variable No Permitted Values Type file nameThe path of the directory to use for creating temporary files. It might be useful if your default
/tmpdirectory resides on a partition that is too small to hold temporary tables. This option accepts several paths that are used in round-robin fashion. Paths should be separated by colon characters (“:”) on Unix and semicolon characters (“;”) on Windows. If the MySQL server is acting as a replication slave, you should not set--tmpdirto point to a directory on a memory-based file system or to a directory that is cleared when the server host restarts. For more information about the storage location of temporary files, see Section C.5.4.4, “Where MySQL Stores Temporary Files”. A replication slave needs some of its temporary files to survive a machine restart so that it can replicate temporary tables orLOAD DATA INFILEoperations. If files in the temporary file directory are lost when the server restarts, replication fails.--user={,user_name|user_id}-u {user_name|user_id}Command-Line Format --user=name-u nameOption-File Format userPermitted Values Type stringRun the mysqld server as the user having the name
user_nameor the numeric user IDuser_id. (“User” in this context refers to a system login account, not a MySQL user listed in the grant tables.)This option is mandatory when starting mysqld as
root. The server changes its user ID during its startup sequence, causing it to run as that particular user rather than asroot. See Section 6.1.1, “Security Guidelines”.To avoid a possible security hole where a user adds a
--user=rootoption to amy.cnffile (thus causing the server to run asroot), mysqld uses only the first--useroption specified and produces a warning if there are multiple--useroptions. Options in/etc/my.cnfand$MYSQL_HOME/my.cnfare processed before command-line options, so it is recommended that you put a--useroption in/etc/my.cnfand specify a value other thanroot. The option in/etc/my.cnfis found before any other--useroptions, which ensures that the server runs as a user other thanroot, and that a warning results if any other--useroption is found.Use this option with the
--helpoption for detailed help.--version,-VDisplay version information and exit.
You can assign a value to a server system variable by using an
option of the form
--.
For example, var_name=value--key_buffer_size=32M
sets the key_buffer_size variable
to a value of 32MB.
Note that when you assign a value to a variable, MySQL might automatically correct the value to stay within a given range, or adjust the value to the closest permissible value if only certain values are permitted.
If you want to restrict the maximum value to which a variable can
be set at runtime with
SET, you
can define this by using the
--maximum-
command-line option.
var_name=value
You can change the values of most system variables for a running
server with the
SET
statement. See Section 13.7.4, “SET Syntax”.
Section 5.1.4, “Server System Variables”, provides a full description for all variables, and additional information for setting them at server startup and runtime. Section 8.11.2, “Tuning Server Parameters”, includes information on optimizing the server by tuning system variables.
The MySQL server maintains many system variables that indicate how
it is configured. Each system variable has a default value. System
variables can be set at server startup using options on the
command line or in an option file. Most of them can be changed
dynamically while the server is running by means of the
SET
statement, which enables you to modify operation of the server
without having to stop and restart it. You can refer to system
variable values in expressions.
There are several ways to see the names and values of system variables:
To see the values that a server will use based on its compiled-in defaults and any option files that it reads, use this command:
mysqld --verbose --help
To see the values that a server will use based on its compiled-in defaults, ignoring the settings in any option files, use this command:
mysqld --no-defaults --verbose --help
To see the current values used by a running server, use the
SHOW VARIABLESstatement.
This section provides a description of each system variable. Variables with no version indicated are present in all MySQL 5.7 releases. For historical information concerning their implementation, please see http://dev.mysql.com/doc/refman/5.0/en/, and http://dev.mysql.com/doc/refman/4.1/en/.
The following table lists all available system variables.
Table 5.2. System Variable Summary
For additional system variable information, see these sections:
Section 5.1.5, “Using System Variables”, discusses the syntax for setting and displaying system variable values.
Section 5.1.5.2, “Dynamic System Variables”, lists the variables that can be set at runtime.
Information on tuning system variables can be found in Section 8.11.2, “Tuning Server Parameters”.
Section 14.2.6, “
InnoDBStartup Options and System Variables”, listsInnoDBsystem variables.MySQL Cluster System Variables, lists system variables which are specific to MySQL Cluster.
For information on server system variables specific to replication, see Section 16.1.4, “Replication and Binary Logging Options and Variables”.
Some of the following variable descriptions refer to
“enabling” or “disabling” a variable.
These variables can be enabled with the
SET
statement by setting them to ON or
1, or disabled by setting them to
OFF or 0. In MySQL
5.7, boolean variables can be set at startup to the
values ON, TRUE,
OFF, and FALSE (not case
sensitive), as well as 1 and
0. See Section 4.2.3.2, “Program Option Modifiers”.
Some system variables control the size of buffers or caches. For a given buffer, the server might need to allocate internal data structures. These structures typically are allocated from the total memory allocated to the buffer, and the amount of space required might be platform dependent. This means that when you assign a value to a system variable that controls a buffer size, the amount of space actually available might differ from the value assigned. In some cases, the amount might be less than the value assigned. It is also possible that the server will adjust a value upward. For example, if you assign a value of 0 to a variable for which the minimal value is 1024, the server will set the value to 1024.
Values for buffer sizes, lengths, and stack sizes are given in bytes unless otherwise specified.
Some system variables take file name values. Unless otherwise
specified, the default file location is the data directory if the
value is a relative path name. To specify the location explicitly,
use an absolute path name. Suppose that the data directory is
/var/mysql/data. If a file-valued variable is
given as a relative path name, it will be located under
/var/mysql/data. If the value is an absolute
path name, its location is as given by the path name.
-
Command-Line Format --autocommit[=#]Option-File Format autocommitSystem Variable Name autocommitVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type booleanDefault ONThe autocommit mode. If set to 1, all changes to a table take effect immediately. If set to 0, you must use
COMMITto accept a transaction orROLLBACKto cancel it. Ifautocommitis 0 and you change it to 1, MySQL performs an automaticCOMMITof any open transaction. Another way to begin a transaction is to use aSTART TRANSACTIONorBEGINstatement. See Section 13.3.1, “START TRANSACTION,COMMIT, andROLLBACKSyntax”.By default, client connections begin with
autocommitset to 1. To cause clients to begin with a default of 0, set the globalautocommitvalue by starting the server with the--autocommit=0option. To set the variable using an option file, include these lines:[mysqld] autocommit=0
-
System Variable Name automatic_sp_privilegesVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanDefault TRUEWhen this variable has a value of 1 (the default), the server automatically grants the
EXECUTEandALTER ROUTINEprivileges to the creator of a stored routine, if the user cannot already execute and alter or drop the routine. (TheALTER ROUTINEprivilege is required to drop the routine.) The server also automatically drops those privileges from the creator when the routine is dropped. Ifautomatic_sp_privilegesis 0, the server does not automatically add or drop these privileges.The creator of a routine is the account used to execute the
CREATEstatement for it. This might not be the same as the account named as theDEFINERin the routine definition.See also Section 18.2.2, “Stored Routines and MySQL Privileges”.
-
System Variable Name back_logVariable Scope Global Dynamic Variable No Permitted Values Type numericDefault -1 (autosized)Range 1 .. 65535The number of outstanding connection requests MySQL can have. This comes into play when the main MySQL thread gets very many connection requests in a very short time. It then takes some time (although very little) for the main thread to check the connection and start a new thread. The
back_logvalue indicates how many requests can be stacked during this short time before MySQL momentarily stops answering new requests. You need to increase this only if you expect a large number of connections in a short period of time.In other words, this value is the size of the listen queue for incoming TCP/IP connections. Your operating system has its own limit on the size of this queue. The manual page for the Unix
listen()system call should have more details. Check your OS documentation for the maximum value for this variable.back_logcannot be set higher than your operating system limit.The default value is based on the following formula, capped to a limit of 900:
50 + (max_connections / 5)
-
Command-Line Format --basedir=path-bOption-File Format basedirSystem Variable Name basedirVariable Scope Global Dynamic Variable No Permitted Values Type file nameThe MySQL installation base directory. This variable can be set with the
--basediroption. Relative path names for other variables usually are resolved relative to the base directory. If set to 1, all temporary tables are stored on disk rather than in memory. This is a little slower, but the error
The tabledoes not occur fortbl_nameis fullSELECToperations that require a large temporary table. The default value for a new connection is 0 (use in-memory temporary tables). Normally, you should never need to set this variable, because in-memory tables are automatically converted to disk-based tables as required.-
Command-Line Format --bind-address=addrOption-File Format bind-addressSystem Variable Name bind_addressVariable Scope Global Dynamic Variable No Permitted Values Type stringDefault *The value of the
--bind-addressoption.This variable has no effect for the embedded server (
libmysqld) and as of MySQL 5.7.2 is no longer visible within the embedded server. -
Command-Line Format --bulk_insert_buffer_size=#Option-File Format bulk_insert_buffer_sizeSystem Variable Name bulk_insert_buffer_sizeVariable Scope Global, Session Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 8388608Range 0 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 8388608Range 0 .. 18446744073709547520MyISAMuses a special tree-like cache to make bulk inserts faster forINSERT ... SELECT,INSERT ... VALUES (...), (...), ..., andLOAD DATA INFILEwhen adding data to nonempty tables. This variable limits the size of the cache tree in bytes per thread. Setting it to 0 disables this optimization. The default value is 8MB. -
System Variable Name character_set_clientVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type stringThe character set for statements that arrive from the client. The session value of this variable is set using the character set requested by the client when the client connects to the server. (Many clients support a
--default-character-setoption to enable this character set to be specified explicitly. See also Section 10.1.4, “Connection Character Sets and Collations”.) The global value of the variable is used to set the session value in cases when the client-requested value is unknown or not available, or the server is configured to ignore client requests:The client is from a version of MySQL older than MySQL 4.1, and thus does not request a character set.
The client requests a character set not known to the server. For example, a Japanese-enabled client requests
sjiswhen connecting to a server not configured withsjissupport.mysqld was started with the
--skip-character-set-client-handshakeoption, which causes it to ignore client character set configuration. This reproduces MySQL 4.0 behavior and is useful should you wish to upgrade the server without upgrading all the clients.
ucs2,utf16,utf16le, andutf32cannot be used as a client character set, which means that they also do not work forSET NAMESorSET CHARACTER SET. -
System Variable Name character_set_connectionVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type stringThe character set used for literals that do not have a character set introducer and for number-to-string conversion.
-
System Variable Name character_set_databaseVariable Scope Global, Session Dynamic Variable Yes Footnote This option is dynamic, but only the server should set this information. You should not set the value of this variable manually. Permitted Values Type stringThe character set used by the default database. The server sets this variable whenever the default database changes. If there is no default database, the variable has the same value as
character_set_server. -
Command-Line Format --character-set-filesystem=nameOption-File Format character-set-filesystemSystem Variable Name character_set_filesystemVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type stringThe file system character set. This variable is used to interpret string literals that refer to file names, such as in the
LOAD DATA INFILEandSELECT ... INTO OUTFILEstatements and theLOAD_FILE()function. Such file names are converted fromcharacter_set_clienttocharacter_set_filesystembefore the file opening attempt occurs. The default value isbinary, which means that no conversion occurs. For systems on which multi-byte file names are permitted, a different value may be more appropriate. For example, if the system represents file names using UTF-8, setcharacter_set_filesystemto'utf8'. -
System Variable Name character_set_resultsVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type stringThe character set used for returning query results such as result sets or error messages to the client.
-
Command-Line Format --character-set-serverOption-File Format character-set-serverSystem Variable Name character_set_serverVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type stringThe server's default character set.
-
System Variable Name character_set_systemVariable Scope Global Dynamic Variable No Permitted Values Type stringThe character set used by the server for storing identifiers. The value is always
utf8. -
Command-Line Format --character-sets-dir=pathOption-File Format character-sets-dirSystem Variable Name character_sets_dirVariable Scope Global Dynamic Variable No Permitted Values Type directory nameThe directory where character sets are installed.
-
System Variable Name collation_connectionVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type stringThe collation of the connection character set.
-
System Variable Name collation_databaseVariable Scope Global, Session Dynamic Variable Yes Footnote This option is dynamic, but only the server should set this information. You should not set the value of this variable manually. Permitted Values Type stringThe collation used by the default database. The server sets this variable whenever the default database changes. If there is no default database, the variable has the same value as
collation_server. -
Command-Line Format --collation-serverOption-File Format collation-serverSystem Variable Name collation_serverVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type stringThe server's default collation.
-
Command-Line Format --completion_type=#Option-File Format completion_typeSystem Variable Name completion_typeVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type enumerationDefault NO_CHAINValid Values NO_CHAINCHAINRELEASE012The transaction completion type. This variable can take the values shown in the following table. The variable can be assigned using either the name values or corresponding integer values.
Value Description NO_CHAIN(or 0)COMMITandROLLBACKare unaffected. This is the default value.CHAIN(or 1)COMMITandROLLBACKare equivalent toCOMMIT AND CHAINandROLLBACK AND CHAIN, respectively. (A new transaction starts immediately with the same isolation level as the just-terminated transaction.)RELEASE(or 2)COMMITandROLLBACKare equivalent toCOMMIT RELEASEandROLLBACK RELEASE, respectively. (The server disconnects after terminating the transaction.)completion_typeaffects transactions that begin withSTART TRANSACTIONorBEGINand end withCOMMITorROLLBACK. It does not apply to implicit commits resulting from execution of the statements listed in Section 13.3.3, “Statements That Cause an Implicit Commit”. It also does not apply forXA COMMIT,XA ROLLBACK, or whenautocommit=1. -
Command-Line Format --concurrent_insert[=#]Option-File Format concurrent_insertSystem Variable Name concurrent_insertVariable Scope Global Dynamic Variable Yes Permitted Values Type enumerationDefault AUTOValid Values NEVERAUTOALWAYS012If
AUTO(the default), MySQL permitsINSERTandSELECTstatements to run concurrently forMyISAMtables that have no free blocks in the middle of the data file. If you start mysqld with--skip-new, this variable is set toNEVER.This variable can take the values shown in the following table. The variable can be assigned using either the name values or corresponding integer values.
Value Description NEVER(or 0)Disables concurrent inserts AUTO(or 1)(Default) Enables concurrent insert for MyISAMtables that do not have holesALWAYS(or 2)Enables concurrent inserts for all MyISAMtables, even those that have holes. For a table with a hole, new rows are inserted at the end of the table if it is in use by another thread. Otherwise, MySQL acquires a normal write lock and inserts the row into the hole.See also Section 8.10.3, “Concurrent Inserts”.
-
Command-Line Format --connect_timeout=#Option-File Format connect_timeoutSystem Variable Name connect_timeoutVariable Scope Global Dynamic Variable Yes Permitted Values Type numericDefault 10The number of seconds that the mysqld server waits for a connect packet before responding with
Bad handshake. The default value is 10 seconds.Increasing the
connect_timeoutvalue might help if clients frequently encounter errors of the formLost connection to MySQL server at '.XXX', system error:errno -
System Variable Name core_fileVariable Scope Global Dynamic Variable No Permitted Values Type booleanDefault OFFWhether to write a core file if the server crashes. This variable is set by the
--core-fileoption. -
Command-Line Format --datadir=path-hOption-File Format datadirSystem Variable Name datadirVariable Scope Global Dynamic Variable No Permitted Values Type file nameThe MySQL data directory. This variable can be set with the
--datadiroption. This variable is unused. It is deprecated and will be removed in a future MySQL release.
This variable is unused. It is deprecated and will be removed in a future MySQL release.
-
Command-Line Format --debug[=debug_options]Option-File Format debugSystem Variable Name debugVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type stringDefault 'd:t:o,/tmp/mysqld.trace'This variable indicates the current debugging settings. It is available only for servers built with debugging support. The initial value comes from the value of instances of the
--debugoption given at server startup. The global and session values may be set at runtime; theSUPERprivilege is required, even for the session value.Assigning a value that begins with
+or-cause the value to added to or subtracted from the current value:mysql>
SET debug = 'T';mysql>SELECT @@debug;+---------+ | @@debug | +---------+ | T | +---------+ mysql>SET debug = '+P';mysql>SELECT @@debug;+---------+ | @@debug | +---------+ | P:T | +---------+ mysql>SET debug = '-P';mysql>SELECT @@debug;+---------+ | @@debug | +---------+ | T | +---------+For more information, see Section 22.4.3, “The DBUG Package”.
-
System Variable Name debug_syncVariable Scope Session Dynamic Variable Yes Permitted Values Type stringThis variable is the user interface to the Debug Sync facility. Use of Debug Sync requires that MySQL be configured with the
-DENABLE_DEBUG_SYNC=1option (see Section 2.9.4, “MySQL Source-Configuration Options”). If Debug Sync is not compiled in, this system variable is not available.The global variable value is read only and indicates whether the facility is enabled. By default, Debug Sync is disabled and the value of
debug_syncisOFF. If the server is started with--debug-sync-timeout=, whereNNis a timeout value greater than 0, Debug Sync is enabled and the value ofdebug_syncisON - current signalfollowed by the signal name. Also,Nbecomes the default timeout for individual synchronization points.The session value can be read by any user and will have the same value as the global variable. The session value can be set by users that have the
SUPERprivilege to control synchronization points.For a description of the Debug Sync facility and how to use synchronization points, see MySQL Internals: Test Synchronization.
-
Command-Line Format --default-storage-engine=nameOption-File Format default-storage-engineSystem Variable Name default_storage_engineVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type enumerationDefault InnoDBThe default storage engine. This variable sets the storage engine for permanent tables only. To set the storage engine for
TEMPORARYtables, set thedefault_tmp_storage_enginesystem variable.default_storage_engineshould be used in preference tostorage_engine, which is deprecated.If you disable the default storage engine at server startup, you must set the default engine for both permanent and
TEMPORARYtables to a different engine or the server will not start. -
Command-Line Format --default_tmp_storage_engine=nameOption-File Format default_tmp_storage_engineSystem Variable Name default_tmp_storage_engineVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type enumerationDefault InnoDBThe default storage engine for
TEMPORARYtables (created withCREATE TEMPORARY TABLE). To set the storage engine for permanent tables, set thedefault_storage_enginesystem variable.If you disable the default storage engine at server startup, you must set the default engine for both permanent and
TEMPORARYtables to a different engine or the server will not start. -
Command-Line Format --default_week_format=#Option-File Format default_week_formatSystem Variable Name default_week_formatVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 0Range 0 .. 7The default mode value to use for the
WEEK()function. See Section 12.7, “Date and Time Functions”. -
Command-Line Format --delay-key-write[=name]Option-File Format delay-key-writeSystem Variable Name delay_key_writeVariable Scope Global Dynamic Variable Yes Permitted Values Type enumerationDefault ONValid Values ONOFFALLThis option applies only to
MyISAMtables. It can have one of the following values to affect handling of theDELAY_KEY_WRITEtable option that can be used inCREATE TABLEstatements.Option Description OFFDELAY_KEY_WRITEis ignored.ONMySQL honors any DELAY_KEY_WRITEoption specified inCREATE TABLEstatements. This is the default value.ALLAll new opened tables are treated as if they were created with the DELAY_KEY_WRITEoption enabled.If
DELAY_KEY_WRITEis enabled for a table, the key buffer is not flushed for the table on every index update, but only when the table is closed. This speeds up writes on keys a lot, but if you use this feature, you should add automatic checking of allMyISAMtables by starting the server with the--myisam-recover-optionsoption (for example,--myisam-recover-options=BACKUP,FORCE). See Section 5.1.3, “Server Command Options”, and Section 14.3.1, “MyISAMStartup Options”.WarningIf you enable external locking with
--external-locking, there is no protection against index corruption for tables that use delayed key writes. -
Deprecated 5.6.7 Command-Line Format --delayed_insert_limit=#Option-File Format delayed_insert_limitSystem Variable Name delayed_insert_limitVariable Scope Global Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 100Range 1 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 100Range 1 .. 18446744073709547520In MySQL 5.7, this system variable is deprecated (because
DELAYEDinserts are not supported), and will be removed in a future release. -
Deprecated 5.6.7 Command-Line Format --delayed_insert_timeout=#Option-File Format delayed_insert_timeoutSystem Variable Name delayed_insert_timeoutVariable Scope Global Dynamic Variable Yes Permitted Values Type numericDefault 300In MySQL 5.7, this system variable is deprecated (because
DELAYEDinserts are not supported), and will be removed in a future release. -
Deprecated 5.6.7 Command-Line Format --delayed_queue_size=#Option-File Format delayed_queue_sizeSystem Variable Name delayed_queue_sizeVariable Scope Global Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 1000Range 1 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 1000Range 1 .. 18446744073709547520In MySQL 5.7, this system variable is deprecated (because
DELAYEDinserts are not supported), and will be removed in a future release. disconnect_on_expired_passwordIntroduced 5.7.1 Command-Line Format --disconnect_on_expired_password=#Option-File Format disconnect_on_expired_passwordSystem Variable Name disconnect_on_expired_passwordVariable Scope Session Dynamic Variable No Permitted Values Type booleanDefault ONThis variable controls how the server handles clients with expired passwords:
If the client indicates that it can handle expires passwords, the value of
disconnect_on_expired_passwordis irrelevant. The server permits the client to connect but puts it in sandbox mode.If the client does not indicate that it can handle expires passwords, the server handles the client according to the value of
disconnect_on_expired_password:If
disconnect_on_expired_password: is enabled, the server disconnects the client.If
disconnect_on_expired_password: is disabled, the server permits the client to connect but puts it in sandbox mode.
For more information about the interaction of client and server settings relating to expired-password handling, see Section 6.3.6, “Password Expiration and Sandbox Mode”.
-
Command-Line Format --div_precision_increment=#Option-File Format div_precision_incrementSystem Variable Name div_precision_incrementVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 4Range 0 .. 30This variable indicates the number of digits by which to increase the scale of the result of division operations performed with the
/operator. The default value is 4. The minimum and maximum values are 0 and 30, respectively. The following example illustrates the effect of increasing the default value.mysql>
SELECT 1/7;+--------+ | 1/7 | +--------+ | 0.1429 | +--------+ mysql>SET div_precision_increment = 12;mysql>SELECT 1/7;+----------------+ | 1/7 | +----------------+ | 0.142857142857 | +----------------+ -
System Variable Name end_markers_in_jsonVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type booleanDefault OFFWhether optimizer JSON output should add end markers.
-
System Variable Name eq_range_index_dive_limitVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 10Range 0 .. 4294967295This variable indicates the number of equality ranges in an equality comparison condition when the optimizer should switch from using index dives to index statistics in estimating the number of qualifying rows. It applies to evaluation of expressions that have either of these equivalent forms, where the optimizer uses a nonunique index to look up
col_namevalues:col_nameIN(val1, ...,valN)col_name=val1OR ... ORcol_name=valNIn both cases, the expression contains
Nequality ranges. The optimizer can make row estimates using index dives or index statistics. Ifeq_range_index_dive_limitis greater than 0, the optimizer uses existing index statistics instead of index dives if there areeq_range_index_dive_limitor more equality ranges. Thus, to permit use of index dives for up toNequality ranges, seteq_range_index_dive_limittoN+ 1. Seteq_range_index_dive_limitto 0 to disable use of index statistics and always use index dives regardless ofN.For more information, see Section 8.2.1.3.3, “Equality Range Optimization of Many-Valued Comparisons”.
To update table index statistics for best estimates, use
ANALYZE TABLE. The number of errors that resulted from the last statement that generated messages. This variable is read only. See Section 13.7.5.16, “
SHOW ERRORSSyntax”.-
Command-Line Format --event-scheduler[=value]Option-File Format event-schedulerSystem Variable Name event_schedulerVariable Scope Global Dynamic Variable Yes Permitted Values Type enumerationDefault OFFValid Values ONOFFDISABLEDThis variable indicates the status of the Event Scheduler; possible values are
ON,OFF, andDISABLED, with the default beingOFF. This variable and its effects on the Event Scheduler's operation are discussed in greater detail in the Overview section of the Events chapter. -
Command-Line Format --expire_logs_days=#Option-File Format expire_logs_daysSystem Variable Name expire_logs_daysVariable Scope Global Dynamic Variable Yes Permitted Values Type numericDefault 0Range 0 .. 99The number of days for automatic binary log file removal. The default is 0, which means “no automatic removal.” Possible removals happen at startup and when the binary log is flushed. Log flushing occurs as indicated in Section 5.2, “MySQL Server Logs”.
To remove binary log files manually, use the
PURGE BINARY LOGSstatement. See Section 13.4.1.1, “PURGE BINARY LOGSSyntax”. explicit_defaults_for_timestampCommand-Line Format --explicit_defaults_for_timestamp=#Option-File Format explicit_defaults_for_timestampSystem Variable Name explicit_defaults_for_timestampVariable Scope Session Dynamic Variable No Permitted Values Type booleanDefault FALSEIn MySQL, the
TIMESTAMPdata type differs in nonstandard ways from other data types:TIMESTAMPcolumns not explicitly declared with theNULLattribute are assigned theNOT NULLattribute. (Columns of other data types, if not explicitly declared asNOT NULL, permitNULLvalues.) Setting such a column toNULLsets it to the current timestamp.The first
TIMESTAMPcolumn in a table, if not declared with theNULLattribute or an explicitDEFAULTorON UPDATEclause, is automatically assigned theDEFAULT CURRENT_TIMESTAMPandON UPDATE CURRENT_TIMESTAMPattributes.TIMESTAMPcolumns following the first one, if not declared with theNULLattribute or an explicitDEFAULTclause, are automatically assignedDEFAULT '0000-00-00 00:00:00'(the “zero” timestamp). For inserted rows that specify no explicit value for such a column, the column is assigned'0000-00-00 00:00:00'and no warning occurs.
Those nonstandard behaviors remain the default for
TIMESTAMPbut as of MySQL 5.6.6 are deprecated and this warning appears at startup:[Warning] TIMESTAMP with implicit DEFAULT value is deprecated. Please use --explicit_defaults_for_timestamp server option (see documentation for more details).
As indicated by the warning, to turn off the nonstandard behaviors, enable the new
explicit_defaults_for_timestampsystem variable at server startup. With this variable enabled, the server handlesTIMESTAMPas follows instead:TIMESTAMPcolumns not explicitly declared asNOT NULLpermitNULLvalues. Setting such a column toNULLsets it toNULL, not the current timestamp.No
TIMESTAMPcolumn is assigned theDEFAULT CURRENT_TIMESTAMPorON UPDATE CURRENT_TIMESTAMPattributes automatically. Those attributes must be explicitly specified.TIMESTAMPcolumns declared asNOT NULLand without an explicitDEFAULTclause are treated as having no default value. For inserted rows that specify no explicit value for such a column, the result depends on the SQL mode. If strict SQL mode is enabled, an error occurs. If strict SQL mode is not enabled, the column is assigned the implicit default of'0000-00-00 00:00:00'and a warning occurs. This is similar to how MySQL treats other temporal types such asDATETIME.
Noteexplicit_defaults_for_timestampis itself deprecated because its only purpose is to permit control over now-deprecatedTIMESTAMPbehaviors that will be removed in a future MySQL release. When that removal occurs,explicit_defaults_for_timestampwill have no purpose and will be removed as well.-
System Variable Name external_userVariable Scope Session Dynamic Variable No Permitted Values Type stringThe external user name used during the authentication process, as set by the plugin used to authenticate the client. With native (built-in) MySQL authentication, or if the plugin does not set the value, this variable is
NULL. See Section 6.3.8, “Proxy Users”. -
Command-Line Format --flushOption-File Format flushSystem Variable Name flushVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanDefault OFFIf
ON, the server flushes (synchronizes) all changes to disk after each SQL statement. Normally, MySQL does a write of all changes to disk only after each SQL statement and lets the operating system handle the synchronizing to disk. See Section C.5.4.2, “What to Do If MySQL Keeps Crashing”. This variable is set toONif you start mysqld with the--flushoption. -
Command-Line Format --flush_time=#Option-File Format flush_timeSystem Variable Name flush_timeVariable Scope Global Dynamic Variable Yes Permitted Values Type (windows) numericDefault 0Min Value 0If this is set to a nonzero value, all tables are closed every
flush_timeseconds to free up resources and synchronize unflushed data to disk. This option is best used only on systems with minimal resources. If set to 1 (the default), foreign key constraints for
InnoDBtables are checked. If set to 0, they are ignored. Typically you leave this setting enabled during normal operation, to enforce referential integrity. Disabling foreign key checking can be useful for reloadingInnoDBtables in an order different from that required by their parent/child relationships. See Section 5.4.5, “InnoDBandFOREIGN KEYConstraints”.Setting
foreign_key_checksto 0 also affects data definition statements:DROP SCHEMAdrops a schema even if it contains tables that have foreign keys that are referred to by tables outside the schema, andDROP TABLEdrops tables that have foreign keys that are referred to by other tables.NoteSetting
foreign_key_checksto 1 does not trigger a scan of the existing table data. Therefore, rows added to the table whileforeign_key_checks = 0will not be verified for consistency.-
Command-Line Format --ft_boolean_syntax=nameOption-File Format ft_boolean_syntaxSystem Variable Name ft_boolean_syntaxVariable Scope Global Dynamic Variable Yes Permitted Values Type stringDefault + -><()~*:""&|The list of operators supported by boolean full-text searches performed using
IN BOOLEAN MODE. See Section 12.9.2, “Boolean Full-Text Searches”.The default variable value is
'+ -><()~*:""&|'. The rules for changing the value are as follows:Operator function is determined by position within the string.
The replacement value must be 14 characters.
Each character must be an ASCII nonalphanumeric character.
Either the first or second character must be a space.
No duplicates are permitted except the phrase quoting operators in positions 11 and 12. These two characters are not required to be the same, but they are the only two that may be.
Positions 10, 13, and 14 (which by default are set to “
:”, “&”, and “|”) are reserved for future extensions.
-
Command-Line Format --ft_max_word_len=#Option-File Format ft_max_word_lenSystem Variable Name ft_max_word_lenVariable Scope Global Dynamic Variable No Permitted Values Type numericMin Value 10The maximum length of the word to be included in a
MyISAMFULLTEXTindex.NoteFULLTEXTindexes onMyISAMtables must be rebuilt after changing this variable. UseREPAIR TABLE.tbl_nameQUICK -
Command-Line Format --ft_min_word_len=#Option-File Format ft_min_word_lenSystem Variable Name ft_min_word_lenVariable Scope Global Dynamic Variable No Permitted Values Type numericDefault 4Min Value 1The minimum length of the word to be included in a
MyISAMFULLTEXTindex.NoteFULLTEXTindexes onMyISAMtables must be rebuilt after changing this variable. UseREPAIR TABLE.tbl_nameQUICK -
Command-Line Format --ft_query_expansion_limit=#Option-File Format ft_query_expansion_limitSystem Variable Name ft_query_expansion_limitVariable Scope Global Dynamic Variable No Permitted Values Type numericDefault 20Range 0 .. 1000The number of top matches to use for full-text searches performed using
WITH QUERY EXPANSION. -
Command-Line Format --ft_stopword_file=file_nameOption-File Format ft_stopword_fileSystem Variable Name ft_stopword_fileVariable Scope Global Dynamic Variable No Permitted Values Type file nameThe file from which to read the list of stopwords for full-text searches on
MyISAMtables. The server looks for the file in the data directory unless an absolute path name is given to specify a different directory. All the words from the file are used; comments are not honored. By default, a built-in list of stopwords is used (as defined in thestorage/myisam/ft_static.cfile). Setting this variable to the empty string ('') disables stopword filtering. See also Section 12.9.4, “Full-Text Stopwords”.NoteFULLTEXTindexes onMyISAMtables must be rebuilt after changing this variable or the contents of the stopword file. UseREPAIR TABLE.tbl_nameQUICK -
Command-Line Format --general-logOption-File Format general-logSystem Variable Name general_logVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanDefault OFFWhether the general query log is enabled. The value can be 0 (or
OFF) to disable the log or 1 (orON) to enable the log. The default value depends on whether the--general_logoption is given. The destination for log output is controlled by thelog_outputsystem variable; if that value isNONE, no log entries are written even if the log is enabled. -
Command-Line Format --general-log-file=file_nameOption-File Format general_log_fileSystem Variable Name general_log_fileVariable Scope Global Dynamic Variable Yes Permitted Values Type file nameDefault host_name.logThe name of the general query log file. The default value is
host_name.log--general_log_fileoption. -
Command-Line Format --group_concat_max_len=#Option-File Format group_concat_max_lenSystem Variable Name group_concat_max_lenVariable Scope Global, Session Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 1024Range 4 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 1024Range 4 .. 18446744073709547520The maximum permitted result length in bytes for the
GROUP_CONCAT()function. The default is 1024. YESif thezlibcompression library is available to the server,NOif not. If not, theCOMPRESS()andUNCOMPRESS()functions cannot be used.YESif thecrypt()system call is available to the server,NOif not. If not, theENCRYPT()function cannot be used.YESif mysqld supports dynamic loading of plugins,NOif not.YESif the server supports spatial data types,NOif not.This variable is an alias for
have_ssl.YESif statement profiling capability is present,NOif not. If present, theprofilingsystem variable controls whether this capability is enabled or disabled. See Section 13.7.5.30, “SHOW PROFILESSyntax”.This variable is deprecated and will be removed in a future MySQL release.
YESif mysqld supports the query cache,NOif not.YESifRTREEindexes are available,NOif not. (These are used for spatial indexes inMyISAMtables.)YESif mysqld supports SSL connections,NOif not.DISABLEDindicates that the server was compiled with SSL support, but but was not started with the appropriate--ssl-options. For more information, see Section 6.3.9.2, “Configuring MySQL for SSL”.xxxYESif symbolic link support is enabled,NOif not. This is required on Unix for support of theDATA DIRECTORYandINDEX DIRECTORYtable options. If the server is started with the--skip-symbolic-linksoption, the value isDISABLED.This variable has no meaning on Windows.
-
System Variable Name host_cache_sizeVariable Scope Global Dynamic Variable Yes Permitted Values Type numericDefault -1 (autosized)Range 0 .. 65536The size of the internal host cache (see Section 8.11.5.2, “DNS Lookup Optimization and the Host Cache”). Setting the size to 0 disables the host cache. Changing the cache size at runtime implicitly causes a
FLUSH HOSTSoperation to clear the host cache and truncate thehost_cachetable.The default value is 128, plus 1 for a value of
max_connectionsup to 500, plus 1 for every increment of 20 over 500 in themax_connectionsvalue, capped to a limit of 2000.Use of
--skip-host-cacheis similar to setting thehost_cache_sizesystem variable to 0, buthost_cache_sizeis more flexible because it can also be used to resize, enable, or disable the host cache at runtime, not just at server startup.If you start the server with
--skip-host-cache, that does not prevent changes to the value ofhost_cache_size, but such changes have no effect and the cache is not re-enabled even ifhost_cache_sizeis set larger than 0. -
System Variable Name hostnameVariable Scope Global Dynamic Variable No Permitted Values Type stringThe server sets this variable to the server host name at startup.
This variable is a synonym for the
last_insert_idvariable. It exists for compatibility with other database systems. You can read its value withSELECT @@identity, and set it usingSET identity.-
System Variable Name ignore_db_dirsVariable Scope Global Dynamic Variable No Permitted Values Type stringA comma-separated list of names that are not considered as database directories in the data directory. The value is set from any instances of
--ignore-db-dirgiven at server startup. -
Command-Line Format --init-connect=nameOption-File Format init_connectSystem Variable Name init_connectVariable Scope Global Dynamic Variable Yes Permitted Values Type stringA string to be executed by the server for each client that connects. The string consists of one or more SQL statements, separated by semicolon characters. For example, each client session begins by default with autocommit mode enabled. For older servers (before MySQL 5.5.8), there is no global
autocommitsystem variable to specify that autocommit should be disabled by default, but as a workaroundinit_connectcan be used to achieve the same effect:SET GLOBAL init_connect='SET autocommit=0';
The
init_connectvariable can also be set on the command line or in an option file. To set the variable as just shown using an option file, include these lines:[mysqld] init_connect='SET autocommit=0'
The content of
init_connectis not executed for users that have theSUPERprivilege. This is done so that an erroneous value forinit_connectdoes not prevent all clients from connecting. For example, the value might contain a statement that has a syntax error, thus causing client connections to fail. Not executinginit_connectfor users that have theSUPERprivilege enables them to open a connection and fix theinit_connectvalue. -
Command-Line Format --init-file=file_nameOption-File Format init-fileSystem Variable Name init_fileVariable Scope Global Dynamic Variable No Permitted Values Type file nameThe name of the file specified with the
--init-fileoption when you start the server. This should be a file containing SQL statements that you want the server to execute when it starts. Each statement must be on a single line and should not include comments. No statement terminator such as;,\g, or\Gshould be given at the end of each statement. innodb_xxxInnoDBsystem variables are listed in Section 14.2.6, “InnoDBStartup Options and System Variables”. These variables control many aspects of storage, memory use, and I/O patterns forInnoDBtables, and are especially important now that InnoDB is the default storage engine.The value to be used by the following
INSERTorALTER TABLEstatement when inserting anAUTO_INCREMENTvalue. This is mainly used with the binary log.-
Command-Line Format --interactive_timeout=#Option-File Format interactive_timeoutSystem Variable Name interactive_timeoutVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 28800Min Value 1The number of seconds the server waits for activity on an interactive connection before closing it. An interactive client is defined as a client that uses the
CLIENT_INTERACTIVEoption tomysql_real_connect(). See alsowait_timeout. -
Command-Line Format --join_buffer_size=#Option-File Format join_buffer_sizeSystem Variable Name join_buffer_sizeVariable Scope Global, Session Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 262144Range 128 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 262144Range 128 .. 18446744073709547520The minimum size of the buffer that is used for plain index scans, range index scans, and joins that do not use indexes and thus perform full table scans. Normally, the best way to get fast joins is to add indexes. Increase the value of
join_buffer_sizeto get a faster full join when adding indexes is not possible. One join buffer is allocated for each full join between two tables. For a complex join between several tables for which indexes are not used, multiple join buffers might be necessary. There is no gain from setting the buffer larger than required to hold each matching row, and all joins allocate at least the minimum size, so use caution in setting this variable to a large value globally. It is better to keep the global setting small and change to a larger setting only in sessions that are doing large joins. Memory allocation time can cause substantial performance drops if the global size is larger than needed by most queries that use it.The default is 256KB. The maximum permissible setting for
join_buffer_sizeis 4GB. Values larger than 4GB are permitted for 64-bit platforms (except 64-bit Windows, for which large values are truncated to 4GB with a warning).For additional information about join buffering, see Section 8.2.1.10, “Nested-Loop Join Algorithms”.
-
Command-Line Format --keep_files_on_create=#Option-File Format keep_files_on_createSystem Variable Name keep_files_on_createVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type booleanDefault OFFIf a
MyISAMtable is created with noDATA DIRECTORYoption, the.MYDfile is created in the database directory. By default, ifMyISAMfinds an existing.MYDfile in this case, it overwrites it. The same applies to.MYIfiles for tables created with noINDEX DIRECTORYoption. To suppress this behavior, set thekeep_files_on_createvariable toON(1), in which caseMyISAMwill not overwrite existing files and returns an error instead. The default value isOFF(0).If a
MyISAMtable is created with aDATA DIRECTORYorINDEX DIRECTORYoption and an existing.MYDor.MYIfile is found, MyISAM always returns an error. It will not overwrite a file in the specified directory. -
Command-Line Format --key_buffer_size=#Option-File Format key_buffer_sizeSystem Variable Name key_buffer_sizeVariable Scope Global Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 8388608Range 8 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 8388608Range 8 .. OS_PER_PROCESS_LIMITIndex blocks for
MyISAMtables are buffered and are shared by all threads.key_buffer_sizeis the size of the buffer used for index blocks. The key buffer is also known as the key cache.The maximum permissible setting for
key_buffer_sizeis 4GB on 32-bit platforms. Values larger than 4GB are permitted for 64-bit platforms. The effective maximum size might be less, depending on your available physical RAM and per-process RAM limits imposed by your operating system or hardware platform. The value of this variable indicates the amount of memory requested. Internally, the server allocates as much memory as possible up to this amount, but the actual allocation might be less.You can increase the value to get better index handling for all reads and multiple writes; on a system whose primary function is to run MySQL using the
MyISAMstorage engine, 25% of the machine's total memory is an acceptable value for this variable. However, you should be aware that, if you make the value too large (for example, more than 50% of the machine's total memory), your system might start to page and become extremely slow. This is because MySQL relies on the operating system to perform file system caching for data reads, so you must leave some room for the file system cache. You should also consider the memory requirements of any other storage engines that you may be using in addition toMyISAM.For even more speed when writing many rows at the same time, use
LOCK TABLES. See Section 8.2.2.1, “Speed ofINSERTStatements”.You can check the performance of the key buffer by issuing a
SHOW STATUSstatement and examining theKey_read_requests,Key_reads,Key_write_requests, andKey_writesstatus variables. (See Section 13.7.5, “SHOWSyntax”.) TheKey_reads/Key_read_requestsratio should normally be less than 0.01. TheKey_writes/Key_write_requestsratio is usually near 1 if you are using mostly updates and deletes, but might be much smaller if you tend to do updates that affect many rows at the same time or if you are using theDELAY_KEY_WRITEtable option.The fraction of the key buffer in use can be determined using
key_buffer_sizein conjunction with theKey_blocks_unusedstatus variable and the buffer block size, which is available from thekey_cache_block_sizesystem variable:1 - ((Key_blocks_unused * key_cache_block_size) / key_buffer_size)
This value is an approximation because some space in the key buffer is allocated internally for administrative structures. Factors that influence the amount of overhead for these structures include block size and pointer size. As block size increases, the percentage of the key buffer lost to overhead tends to decrease. Larger blocks results in a smaller number of read operations (because more keys are obtained per read), but conversely an increase in reads of keys that are not examined (if not all keys in a block are relevant to a query).
It is possible to create multiple
MyISAMkey caches. The size limit of 4GB applies to each cache individually, not as a group. See Section 8.9.2, “TheMyISAMKey Cache”. -
Command-Line Format --key_cache_age_threshold=#Option-File Format key_cache_age_thresholdSystem Variable Name key_cache_age_thresholdVariable Scope Global Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 300Range 100 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 300Range 100 .. 18446744073709547520This value controls the demotion of buffers from the hot sublist of a key cache to the warm sublist. Lower values cause demotion to happen more quickly. The minimum value is 100. The default value is 300. See Section 8.9.2, “The
MyISAMKey Cache”. -
Command-Line Format --key_cache_block_size=#Option-File Format key_cache_block_sizeSystem Variable Name key_cache_block_sizeVariable Scope Global Dynamic Variable Yes Permitted Values Type numericDefault 1024Range 512 .. 16384The size in bytes of blocks in the key cache. The default value is 1024. See Section 8.9.2, “The
MyISAMKey Cache”. -
Command-Line Format --key_cache_division_limit=#Option-File Format key_cache_division_limitSystem Variable Name key_cache_division_limitVariable Scope Global Dynamic Variable Yes Permitted Values Type numericDefault 100Range 1 .. 100The division point between the hot and warm sublists of the key cache buffer list. The value is the percentage of the buffer list to use for the warm sublist. Permissible values range from 1 to 100. The default value is 100. See Section 8.9.2, “The
MyISAMKey Cache”. -
System Variable Name large_files_supportVariable Scope Global Dynamic Variable No Whether mysqld was compiled with options for large file support.
-
Command-Line Format --large-pagesOption-File Format large-pagesSystem Variable Name large_pagesVariable Scope Global Dynamic Variable No Platform Specific linux Permitted Values Type (linux) booleanDefault FALSEWhether large page support is enabled (via the
--large-pagesoption). See Section 8.11.4.2, “Enabling Large Page Support”. -
System Variable Name large_page_sizeVariable Scope Global Dynamic Variable No Permitted Values Type (linux) numericDefault 0If large page support is enabled, this shows the size of memory pages. Currently, large memory pages are supported only on Linux; on other platforms, the value of this variable is always 0. See Section 8.11.4.2, “Enabling Large Page Support”.
The value to be returned from
LAST_INSERT_ID(). This is stored in the binary log when you useLAST_INSERT_ID()in a statement that updates a table. Setting this variable does not update the value returned by themysql_insert_id()C API function.-
Command-Line Format --lc-messages=nameOption-File Format lc-messagesSystem Variable Name lc_messagesVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type stringThe locale to use for error messages. The server converts the value to a language name and combines it with the value of the
lc_messages_dirto produce the location for the error message file. See Section 10.2, “Setting the Error Message Language”. -
Command-Line Format --lc-messages-dir=pathOption-File Format lc-messages-dirSystem Variable Name lc_messages_dirVariable Scope Global Dynamic Variable No Permitted Values Type directory nameThe directory where error messages are located. The value is used together with the value of
lc_messagesto produce the location for the error message file. See Section 10.2, “Setting the Error Message Language”. -
System Variable Name lc_time_namesVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type stringThis variable specifies the locale that controls the language used to display day and month names and abbreviations. This variable affects the output from the
DATE_FORMAT(),DAYNAME()andMONTHNAME()functions. Locale names are POSIX-style values such as'ja_JP'or'pt_BR'. The default value is'en_US'regardless of your system's locale setting. For further information, see Section 10.7, “MySQL Server Locale Support”. -
System Variable Name licenseVariable Scope Global Dynamic Variable No Permitted Values Type stringDefault GPLThe type of license the server has.
-
System Variable Name local_infileVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanWhether
LOCALis supported forLOAD DATA INFILEstatements. If this variable is disabled, clients cannot useLOCALinLOAD DATAstatements. See Section 6.1.6, “Security Issues withLOAD DATA LOCAL”. -
Command-Line Format --lock_wait_timeout=#Option-File Format lock_wait_timeoutSystem Variable Name lock_wait_timeoutVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 31536000Range 1 .. 31536000This variable specifies the timeout in seconds for attempts to acquire metadata locks. The permissible values range from 1 to 31536000 (1 year). The default is 31536000.
This timeout applies to all statements that use metadata locks. These include DML and DDL operations on tables, views, stored procedures, and stored functions, as well as
LOCK TABLES,FLUSH TABLES WITH READ LOCK, andHANDLERstatements.This timeout does not apply to implicit accesses to system tables in the
mysqldatabase, such as grant tables modified byGRANTorREVOKEstatements or table logging statements. The timeout does apply to system tables accessed directly, such as withSELECTorUPDATE.The timeout value applies separately for each metadata lock attempt. A given statement can require more than one lock, so it is possible for the statement to block for longer than the
lock_wait_timeoutvalue before reporting a timeout error. When lock timeout occurs,ER_LOCK_WAIT_TIMEOUTis reported.lock_wait_timeoutdoes not apply to delayed inserts, which always execute with a timeout of 1 year. This is done to avoid unnecessary timeouts because a session that issues a delayed insert receives no notification of delayed insert timeouts. -
System Variable Name locked_in_memoryVariable Scope Global Dynamic Variable No log_bin_trust_function_creatorsCommand-Line Format --log-bin-trust-function-creatorsOption-File Format log-bin-trust-function-creatorsSystem Variable Name log_bin_trust_function_creatorsVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanDefault FALSEThis variable applies when binary logging is enabled. It controls whether stored function creators can be trusted not to create stored functions that will cause unsafe events to be written to the binary log. If set to 0 (the default), users are not permitted to create or alter stored functions unless they have the
SUPERprivilege in addition to theCREATE ROUTINEorALTER ROUTINEprivilege. A setting of 0 also enforces the restriction that a function must be declared with theDETERMINISTICcharacteristic, or with theREADS SQL DATAorNO SQLcharacteristic. If the variable is set to 1, MySQL does not enforce these restrictions on stored function creation. This variable also applies to trigger creation. See Section 18.7, “Binary Logging of Stored Programs”.-
Command-Line Format --log-error[=name]Option-File Format log-errorSystem Variable Name log_errorVariable Scope Global Dynamic Variable No Permitted Values Type file nameThe location of the error log, or
stderrif the server is writing error message to the standard error output. See Section 5.2.2, “The Error Log”. -
Command-Line Format --log-output=nameOption-File Format log-outputSystem Variable Name log_outputVariable Scope Global Dynamic Variable Yes Permitted Values Type setDefault FILEValid Values TABLEFILENONEThe destination for general query log and slow query log output. The value can be a comma-separated list of one or more of the words
TABLE(log to tables),FILE(log to files), orNONE(do not log to tables or files). The default value isFILE.NONE, if present, takes precedence over any other specifiers. If the value isNONElog entries are not written even if the logs are enabled. If the logs are not enabled, no logging occurs even if the value oflog_outputis notNONE. For more information, see Section 5.2.1, “Selecting General Query and Slow Query Log Output Destinations”. -
Command-Line Format --log-queries-not-using-indexesOption-File Format log-queries-not-using-indexesSystem Variable Name log_queries_not_using_indexesVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanDefault OFFWhether queries that do not use indexes are logged to the slow query log. See Section 5.2.5, “The Slow Query Log”.
log_throttle_queries_not_using_indexesSystem Variable Name log_throttle_queries_not_using_indexesVariable Scope Global Dynamic Variable Yes Permitted Values Type numericDefault 0If
log_queries_not_using_indexesis enabled, thelog_throttle_queries_not_using_indexesvariable limits the number of such queries per minute that can be written to the slow query log. A value of 0 (the default) means “no limit”. For more information, see Section 5.2.5, “The Slow Query Log”.-
Introduced 5.7.1 System Variable Name log_slow_admin_statementsVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanDefault OFFInclude slow administrative statements in the statements written to the slow query log. Administrative statements include
ALTER TABLE,ANALYZE TABLE,CHECK TABLE,CREATE INDEX,DROP INDEX,OPTIMIZE TABLE, andREPAIR TABLE.This variable was added in MySQL 5.7.1.
-
Command-Line Format --log-warnings[=#]-W [#]Option-File Format log-warnings[=#]System Variable Name log_warningsVariable Scope Global Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 1Range 0 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 1Range 0 .. 18446744073709547520Whether to produce additional warning messages to the error log. This variable is enabled (1) by default and can be disabled by setting it to 0. The server logs messages about statements that are unsafe for statement-based logging if the value is greater than 0. Aborted connections and access-denied errors for new connection attempts are logged if the value is greater than 1.
-
Command-Line Format --long_query_time=#Option-File Format long_query_timeSystem Variable Name long_query_timeVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 10Min Value 0If a query takes longer than this many seconds, the server increments the
Slow_queriesstatus variable. If the slow query log is enabled, the query is logged to the slow query log file. This value is measured in real time, not CPU time, so a query that is under the threshold on a lightly loaded system might be above the threshold on a heavily loaded one. The minimum and default values oflong_query_timeare 0 and 10, respectively. The value can be specified to a resolution of microseconds. For logging to a file, times are written including the microseconds part. For logging to tables, only integer times are written; the microseconds part is ignored. See Section 5.2.5, “The Slow Query Log”. -
Command-Line Format --low-priority-updatesOption-File Format low-priority-updatesSystem Variable Name low_priority_updatesVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type booleanDefault FALSEIf set to
1, allINSERT,UPDATE,DELETE, andLOCK TABLE WRITEstatements wait until there is no pendingSELECTorLOCK TABLE READon the affected table. This affects only storage engines that use only table-level locking (such asMyISAM,MEMORY, andMERGE). -
System Variable Name lower_case_file_systemVariable Scope Global Dynamic Variable No Permitted Values Type booleanThis variable describes the case sensitivity of file names on the file system where the data directory is located.
OFFmeans file names are case sensitive,ONmeans they are not case sensitive. This variable is read only because it reflects a file system attribute and setting it would have no effect on the file system. -
Command-Line Format --lower_case_table_names[=#]Option-File Format lower_case_table_namesSystem Variable Name lower_case_table_namesVariable Scope Global Dynamic Variable No Permitted Values Type numericDefault 0Range 0 .. 2If set to 0, table names are stored as specified and comparisons are case sensitive. If set to 1, table names are stored in lowercase on disk and comparisons are not case sensitive. If set to 2, table names are stored as given but compared in lowercase. This option also applies to database names and table aliases. For additional information, see Section 9.2.2, “Identifier Case Sensitivity”.
You should not set this variable to 0 if you are running MySQL on a system that has case-insensitive file names (such as Windows or Mac OS X). If you set this variable to 0 on such a system and access
MyISAMtablenames using different lettercases, index corruption may result. On Windows the default value is 1. On Mac OS X, the default value is 2.If you are using
InnoDBtables, you should set this variable to 1 on all platforms to force names to be converted to lowercase.The setting of this variable in MySQL 5.7 affects the behavior of replication filtering options with regard to case sensitivity. (Bug #51639) See Section 16.2.3, “How Servers Evaluate Replication Filtering Rules”, for more information.
-
Command-Line Format --max_allowed_packet=#Option-File Format max_allowed_packetSystem Variable Name max_allowed_packetVariable Scope Global Dynamic Variable Yes Permitted Values Type numericDefault 4194304Range 1024 .. 1073741824The maximum size of one packet or any generated/intermediate string, or any parameter sent by the
mysql_stmt_send_long_data()C API function. The default is 4MB.The packet message buffer is initialized to
net_buffer_lengthbytes, but can grow up tomax_allowed_packetbytes when needed. This value by default is small, to catch large (possibly incorrect) packets.You must increase this value if you are using large
BLOBcolumns or long strings. It should be as big as the largestBLOByou want to use. The protocol limit formax_allowed_packetis 1GB. The value should be a multiple of 1024; nonmultiples are rounded down to the nearest multiple.When you change the message buffer size by changing the value of the
max_allowed_packetvariable, you should also change the buffer size on the client side if your client program permits it. The defaultmax_allowed_packetvalue built in to the client library is 1GB, but individual client programs might override this. For example, mysql and mysqldump have defaults of 16MB and 24MB, respectively. They also enable you to change the client-side value by settingmax_allowed_packeton the command line or in an option file.The session value of this variable is read only.
-
Command-Line Format --max_connect_errors=#Option-File Format max_connect_errorsSystem Variable Name max_connect_errorsVariable Scope Global Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 100Range 1 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 100Range 1 .. 18446744073709547520If more than this many successive connection requests from a host are interrupted without a successful connection, the server blocks that host from further connections. You can unblock blocked hosts by flushing the host cache. To do so, issue a
FLUSH HOSTSstatement or execute a mysqladmin flush-hosts command. If a connection is established successfully within fewer thanmax_connect_errorsattempts after a previous connection was interrupted, the error count for the host is cleared to zero. However, once a host is blocked, flushing the host cache is the only way to unblock it. The default is 100. -
Command-Line Format --max_connections=#Option-File Format max_connectionsSystem Variable Name max_connectionsVariable Scope Global Dynamic Variable Yes Permitted Values Type numericDefault 151Range 1 .. 100000The maximum permitted number of simultaneous client connections. By default, this is 151. See Section C.5.2.7, “
Too many connections”, for more information.Increasing this value increases the number of file descriptors that mysqld requires. See Section 8.4.3.1, “How MySQL Opens and Closes Tables”, for comments on file descriptor limits.
Connections refused because the
max_connectionslimit is reached increment theConnection_errors_max_connectionsstatus variable. -
Deprecated 5.6.7 Command-Line Format --max_delayed_threads=#Option-File Format max_delayed_threadsSystem Variable Name max_delayed_threadsVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 20Range 0 .. 16384In MySQL 5.7, this system variable is deprecated (because
DELAYEDinserts are not supported), and will be removed in a future release. -
Command-Line Format --max_error_count=#Option-File Format max_error_countSystem Variable Name max_error_countVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 64Range 0 .. 65535The maximum number of error, warning, and note messages to be stored for display by the
SHOW ERRORSandSHOW WARNINGSstatements. This is the same as the number of condition areas in the diagnostics area, and thus the number of conditions that can be inspected byGET DIAGNOSTICS. -
Command-Line Format --max_heap_table_size=#Option-File Format max_heap_table_sizeSystem Variable Name max_heap_table_sizeVariable Scope Global, Session Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 16777216Range 16384 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 16777216Range 16384 .. 1844674407370954752This variable sets the maximum size to which user-created
MEMORYtables are permitted to grow. The value of the variable is used to calculateMEMORYtableMAX_ROWSvalues. Setting this variable has no effect on any existingMEMORYtable, unless the table is re-created with a statement such asCREATE TABLEor altered withALTER TABLEorTRUNCATE TABLE. A server restart also sets the maximum size of existingMEMORYtables to the globalmax_heap_table_sizevalue.This variable is also used in conjunction with
tmp_table_sizeto limit the size of internal in-memory tables. See Section 8.4.3.3, “How MySQL Uses Internal Temporary Tables”.max_heap_table_sizeis not replicated. See Section 16.4.1.21, “Replication andMEMORYTables”, and Section 16.4.1.33, “Replication and Variables”, for more information. -
Deprecated 5.6.7 System Variable Name max_insert_delayed_threadsVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericThis variable is a synonym for
max_delayed_threads.In MySQL 5.7, this system variable is deprecated (because
DELAYEDinserts are not supported), and will be removed in a future release. -
Command-Line Format --max_join_size=#Option-File Format max_join_sizeSystem Variable Name max_join_sizeVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 18446744073709551615Range 1 .. 18446744073709551615Do not permit statements that probably need to examine more than
max_join_sizerows (for single-table statements) or row combinations (for multiple-table statements) or that are likely to do more thanmax_join_sizedisk seeks. By setting this value, you can catch statements where keys are not used properly and that would probably take a long time. Set it if your users tend to perform joins that lack aWHEREclause, that take a long time, or that return millions of rows.Setting this variable to a value other than
DEFAULTresets the value ofsql_big_selectsto0. If you set thesql_big_selectsvalue again, themax_join_sizevariable is ignored.If a query result is in the query cache, no result size check is performed, because the result has previously been computed and it does not burden the server to send it to the client.
-
Command-Line Format --max_length_for_sort_data=#Option-File Format max_length_for_sort_dataSystem Variable Name max_length_for_sort_dataVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 1024Range 4 .. 8388608The cutoff on the size of index values that determines which
filesortalgorithm to use. See Section 8.2.1.15, “ORDER BYOptimization”. -
Command-Line Format --max_prepared_stmt_count=#Option-File Format max_prepared_stmt_countSystem Variable Name max_prepared_stmt_countVariable Scope Global Dynamic Variable Yes Permitted Values Type numericDefault 16382Range 0 .. 1048576This variable limits the total number of prepared statements in the server. It can be used in environments where there is the potential for denial-of-service attacks based on running the server out of memory by preparing huge numbers of statements. If the value is set lower than the current number of prepared statements, existing statements are not affected and can be used, but no new statements can be prepared until the current number drops below the limit. The default value is 16,382. The permissible range of values is from 0 to 1 million. Setting the value to 0 disables prepared statements.
-
Command-Line Format --max_relay_log_size=#Option-File Format max_relay_log_sizeSystem Variable Name max_relay_log_sizeVariable Scope Global Dynamic Variable Yes Permitted Values Type numericDefault 0Range 0 .. 1073741824If a write by a replication slave to its relay log causes the current log file size to exceed the value of this variable, the slave rotates the relay logs (closes the current file and opens the next one). If
max_relay_log_sizeis 0, the server usesmax_binlog_sizefor both the binary log and the relay log. Ifmax_relay_log_sizeis greater than 0, it constrains the size of the relay log, which enables you to have different sizes for the two logs. You must setmax_relay_log_sizeto between 4096 bytes and 1GB (inclusive), or to 0. The default value is 0. See Section 16.2.1, “Replication Implementation Details”. -
Command-Line Format --max_seeks_for_key=#Option-File Format max_seeks_for_keySystem Variable Name max_seeks_for_keyVariable Scope Global, Session Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 4294967295Range 1 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 18446744073709547520Range 1 .. 18446744073709547520Limit the assumed maximum number of seeks when looking up rows based on a key. The MySQL optimizer assumes that no more than this number of key seeks are required when searching for matching rows in a table by scanning an index, regardless of the actual cardinality of the index (see Section 13.7.5.21, “
SHOW INDEXSyntax”). By setting this to a low value (say, 100), you can force MySQL to prefer indexes instead of table scans. -
Command-Line Format --max_sort_length=#Option-File Format max_sort_lengthSystem Variable Name max_sort_lengthVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 1024Range 4 .. 8388608The number of bytes to use when sorting data values. Only the first
max_sort_lengthbytes of each value are used; the rest are ignored.max_sort_lengthis ignored for integer, decimal, floating-point, and temporal data types. -
Command-Line Format --max_sp_recursion_depth[=#]Option-File Format max_sp_recursion_depthSystem Variable Name max_sp_recursion_depthVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 0Max Value 255The number of times that any given stored procedure may be called recursively. The default value for this option is 0, which completely disables recursion in stored procedures. The maximum value is 255.
Stored procedure recursion increases the demand on thread stack space. If you increase the value of
max_sp_recursion_depth, it may be necessary to increase thread stack size by increasing the value ofthread_stackat server startup. This variable is unused. It is deprecated and will be removed in a future MySQL release.
-
Command-Line Format --max_user_connections=#Option-File Format max_user_connectionsSystem Variable Name max_user_connectionsVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 0Range 0 .. 4294967295The maximum number of simultaneous connections permitted to any given MySQL user account. A value of 0 (the default) means “no limit.”
This variable has a global value that can be set at server startup or runtime. It also has a read-only session value that indicates the effective simultaneous-connection limit that applies to the account associated with the current session. The session value is initialized as follows:
If the user account has a nonzero
MAX_USER_CONNECTIONSresource limit, the sessionmax_user_connectionsvalue is set to that limit.Otherwise, the session
max_user_connectionsvalue is set to the global value.
Account resource limits are specified using the
GRANTstatement. See Section 6.3.4, “Setting Account Resource Limits”, and Section 13.7.1.4, “GRANTSyntax”. -
Command-Line Format --max_write_lock_count=#Option-File Format max_write_lock_countSystem Variable Name max_write_lock_countVariable Scope Global Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 4294967295Range 1 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 18446744073709547520Range 1 .. 18446744073709547520After this many write locks, permit some pending read lock requests to be processed in between.
-
System Variable Name metadata_locks_cache_sizeVariable Scope Global Dynamic Variable No Permitted Values Type numericDefault 1024Range 1 .. 1048576The size of the metadata locks cache. The server uses this cache to avoid creation and destruction of synchronization objects. This is particularly helpful on systems where such operations are expensive, such as Windows XP.
-
System Variable Name metadata_locks_hash_instancesVariable Scope Global Dynamic Variable No Permitted Values Type numericDefault 8Range 1 .. 1024The set of metadata locks can be partitioned into separate hashes to permit connections accessing different objects to use different locking hashes and reduce contention. The
metadata_locks_hash_instancessystem variable specifies the number of hashes (default 8). -
Command-Line Format --min-examined-row-limit=#Option-File Format min-examined-row-limitSystem Variable Name min_examined_row_limitVariable Scope Global, Session Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 0Range 0 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 0Range 0 .. 18446744073709547520Queries that examine fewer than this number of rows are not logged to the slow query log.
-
Command-Line Format --myisam_data_pointer_size=#Option-File Format myisam_data_pointer_sizeSystem Variable Name myisam_data_pointer_sizeVariable Scope Global Dynamic Variable Yes Permitted Values Type numericDefault 6Range 2 .. 7The default pointer size in bytes, to be used by
CREATE TABLEforMyISAMtables when noMAX_ROWSoption is specified. This variable cannot be less than 2 or larger than 7. The default value is 6. See Section C.5.2.12, “The table is full”. -
Command-Line Format --myisam_max_sort_file_size=#Option-File Format myisam_max_sort_file_sizeSystem Variable Name myisam_max_sort_file_sizeVariable Scope Global Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 2147483648Permitted Values Platform Bit Size 64Type numericDefault 9223372036854775807The maximum size of the temporary file that MySQL is permitted to use while re-creating a
MyISAMindex (duringREPAIR TABLE,ALTER TABLE, orLOAD DATA INFILE). If the file size would be larger than this value, the index is created using the key cache instead, which is slower. The value is given in bytes.The default value is 2GB. If
MyISAMindex files exceed this size and disk space is available, increasing the value may help performance. The space must be available in the file system containing the directory where the original index file is located. -
Command-Line Format --myisam_mmap_size=#Option-File Format myisam_mmap_sizeSystem Variable Name myisam_mmap_sizeVariable Scope Global Dynamic Variable No Permitted Values Platform Bit Size 32Type numericDefault 4294967295Range 7 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 18446744073709547520Range 7 .. 18446744073709547520The maximum amount of memory to use for memory mapping compressed
MyISAMfiles. If many compressedMyISAMtables are used, the value can be decreased to reduce the likelihood of memory-swapping problems. -
System Variable Name myisam_recover_optionsVariable Scope Global Dynamic Variable No The value of the
--myisam-recover-optionsoption. See Section 5.1.3, “Server Command Options”. -
Command-Line Format --myisam_repair_threads=#Option-File Format myisam_repair_threadsSystem Variable Name myisam_repair_threadsVariable Scope Global, Session Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 1Range 1 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 1Range 1 .. 18446744073709547520If this value is greater than 1,
MyISAMtable indexes are created in parallel (each index in its own thread) during theRepair by sortingprocess. The default value is 1.NoteMulti-threaded repair is still beta-quality code.
-
Command-Line Format --myisam_sort_buffer_size=#Option-File Format myisam_sort_buffer_sizeSystem Variable Name myisam_sort_buffer_sizeVariable Scope Global, Session Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 8388608Range 4096 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 8388608Range 4096 .. 18446744073709547520The size of the buffer that is allocated when sorting
MyISAMindexes during aREPAIR TABLEor when creating indexes withCREATE INDEXorALTER TABLE.The maximum permissible setting for
myisam_sort_buffer_sizeis 4GB. Values larger than 4GB are permitted for 64-bit platforms (except 64-bit Windows, for which large values are truncated to 4GB with a warning). -
Command-Line Format --myisam_stats_method=nameOption-File Format myisam_stats_methodSystem Variable Name myisam_stats_methodVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type enumerationValid Values nulls_equalnulls_unequalnulls_ignoredHow the server treats
NULLvalues when collecting statistics about the distribution of index values forMyISAMtables. This variable has three possible values,nulls_equal,nulls_unequal, andnulls_ignored. Fornulls_equal, allNULLindex values are considered equal and form a single value group that has a size equal to the number ofNULLvalues. Fornulls_unequal,NULLvalues are considered unequal, and eachNULLforms a distinct value group of size 1. Fornulls_ignored,NULLvalues are ignored.The method that is used for generating table statistics influences how the optimizer chooses indexes for query execution, as described in Section 8.3.7, “
InnoDBandMyISAMIndex Statistics Collection”. -
Command-Line Format --myisam_use_mmapOption-File Format myisam_use_mmapSystem Variable Name myisam_use_mmapVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanDefault OFFUse memory mapping for reading and writing
MyISAMtables. -
System Variable Name named_pipeVariable Scope Global Dynamic Variable No Platform Specific windows Permitted Values Type (windows) booleanDefault OFF(Windows only.) Indicates whether the server supports connections over named pipes.
-
Command-Line Format --net_buffer_length=#Option-File Format net_buffer_lengthSystem Variable Name net_buffer_lengthVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 16384Range 1024 .. 1048576Each client thread is associated with a connection buffer and result buffer. Both begin with a size given by
net_buffer_lengthbut are dynamically enlarged up tomax_allowed_packetbytes as needed. The result buffer shrinks tonet_buffer_lengthafter each SQL statement.This variable should not normally be changed, but if you have very little memory, you can set it to the expected length of statements sent by clients. If statements exceed this length, the connection buffer is automatically enlarged. The maximum value to which
net_buffer_lengthcan be set is 1MB.The session value of this variable is read only.
-
Command-Line Format --net_read_timeout=#Option-File Format net_read_timeoutSystem Variable Name net_read_timeoutVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 30Min Value 1The number of seconds to wait for more data from a connection before aborting the read. When the server is reading from the client,
net_read_timeoutis the timeout value controlling when to abort. When the server is writing to the client,net_write_timeoutis the timeout value controlling when to abort. See alsoslave_net_timeout. -
Command-Line Format --net_retry_count=#Option-File Format net_retry_countSystem Variable Name net_retry_countVariable Scope Global, Session Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 10Range 1 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 10Range 1 .. 18446744073709547520If a read or write on a communication port is interrupted, retry this many times before giving up. This value should be set quite high on FreeBSD because internal interrupts are sent to all threads.
-
Command-Line Format --net_write_timeout=#Option-File Format net_write_timeoutSystem Variable Name net_write_timeoutVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 60Min Value 1The number of seconds to wait for a block to be written to a connection before aborting the write. See also
net_read_timeout. -
Command-Line Format --new-nOption-File Format newSystem Variable Name newVariable Scope Global, Session Dynamic Variable Yes Disabled by skip-newPermitted Values Type booleanDefault FALSEThis variable was used in MySQL 4.0 to turn on some 4.1 behaviors, and is retained for backward compatibility. In MySQL 5.7, its value is always
OFF. -
Command-Line Format --oldOption-File Format oldSystem Variable Name oldVariable Scope Global Dynamic Variable No oldis a compatibility variable. It is disabled by default, but can be enabled at startup to revert the server to behaviors present in older versions.Currently, when
oldis enabled, it changes the default scope of index hints to that used prior to MySQL 5.1.17. That is, index hints with noFORclause apply only to how indexes are used for row retrieval and not to resolution ofORDER BYorGROUP BYclauses. (See Section 13.2.9.3, “Index Hint Syntax”.) Take care about enabling this in a replication setup. With statement-based binary logging, having different modes for the master and slaves might lead to replication errors. -
Command-Line Format --old-alter-tableOption-File Format old-alter-tableSystem Variable Name old_alter_tableVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type booleanDefault OFFWhen this variable is enabled, the server does not use the optimized method of processing an
ALTER TABLEoperation. It reverts to using a temporary table, copying over the data, and then renaming the temporary table to the original, as used by MySQL 5.0 and earlier. For more information on the operation ofALTER TABLE, see Section 13.1.6, “ALTER TABLESyntax”. -
System Variable Name old_passwordsVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type enumerationDefault 0Valid Values 012This variable controls the password hashing method used by the
PASSWORD()function and for theIDENTIFIED BYclause of theCREATE USERandGRANTstatements.If the
--default-authentication-pluginoption is given at server startup, the server setsold_passwordsto the value that is consistent with the password hashing method required by the default plugin.The following table shows the permitted values of
old_passwords, the password hashing method for each value, and which authentication plugins use passwords hashed with each method.Value Password Hashing Format Intended Use 0 MySQL 4.1 native hashing Accounts that authenticate with the mysql_native_passwordplugin1 Pre-4.1 (“old”) hashing Accounts that authenticate with the mysql_old_passwordplugin2 SHA-256 hashing Accounts that authenticate with the sha256_passwordpluginIf
old_passwords=1,PASSWORD('returns the same value asstr')OLD_PASSWORD('. The latter function is not affected by the value ofstr')old_passwords.If you set
old_passwords=2, follow the instructions for using thesha256_passwordplugin at Section 6.3.7.4, “The SHA-256 Authentication Plugin”.For information about authentication plugins and hashing formats, see Section 6.3.7, “Pluggable Authentication”, and Section 6.1.2.4, “Password Hashing in MySQL”.
NotePasswords that use the pre-4.1 hashing method are less secure than passwords that use the native password hashing method and should be avoided. Pre-4.1 passwords are deprecated and support for them will be removed in a future MySQL release.
-
Command-Line Format --open-files-limit=#Option-File Format open-files-limitSystem Variable Name open_files_limitVariable Scope Global Dynamic Variable No Permitted Values Type numericDefault -1 (autosized)Range 0 .. 65535The number of files that the operating system permits mysqld to open. The value of this variable at runtime is the real value permitted by the system and might be different from the value you specify at server startup. The value is 0 on systems where MySQL cannot change the number of open files.
The effective
open_files_limitvalue is based on the value specified at system startup (if any) and the values ofmax_connectionsandtable_open_cache, using these formulas:1) 10 + max_connections + (table_open_cache * 2) 2) max_connections * 5 3) open_files_limit value specified at startup, 5000 if none
The server attempts to obtain the number of file descriptors using the maximum of those three values. If that many descriptors cannot be obtained, the server attempts to obtain as many as the system will permit.
-
Command-Line Format --optimizer_prune_level[=#]Option-File Format optimizer_prune_levelSystem Variable Name optimizer_prune_levelVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type booleanDefault 1Controls the heuristics applied during query optimization to prune less-promising partial plans from the optimizer search space. A value of 0 disables heuristics so that the optimizer performs an exhaustive search. A value of 1 causes the optimizer to prune plans based on the number of rows retrieved by intermediate plans.
-
Command-Line Format --optimizer_search_depth[=#]Option-File Format optimizer_search_depthSystem Variable Name optimizer_search_depthVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 62Range 0 .. 62The maximum depth of search performed by the query optimizer. Values larger than the number of relations in a query result in better query plans, but take longer to generate an execution plan for a query. Values smaller than the number of relations in a query return an execution plan quicker, but the resulting plan may be far from being optimal. If set to 0, the system automatically picks a reasonable value.
-
Command-Line Format --optimizer_switch=valueOption-File Format optimizer_switchSystem Variable Name optimizer_switchVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type setValid Values batched_key_access={on|off}block_nested_loop={on|off}engine_condition_pushdown={on|off}firstmatch={on|off}index_condition_pushdown={on|off}index_merge={on|off}index_merge_intersection={on|off}index_merge_sort_union={on|off}index_merge_union={on|off}loosescan={on|off}materialization={on|off}mrr={on|off}mrr_cost_based={on|off}semijoin={on|off}subquery_materialization_cost_based={on|off}use_index_extensions={on|off}The
optimizer_switchsystem variable enables control over optimizer behavior. The value of this variable is a set of flags, each of which has a value ofonoroffto indicate whether the corresponding optimizer behavior is enabled or disabled. This variable has global and session values and can be changed at runtime. The global default can be set at server startup.To see the current set of optimizer flags, select the variable value:
mysql>
SELECT @@optimizer_switch\G*************************** 1. row *************************** @@optimizer_switch: index_merge=on,index_merge_union=on, index_merge_sort_union=on, index_merge_intersection=on, engine_condition_pushdown=on, index_condition_pushdown=on, mrr=on,mrr_cost_based=on, block_nested_loop=on,batched_key_access=off, materialization=on,semijoin=on,loosescan=on, firstmatch=on, subquery_materialization_cost_based=on, use_index_extensions=onFor more information about the syntax of this variable and the optimizer behaviors that it controls, see Section 8.8.6.2, “Controlling Switchable Optimizations”.
-
System Variable Name optimizer_traceVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type stringThis variable controls optimizer tracing. For details, see MySQL Internals: Tracing the Optimizer.
-
System Variable Name optimizer_trace_featuresVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type stringThis variable enables or disables selected optimizer tracing features. For details, see MySQL Internals: Tracing the Optimizer.
-
System Variable Name optimizer_trace_limitVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 1The maximum number of optimizer traces to display. For details, see MySQL Internals: Tracing the Optimizer.
-
System Variable Name optimizer_trace_max_mem_sizeVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 16384The maximum cumulative size of stored optimizer traces. For details, see MySQL Internals: Tracing the Optimizer.
-
System Variable Name optimizer_trace_offsetVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault -1The offset of optimizer traces to display. For details, see MySQL Internals: Tracing the Optimizer.
performance_schema_xxxPerformance Schema system variables are listed in Section 20.12, “Performance Schema System Variables”. These variables may be used to configure Performance Schema operation.
-
Command-Line Format --pid-file=file_nameOption-File Format pid-fileSystem Variable Name pid_fileVariable Scope Global Dynamic Variable No Permitted Values Type file nameThe path name of the process ID (PID) file. This variable can be set with the
--pid-fileoption. -
Command-Line Format --plugin_dir=pathOption-File Format plugin_dirSystem Variable Name plugin_dirVariable Scope Global Dynamic Variable No Permitted Values Type directory nameDefault BASEDIR/lib/pluginThe path name of the plugin directory.
If the plugin directory is writable by the server, it may be possible for a user to write executable code to a file in the directory using
SELECT ... INTO DUMPFILE. This can be prevented by makingplugin_dirread only to the server or by setting--secure-file-privto a directory whereSELECTwrites can be made safely. -
Command-Line Format --port=#-POption-File Format portSystem Variable Name portVariable Scope Global Dynamic Variable No Permitted Values Type numericDefault 3306Range 0 .. 65535The number of the port on which the server listens for TCP/IP connections. This variable can be set with the
--portoption. -
Command-Line Format --preload_buffer_size=#Option-File Format preload_buffer_sizeSystem Variable Name preload_buffer_sizeVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 32768Range 1024 .. 1073741824The size of the buffer that is allocated when preloading indexes.
If set to 0 or
OFF(the default), statement profiling is disabled. If set to 1 orON, statement profiling is enabled and theSHOW PROFILEandSHOW PROFILESstatements provide access to profiling information. See Section 13.7.5.30, “SHOW PROFILESSyntax”.This variable is deprecated and will be removed in a future MySQL release.
The number of statements for which to maintain profiling information if
profilingis enabled. The default value is 15. The maximum value is 100. Setting the value to 0 effectively disables profiling. See Section 13.7.5.30, “SHOW PROFILESSyntax”.This variable is deprecated and will be removed in a future MySQL release.
-
System Variable Name protocol_versionVariable Scope Global Dynamic Variable No Permitted Values Type numericThe version of the client/server protocol used by the MySQL server.
-
System Variable Name proxy_userVariable Scope Session Dynamic Variable No Permitted Values Type stringIf the current client is a proxy for another user, this variable is the proxy user account name. Otherwise, this variable is
NULL. See Section 6.3.8, “Proxy Users”. -
System Variable Name pseudo_slave_modeVariable Scope Session Dynamic Variable Yes Permitted Values Type numericThis variable is for internal server use.
-
System Variable Name pseudo_thread_idVariable Scope Session Dynamic Variable Yes Permitted Values Type numericThis variable is for internal server use.
-
Command-Line Format --query_alloc_block_size=#Option-File Format query_alloc_block_sizeSystem Variable Name query_alloc_block_sizeVariable Scope Global, Session Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 8192Range 1024 .. 4294967295Block Size 1024Permitted Values Platform Bit Size 64Type numericDefault 8192Range 1024 .. 18446744073709547520Block Size 1024The allocation size of memory blocks that are allocated for objects created during statement parsing and execution. If you have problems with memory fragmentation, it might help to increase this parameter.
-
Command-Line Format --query_cache_limit=#Option-File Format query_cache_limitSystem Variable Name query_cache_limitVariable Scope Global Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 1048576Range 0 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 1048576Range 0 .. 18446744073709547520Do not cache results that are larger than this number of bytes. The default value is 1MB.
-
Command-Line Format --query_cache_min_res_unit=#Option-File Format query_cache_min_res_unitSystem Variable Name query_cache_min_res_unitVariable Scope Global Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 4096Range 512 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 4096Range 512 .. 18446744073709547520The minimum size (in bytes) for blocks allocated by the query cache. The default value is 4096 (4KB). Tuning information for this variable is given in Section 8.9.3.3, “Query Cache Configuration”.
-
Command-Line Format --query_cache_size=#Option-File Format query_cache_sizeSystem Variable Name query_cache_sizeVariable Scope Global Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 1048576Range 0 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 1048576Range 0 .. 18446744073709547520The amount of memory allocated for caching query results. By default, the query cache is disabled. This is achieved using a default value of 1M, with a default for
query_cache_typeof 0. (To reduce overhead significantly if you set the size to 0, you should also start the server withquery_cache_type=0.The permissible values are multiples of 1024; other values are rounded down to the nearest multiple. Note that
query_cache_sizebytes of memory are allocated even ifquery_cache_typeis set to 0. See Section 8.9.3.3, “Query Cache Configuration”, for more information.The query cache needs a minimum size of about 40KB to allocate its structures. (The exact size depends on system architecture.) If you set the value of
query_cache_sizetoo small, a warning will occur, as described in Section 8.9.3.3, “Query Cache Configuration”. -
Command-Line Format --query_cache_type=#Option-File Format query_cache_typeSystem Variable Name query_cache_typeVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type enumerationDefault 0Valid Values 012Set the query cache type. Setting the
GLOBALvalue sets the type for all clients that connect thereafter. Individual clients can set theSESSIONvalue to affect their own use of the query cache. Possible values are shown in the following table.Option Description 0orOFFDo not cache results in or retrieve results from the query cache. Note that this does not deallocate the query cache buffer. To do that, you should set query_cache_sizeto 0.1orONCache all cacheable query results except for those that begin with SELECT SQL_NO_CACHE.2orDEMANDCache results only for cacheable queries that begin with SELECT SQL_CACHE.This variable defaults to
OFF.If the server is started with
query_cache_typeset to 0, it does not acquire the query cache mutex at all, which means that the query cache cannot be enabled at runtime and there is reduced overhead in query execution. -
Command-Line Format --query_cache_wlock_invalidateOption-File Format query_cache_wlock_invalidateSystem Variable Name query_cache_wlock_invalidateVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type booleanDefault FALSENormally, when one client acquires a
WRITElock on aMyISAMtable, other clients are not blocked from issuing statements that read from the table if the query results are present in the query cache. Setting this variable to 1 causes acquisition of aWRITElock for a table to invalidate any queries in the query cache that refer to the table. This forces other clients that attempt to access the table to wait while the lock is in effect. -
Command-Line Format --query_prealloc_size=#Option-File Format query_prealloc_sizeSystem Variable Name query_prealloc_sizeVariable Scope Global, Session Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 8192Range 8192 .. 4294967295Block Size 1024Permitted Values Platform Bit Size 64Type numericDefault 8192Range 8192 .. 18446744073709547520Block Size 1024The size of the persistent buffer used for statement parsing and execution. This buffer is not freed between statements. If you are running complex queries, a larger
query_prealloc_sizevalue might be helpful in improving performance, because it can reduce the need for the server to perform memory allocation during query execution operations. The
rand_seed1andrand_seed2variables exist as session variables only, and can be set but not read. The variables—but not their values—are shown in the output ofSHOW VARIABLES.The purpose of these variables is to support replication of the
RAND()function. For statements that invokeRAND(), the master passes two values to the slave, where they are used to seed the random number generator. The slave uses these values to set the session variablesrand_seed1andrand_seed2so thatRAND()on the slave generates the same value as on the master.See the description for
rand_seed1.-
Command-Line Format --range_alloc_block_size=#Option-File Format range_alloc_block_sizeSystem Variable Name range_alloc_block_sizeVariable Scope Global, Session Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 4096Range 4096 .. 4294967295Block Size 1024Permitted Values Platform Bit Size 64Type numericDefault 4096Range 4096 .. 18446744073709547520Block Size 1024The size of blocks that are allocated when doing range optimization.
-
Command-Line Format --read_buffer_size=#Option-File Format read_buffer_sizeSystem Variable Name read_buffer_sizeVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 131072Range 8200 .. 2147479552Each thread that does a sequential scan for a
MyISAMtable allocates a buffer of this size (in bytes) for each table it scans. If you do many sequential scans, you might want to increase this value, which defaults to 131072. The value of this variable should be a multiple of 4KB. If it is set to a value that is not a multiple of 4KB, its value will be rounded down to the nearest multiple of 4KB.This option is also used in the following context for all search engines:
For caching the indexes in a temporary file (not a temporary table), when sorting rows for
ORDER BY.For bulk insert into partitions.
For caching results of nested queries.
and in one other storage engine-specific way: to determine the memory block size for
MEMORYtables.The maximum permissible setting for
read_buffer_sizeis 2GB.For more information about memory use during different operations, see Section 8.11.4.1, “How MySQL Uses Memory”.
-
Command-Line Format --read-onlyOption-File Format read_onlySystem Variable Name read_onlyVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanDefault falseThis variable is off by default. When it is enabled, the server permits no updates except from users that have the
SUPERprivilege or (on a slave server) from updates performed by slave threads. In replication setups, it can be useful to enableread_onlyon slave servers to ensure that slaves accept updates only from the master server and not from clients.read_onlydoes not apply toTEMPORARYtables, nor does it prevent the server from inserting rows into the log tables (see Section 5.2.1, “Selecting General Query and Slow Query Log Output Destinations”). This variable does not prevent the use ofANALYZE TABLEorOPTIMIZE TABLEstatements because its purpose is to prevent changes to table structure or contents. Analysis and optimization do not qualify as such changes. This means, for example, that consistency checks on read-only slaves can be performed with mysqlcheck --all-databases --analyze.read_onlyexists only as aGLOBALvariable, so changes to its value require theSUPERprivilege. Changes toread_onlyon a master server are not replicated to slave servers. The value can be set on a slave server independent of the setting on the master.ImportantIn MySQL 5.7, enabling
read_onlyprevents the use of theSET PASSWORDstatement by any user not having theSUPERprivilege. This is not necessarily the case for all MySQL release series. When replicating from one MySQL release series to another (for example, from a MySQL 5.0 master to a MySQL 5.1 or later slave), you should check the documentation for the versions running on both master and slave to determine whether the behavior ofread_onlyin this regard is or is not the same, and, if it is different, whether this has an impact on your applications.The following conditions apply:
If you attempt to enable
read_onlywhile you have any explicit locks (acquired withLOCK TABLES) or have a pending transaction, an error occurs.If you attempt to enable
read_onlywhile other clients hold explicit table locks or have pending transactions, the attempt blocks until the locks are released and the transactions end. While the attempt to enableread_onlyis pending, requests by other clients for table locks or to begin transactions also block untilread_onlyhas been set.read_onlycan be enabled while you hold a global read lock (acquired withFLUSH TABLES WITH READ LOCK) because that does not involve table locks.
In MySQL 5.7, attempts to set
read_onlyblock for active transactions that hold metadata locks until those transactions end. -
Command-Line Format --read_rnd_buffer_size=#Option-File Format read_rnd_buffer_sizeSystem Variable Name read_rnd_buffer_sizeVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 262144Range 8200 .. 4294967295When reading rows from a
MyISAMtable in sorted order following a key-sorting operation, the rows are read through this buffer to avoid disk seeks. See Section 8.2.1.15, “ORDER BYOptimization”. Setting the variable to a large value can improveORDER BYperformance by a lot. However, this is a buffer allocated for each client, so you should not set the global variable to a large value. Instead, change the session variable only from within those clients that need to run large queries.The maximum permissible setting for
read_rnd_buffer_sizeis 2GB.For more information about memory use during different operations, see Section 8.11.4.1, “How MySQL Uses Memory”.
-
Command-Line Format --relay_log_purgeOption-File Format relay_log_purgeSystem Variable Name relay_log_purgeVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanDefault TRUEDisables or enables automatic purging of relay log files as soon as they are not needed any more. The default value is 1 (
ON). -
Command-Line Format --relay_log_space_limit=#Option-File Format relay_log_space_limitSystem Variable Name relay_log_space_limitVariable Scope Global Dynamic Variable No Permitted Values Platform Bit Size 32Type numericDefault 0Range 0 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 0Range 0 .. 18446744073709547520The maximum amount of space to use for all relay logs.
-
Command-Line Format --report-host=host_nameOption-File Format report-hostSystem Variable Name report_hostVariable Scope Global Dynamic Variable No Permitted Values Type stringThe value of the
--report-hostoption. -
Command-Line Format --report-password=nameOption-File Format report-passwordSystem Variable Name report_passwordVariable Scope Global Dynamic Variable No Permitted Values Type stringThe value of the
--report-passwordoption. Not the same as the password used for the MySQL replication user account. -
Command-Line Format --report-port=#Option-File Format report-portSystem Variable Name report_portVariable Scope Global Dynamic Variable No Permitted Values Type numericDefault [slave_port]Range 0 .. 65535The value of the
--report-portoption. -
Command-Line Format --report-user=nameOption-File Format report-userSystem Variable Name report_userVariable Scope Global Dynamic Variable No Permitted Values Type stringThe value of the
--report-useroption. Not the same as the name for the MySQL replication user account. -
System Variable Name rpl_semi_sync_master_enabledVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanDefault OFFControls whether semisynchronous replication is enabled on the master. To enable or disable the plugin, set this variable to
ONorOFF(or 1 or 0), respectively. The default isOFF.This variable is available only if the master-side semisynchronous replication plugin is installed.
-
System Variable Name rpl_semi_sync_master_timeoutVariable Scope Global Dynamic Variable Yes Permitted Values Type numericDefault 10000A value in milliseconds that controls how long the master waits on a commit for acknowledgment from a slave before timing out and reverting to asynchronous replication. The default value is 10000 (10 seconds).
This variable is available only if the master-side semisynchronous replication plugin is installed.
rpl_semi_sync_master_trace_levelSystem Variable Name rpl_semi_sync_master_trace_levelVariable Scope Global Dynamic Variable Yes Permitted Values Type numericDefault 32The semisynchronous replication debug trace level on the master. Currently, four levels are defined:
1 = general level (for example, time function failures)
16 = detail level (more verbose information)
32 = net wait level (more information about network waits)
64 = function level (information about function entry and exit)
This variable is available only if the master-side semisynchronous replication plugin is installed.
rpl_semi_sync_master_wait_no_slaveSystem Variable Name rpl_semi_sync_master_wait_no_slaveVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanDefault ONWith semisynchronous replication, for each transaction, the master waits until timeout for acknowledgment of receipt from some semisynchronous slave. If no response occurs during this period, the master reverts to normal replication. This variable controls whether the master waits for the timeout to expire before reverting to normal replication even if the slave count drops to zero during the timeout period.
If the value is
ON(the default), it is permissible for the slave count to drop to zero during the timeout period (for example, if slaves disconnect). The master still waits for the timeout, so as long as some slave reconnects and acknowledges the transaction within the timeout interval, semisynchronous replication continues.If the value is
OFF, the master reverts to normal replication if the slave count drops to zero during the timeout period.This variable is available only if the master-side semisynchronous replication plugin is installed.
rpl_semi_sync_master_wait_pointIntroduced 5.7.2 System Variable Name rpl_semi_sync_master_wait_pointVariable Scope Global Dynamic Variable Yes Permitted Values Type enumerationDefault AFTER_SYNCValid Values AFTER_SYNCAFTER_COMMITThis variable controls the point at which a semisynchronous replication master waits for slave acknowledgment of transaction receipt before returning a status to the client that committed the transaction. These values are permitted:
AFTER_SYNC(the default): The master writes each transaction to its binary log and the slave, and syncs the binary log to disk. The master waits for slave acknowledgment of transaction receipt after the sync. Upon receiving acknowledgment, the master commits the transaction to the storage engine and returns a result to the client, which then can proceed.AFTER_COMMIT: The master writes each transaction to its binary log and the slave, syncs the binary log, and commits the transaction to the storage engine. The master waits for slave acknowledgment of transaction receipt after the commit. Upon receiving acknowledgment, the master returns a result to the client, which then can proceed.
The replication characteristics of these settings differ as follows:
With
AFTER_SYNC, all clients see the committed transaction at the same time: After it has been acknowledged by the slave and committed to the storage engine on the master. Thus, all clients see the same data on the master.In the event of master failure, all transactions committed on the master have been replicated to the slave (saved to its relay log). A crash of the master and failover to the slave is lossless because the slave is up to date.
With
AFTER_COMMIT, the client issuing the transaction gets a return status only after the server commits to the storage engine and receives slave acknowledgement. After the commit and before slave acknowledgment, other clients can see the committed transaction before the committing client.If something goes wrong such that the slave does not process the transaction, then in the event of a master crash and failover to the slave, it is possible that such clients will see a loss of data relative to what they saw on the master.
This variable is available only if the master-side semisynchronous replication plugin is installed.
rpl_semi_sync_master_wait_pointwas added in MySQL 5.7.2. For older versions, semisynchronous master behavior is equivalent to a setting ofAFTER_COMMIT.This change introduces a version compatibility constraint because it increments the semisynchronous interface version: Servers for MySQL 5.7.2 and up do not work with semisynchronous replication plugins from older versions, nor do servers from older versions work with semisynchronous replication plugins for MySQL 5.7.2 and up.
-
System Variable Name rpl_semi_sync_slave_enabledVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanDefault OFFControls whether semisynchronous replication is enabled on the slave. To enable or disable the plugin, set this variable to
ONorOFF(or 1 or 0), respectively. The default isOFF.This variable is available only if the slave-side semisynchronous replication plugin is installed.
rpl_semi_sync_slave_trace_levelSystem Variable Name rpl_semi_sync_slave_trace_levelVariable Scope Global Dynamic Variable Yes Permitted Values Type numericDefault 32The semisynchronous replication debug trace level on the slave. See
rpl_semi_sync_master_trace_levelfor the permissible values.This variable is available only if the slave-side semisynchronous replication plugin is installed.
-
Command-Line Format --secure-authOption-File Format secure-authSystem Variable Name secure_authVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanDefault ONIf this variable is enabled, the server blocks connections by clients that attempt to use accounts that have passwords stored in the old (pre-4.1) format.
Enable this variable to prevent all use of passwords employing the old format (and hence insecure communication over the network). This variable enabled by default.
Server startup fails with an error if this variable is enabled and the privilege tables are in pre-4.1 format. See Section C.5.2.4, “
Client does not support authentication protocol”.NotePasswords that use the pre-4.1 hashing method are less secure than passwords that use the native password hashing method and should be avoided. Pre-4.1 passwords are deprecated and support for them will be removed in a future MySQL release.
-
Command-Line Format --secure-file-priv=pathOption-File Format secure-file-privSystem Variable Name secure_file_privVariable Scope Global Dynamic Variable No Permitted Values Type stringBy default, this variable is empty. If set to the name of a directory, it limits the effect of the
LOAD_FILE()function and theLOAD DATAandSELECT ... INTO OUTFILEstatements to work only with files in that directory. -
Command-Line Format --server-id=#Option-File Format server-idSystem Variable Name server_idVariable Scope Global Dynamic Variable Yes Permitted Values Type numericDefault 0Range 0 .. 4294967295The server ID, used in replication to give each master and slave a unique identity. This variable is set by the
--server-idoption. For each server participating in replication, you should pick a positive integer in the range from 1 to 232 – 1 to act as that server's ID. sha256_password_private_key_pathSystem Variable Name sha256_password_private_key_pathVariable Scope Global Dynamic Variable No Permitted Values Type file nameDefault private_key.pemThe path name of the RSA private key file for the
sha256_passwordauthentication plugin. If the file is named as a relative path, it is interpreted relative to the server data directory. The file must be in PEM format. Because this file stores a private key, its access mode should be restricted so that only the MySQL server can read it.For information about
sha256_password, including instructions for creating the RSA key files, see Section 6.3.7.4, “The SHA-256 Authentication Plugin”.This variable is available only if MySQL was built using OpenSSL.
sha256_password_public_key_pathSystem Variable Name sha256_password_public_key_pathVariable Scope Global Dynamic Variable No Permitted Values Type file nameDefault public_key.pemThe path name of the RSA public key file for the
sha256_passwordauthentication plugin. If the file is named as a relative path, it is interpreted relative to the server data directory. The file must be in PEM format. Because this file stores a public key, copies can be freely distributed to client users. (Clients that explicitly specify a public key when connecting to the server using RSA password encryption must use the same public key as that used by the server.)For information about
sha256_password, including instructions for creating the RSA key files and how clients specify the RSA public key, see Section 6.3.7.4, “The SHA-256 Authentication Plugin”.This variable is available only if MySQL was built using OpenSSL.
-
System Variable Name shared_memoryVariable Scope Global Dynamic Variable No Platform Specific windows (Windows only.) Whether the server permits shared-memory connections.
-
System Variable Name shared_memory_base_nameVariable Scope Global Dynamic Variable No Platform Specific windows (Windows only.) The name of shared memory to use for shared-memory connections. This is useful when running multiple MySQL instances on a single physical machine. The default name is
MYSQL. The name is case sensitive. -
Command-Line Format --skip-external-lockingOption-File Format skip_external_lockingSystem Variable Name skip_external_lockingVariable Scope Global Dynamic Variable No Permitted Values Type booleanDefault ONThis is
OFFif mysqld uses external locking (system locking),ONif external locking is disabled. This affects onlyMyISAMtable access.This variable is set by the
--external-lockingor--skip-external-lockingoption. External locking has been disabled by default as of MySQL 4.0.External locking affects only
MyISAMtable access. For more information, including conditions under which it can and cannot be used, see Section 8.10.5, “External Locking”. -
Command-Line Format --skip-name-resolveOption-File Format skip-name-resolveSystem Variable Name skip_name_resolveVariable Scope Global Dynamic Variable No Permitted Values Type booleanDefault OFFThis variable is set from the value of the
--skip-name-resolveoption. If it isON, mysqld resolves host names when checking client connections. IfOFF, mysqld uses only IP numbers and allHostcolumn values in the grant tables must be IP addresses orlocalhost. See Section 8.11.5.2, “DNS Lookup Optimization and the Host Cache”. -
Command-Line Format --skip-networkingOption-File Format skip-networkingSystem Variable Name skip_networkingVariable Scope Global Dynamic Variable No This is
ONif the server permits only local (non-TCP/IP) connections. On Unix, local connections use a Unix socket file. On Windows, local connections use a named pipe or shared memory. This variable can be set toONwith the--skip-networkingoption. -
Command-Line Format --skip-show-databaseOption-File Format skip-show-databaseSystem Variable Name skip_show_databaseVariable Scope Global Dynamic Variable No This prevents people from using the
SHOW DATABASESstatement if they do not have theSHOW DATABASESprivilege. This can improve security if you have concerns about users being able to see databases belonging to other users. Its effect depends on theSHOW DATABASESprivilege: If the variable value isON, theSHOW DATABASESstatement is permitted only to users who have theSHOW DATABASESprivilege, and the statement displays all database names. If the value isOFF,SHOW DATABASESis permitted to all users, but displays the names of only those databases for which the user has theSHOW DATABASESor other privilege. (Note that any global privilege is considered a privilege for the database.) -
Command-Line Format --slow_launch_time=#Option-File Format slow_launch_timeSystem Variable Name slow_launch_timeVariable Scope Global Dynamic Variable Yes Permitted Values Type numericDefault 2If creating a thread takes longer than this many seconds, the server increments the
Slow_launch_threadsstatus variable. -
Command-Line Format --slow-query-logOption-File Format slow-query-logSystem Variable Name slow_query_logVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanDefault OFFWhether the slow query log is enabled. The value can be 0 (or
OFF) to disable the log or 1 (orON) to enable the log. The default value depends on whether the--slow_query_logoption is given. The destination for log output is controlled by thelog_outputsystem variable; if that value isNONE, no log entries are written even if the log is enabled.“Slow” is determined by the value of the
long_query_timevariable. See Section 5.2.5, “The Slow Query Log”. -
Command-Line Format --slow-query-log-file=file_nameOption-File Format slow_query_log_fileSystem Variable Name slow_query_log_fileVariable Scope Global Dynamic Variable Yes Permitted Values Type file nameThe name of the slow query log file. The default value is
host_name-slow.log--slow_query_log_fileoption. -
Command-Line Format --socket=nameOption-File Format socketSystem Variable Name socketVariable Scope Global Dynamic Variable No Permitted Values Type file nameDefault /tmp/mysql.sockOn Unix platforms, this variable is the name of the socket file that is used for local client connections. The default is
/tmp/mysql.sock. (For some distribution formats, the directory might be different, such as/var/lib/mysqlfor RPMs.)On Windows, this variable is the name of the named pipe that is used for local client connections. The default value is
MySQL(not case sensitive). -
Command-Line Format --sort_buffer_size=#Option-File Format sort_buffer_sizeSystem Variable Name sort_buffer_sizeVariable Scope Global, Session Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 262144Range 32768 .. 4294967295Permitted Values Platform Bit Size 64Type numericDefault 262144Range 32768 .. 18446744073709547520Each session that needs to do a sort allocates a buffer of this size.
sort_buffer_sizeis not specific to any storage engine and applies in a general manner for optimization. See Section 8.2.1.15, “ORDER BYOptimization”, for example.If you see many
Sort_merge_passesper second inSHOW GLOBAL STATUSoutput, you can consider increasing thesort_buffer_sizevalue to speed upORDER BYorGROUP BYoperations that cannot be improved with query optimization or improved indexing.The optimizer tries to work out how much space is needed but can allocate more, up to the limit. Setting it larger than required globally will slow down most queries that sort. It is best to increase it as a session setting, and only for the sessions that need a larger size. On Linux, there are thresholds of 256KB and 2MB where larger values may significantly slow down memory allocation, so you should consider staying below one of those values. Experiment to find the best value for your workload. See Section C.5.4.4, “Where MySQL Stores Temporary Files”.
The maximum permissible setting for
sort_buffer_sizeis 4GB. Values larger than 4GB are permitted for 64-bit platforms (except 64-bit Windows, for which large values are truncated to 4GB with a warning). -
System Variable Name sql_auto_is_nullVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type booleanDefault 0If this variable is set to 1, then after a statement that successfully inserts an automatically generated
AUTO_INCREMENTvalue, you can find that value by issuing a statement of the following form:SELECT * FROM
tbl_nameWHEREauto_colIS NULLIf the statement returns a row, the value returned is the same as if you invoked the
LAST_INSERT_ID()function. For details, including the return value after a multiple-row insert, see Section 12.14, “Information Functions”. If noAUTO_INCREMENTvalue was successfully inserted, theSELECTstatement returns no row.The behavior of retrieving an
AUTO_INCREMENTvalue by using anIS NULLcomparison is used by some ODBC programs, such as Access. See Section 21.1.7.1.1, “Obtaining Auto-Increment Values”. This behavior can be disabled by settingsql_auto_is_nullto 0.The default value of
sql_auto_is_nullis 0 in MySQL 5.7. -
System Variable Name sql_big_selectsVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type booleanDefault 1If set to 0, MySQL aborts
SELECTstatements that are likely to take a very long time to execute (that is, statements for which the optimizer estimates that the number of examined rows exceeds the value ofmax_join_size). This is useful when an inadvisableWHEREstatement has been issued. The default value for a new connection is 1, which permits allSELECTstatements.If you set the
max_join_sizesystem variable to a value other thanDEFAULT,sql_big_selectsis set to 0. -
System Variable Name sql_buffer_resultVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type booleanDefault 0If set to 1,
sql_buffer_resultforces results fromSELECTstatements to be put into temporary tables. This helps MySQL free the table locks early and can be beneficial in cases where it takes a long time to send results to the client. The default value is 0. -
System Variable Name sql_log_binVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type booleanThis variable controls whether logging to the binary log is done. The default value is 1 (do logging). To change logging for the current session, change the session value of this variable. The session user must have the
SUPERprivilege to set this variable.In MySQL 5.7, it is not possible to set
@@session.sql_log_binwithin a transaction or subquery. (Bug #53437) -
System Variable Name sql_log_offVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type booleanDefault 0This variable controls whether logging to the general query log is done. The default value is 0 (do logging). To change logging for the current session, change the session value of this variable. The session user must have the
SUPERprivilege to set this option. The default value is 0. -
Command-Line Format --sql-mode=nameOption-File Format sql-modeSystem Variable Name sql_modeVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type setDefault NO_ENGINE_SUBSTITUTIONValid Values ALLOW_INVALID_DATESANSI_QUOTESERROR_FOR_DIVISION_BY_ZEROHIGH_NOT_PRECEDENCEIGNORE_SPACENO_AUTO_CREATE_USERNO_AUTO_VALUE_ON_ZERONO_BACKSLASH_ESCAPESNO_DIR_IN_CREATENO_ENGINE_SUBSTITUTIONNO_FIELD_OPTIONSNO_KEY_OPTIONSNO_TABLE_OPTIONSNO_UNSIGNED_SUBTRACTIONNO_ZERO_DATENO_ZERO_IN_DATEONLY_FULL_GROUP_BYPAD_CHAR_TO_FULL_LENGTHPIPES_AS_CONCATREAL_AS_FLOATSTRICT_ALL_TABLESSTRICT_TRANS_TABLESThe current server SQL mode, which can be set dynamically. See Section 5.1.7, “Server SQL Modes”.
NoteMySQL installation programs may configure the SQL mode during the installation process. For example, mysql_install_db creates a default option file named
my.cnfin the base installation directory. This file contains a line that sets the SQL mode; see Section 4.4.3, “mysql_install_db — Initialize MySQL Data Directory”.If the SQL mode differs from the default or from what you expect, check for a setting in an option file that the server reads at startup.
If set to 1 (the default), warnings of
Notelevel incrementwarning_countand the server records them. If set to 0,Notewarnings do not incrementwarning_countand the server does not record them. mysqldump includes output to set this variable to 0 so that reloading the dump file does not produce warnings for events that do not affect the integrity of the reload operation.If set to 1 (the default), the server quotes identifiers for
SHOW CREATE TABLEandSHOW CREATE DATABASEstatements. If set to 0, quoting is disabled. This option is enabled by default so that replication works for identifiers that require quoting. See Section 13.7.5.10, “SHOW CREATE TABLESyntax”, and Section 13.7.5.6, “SHOW CREATE DATABASESyntax”.If set to 1, MySQL aborts
UPDATEorDELETEstatements that do not use a key in theWHEREclause or aLIMITclause. This makes it possible to catchUPDATEorDELETEstatements where keys are not used properly and that would probably change or delete a large number of rows. The default value is 0.-
System Variable Name sql_select_limitVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericThe maximum number of rows to return from
SELECTstatements. The default value for a new connection is the maximum number of rows that the server permits per table. Typical default values are (232)–1 or (264)–1. If you have changed the limit, the default value can be restored by assigning a value ofDEFAULT.If a
SELECThas aLIMITclause, theLIMITtakes precedence over the value ofsql_select_limit. This variable controls whether single-row
INSERTstatements produce an information string if warnings occur. The default is 0. Set the value to 1 to produce an information string.-
Command-Line Format --ssl-ca=nameOption-File Format ssl-caSystem Variable Name ssl_caVariable Scope Global Dynamic Variable No Permitted Values Type file nameThe path to a file with a list of trusted SSL CAs.
-
Command-Line Format --ssl-capath=nameOption-File Format ssl-capathSystem Variable Name ssl_capathVariable Scope Global Dynamic Variable No Permitted Values Type directory nameThe path to a directory that contains trusted SSL CA certificates in PEM format.
-
Command-Line Format --ssl-cert=nameOption-File Format ssl-certSystem Variable Name ssl_certVariable Scope Global Dynamic Variable No Permitted Values Type file nameThe name of the SSL certificate file to use for establishing a secure connection.
-
Command-Line Format --ssl-cipher=nameOption-File Format ssl-cipherSystem Variable Name ssl_cipherVariable Scope Global Dynamic Variable No Permitted Values Type file nameA list of permissible ciphers to use for SSL encryption.
-
Command-Line Format --ssl-crl=nameOption-File Format ssl-crlSystem Variable Name ssl_crlVariable Scope Global Dynamic Variable No Permitted Values Type file nameThe path to a file containing certificate revocation lists in PEM format. Revocation lists work for MySQL distributions compiled against OpenSSL (but not yaSSL).
-
Command-Line Format --ssl-crlpath=nameOption-File Format ssl-crlpathSystem Variable Name ssl_crlpathVariable Scope Global Dynamic Variable No Permitted Values Type directory nameThe path to a directory that contains files containing certificate revocation lists in PEM format. Revocation lists work for MySQL distributions compiled against OpenSSL (but not yaSSL).
-
Command-Line Format --ssl-key=nameOption-File Format ssl-keySystem Variable Name ssl_keyVariable Scope Global Dynamic Variable No Permitted Values Type stringThe name of the SSL key file to use for establishing a secure connection.
-
System Variable Name storage_engineVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type enumerationDefault InnoDBThe default storage engine (table type). To set the storage engine at server startup, use the
--default-storage-engineoption. See Section 5.1.3, “Server Command Options”.This variable is deprecated. Use
default_storage_engineinstead. -
Command-Line Format --stored-program-cache=#Option-File Format stored_program_cacheSystem Variable Name stored_program_cacheVariable Scope Global Dynamic Variable Yes Permitted Values Type numericDefault 256Range 256 .. 524288Sets a soft upper limit for the number of cached stored routines per connection. The value of this variable is specified in terms of the number of stored routines held in each of the two caches maintained by the MySQL Server for, respectively, stored procedures and stored functions.
Whenever a stored routine is executed this cache size is checked before the first or top-level statement in the routine is parsed; if the number of routines of the same type (stored procedures or stored functions according to which is being executed) exceeds the limit specified by this variable, the corresponding cache is flushed and memory previously allocated for cached objects is freed. This allows the cache to be flushed safely, even when there are dependencies between stored routines.
-
Command-Line Format --sync-frmOption-File Format sync_frmSystem Variable Name sync_frmVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanDefault TRUEIf this variable is set to 1, when any nontemporary table is created its
.frmfile is synchronized to disk (usingfdatasync()). This is slower but safer in case of a crash. The default is 1. -
System Variable Name system_time_zoneVariable Scope Global Dynamic Variable No Permitted Values Type stringThe server system time zone. When the server begins executing, it inherits a time zone setting from the machine defaults, possibly modified by the environment of the account used for running the server or the startup script. The value is used to set
system_time_zone. Typically the time zone is specified by theTZenvironment variable. It also can be specified using the--timezoneoption of the mysqld_safe script.The
system_time_zonevariable differs fromtime_zone. Although they might have the same value, the latter variable is used to initialize the time zone for each client that connects. See Section 10.6, “MySQL Server Time Zone Support”. -
System Variable Name table_definition_cacheVariable Scope Global Dynamic Variable Yes Permitted Values Type numericDefault -1 (autosized)Range 400 .. 524288The number of table definitions (from
.frmfiles) that can be stored in the definition cache. If you use a large number of tables, you can create a large table definition cache to speed up opening of tables. The table definition cache takes less space and does not use file descriptors, unlike the normal table cache. The minimum value is 400. The default value is based on the following formula, capped to a limit of 2000:400 + (table_open_cache / 2)
For
InnoDB,table_definition_cacheacts as a soft limit for the number of open tables instances in theInnoDBdata dictionary cache. If the number of open table instances exceeds thetable_definition_cachesetting, the LRU mechanism begins to mark table instances for eviction and eventually removes them from the data dictionary cache. The limit helps address situations in which significant amounts of memory would be used to cache rarely used table instances until the next server restart. Table instances with foreign key relationships are not placed on the LRU list and are not subject to eviction.Additionally,
table_definition_cachedefines a soft limit for the number ofInnoDBfile-per-table tablespaces that can be open at one time, which is also controlled byinnodb_open_files. If bothtable_definition_cacheandinnodb_open_filesare set, the highest setting is used. If neither variable is set,table_definition_cache, which has a higher default value, is used. If the number of open tablespace file handles exceeds the limit defined bytable_definition_cacheorinnodb_open_files, the LRU mechanism searches the tablespace file LRU list for files that are fully flushed and are not currently being extended. This process is performed each time a new tablespace is opened. If there are no “inactive” tablespaces, no tablespace files are closed. -
System Variable Name table_open_cacheVariable Scope Global Dynamic Variable Yes Permitted Values Type numericDefault 2000Range 400 .. 524288The number of open tables for all threads. Increasing this value increases the number of file descriptors that mysqld requires. You can check whether you need to increase the table cache by checking the
Opened_tablesstatus variable. See Section 5.1.6, “Server Status Variables”. If the value ofOpened_tablesis large and you do not useFLUSH TABLESoften (which just forces all tables to be closed and reopened), then you should increase the value of thetable_open_cachevariable. For more information about the table cache, see Section 8.4.3.1, “How MySQL Opens and Closes Tables”. -
System Variable Name table_open_cache_instancesVariable Scope Global Dynamic Variable No Permitted Values Type numericDefault 1The number of open tables cache instances (default 1). To improve scalability by reducing contention among sessions, the open tables cache can be partitioned into several smaller cache instances of size
table_open_cache/table_open_cache_instances. A session need lock only one instance to access it for DML statements. This segments cache access among instances, permitting higher performance for operations that need to use the cache when many there are many sessions accessing tables. (DDL statements still require a lock on the entire cache, but such statements are much less frequent than DML statements.)A value of 8 or 16 is recommended on systems that routinely use 16 or more cores.
-
Command-Line Format --thread_cache_size=#Option-File Format thread_cache_sizeSystem Variable Name thread_cache_sizeVariable Scope Global Dynamic Variable Yes Permitted Values Type numericDefault -1 (autosized)Range 0 .. 16384How many threads the server should cache for reuse. When a client disconnects, the client's threads are put in the cache if there are fewer than
thread_cache_sizethreads there. Requests for threads are satisfied by reusing threads taken from the cache if possible, and only when the cache is empty is a new thread created. This variable can be increased to improve performance if you have a lot of new connections. Normally, this does not provide a notable performance improvement if you have a good thread implementation. However, if your server sees hundreds of connections per second you should normally setthread_cache_sizehigh enough so that most new connections use cached threads. By examining the difference between theConnectionsandThreads_createdstatus variables, you can see how efficient the thread cache is. For details, see Section 5.1.6, “Server Status Variables”.The default value is based on the following formula, capped to a limit of 100:
8 + (max_connections / 100)
This variable has no effect for the embedded server (
libmysqld) and as of MySQL 5.7.2 is no longer visible within the embedded server. -
Deprecated 5.6.1 Removed 5.7.2 Command-Line Format --thread_concurrency=#Option-File Format thread_concurrencySystem Variable Name thread_concurrencyVariable Scope Global Dynamic Variable No Permitted Values Type numericDefault 10Range 1 .. 512This variable is specific to Solaris 8 and earlier systems, for which mysqld invokes the
thr_setconcurrency()function with the variable value. This function enables applications to give the threads system a hint about the desired number of threads that should be run at the same time. Current Solaris versions document this as having no effect.This variable was removed in MySQL 5.7.2.
-
Command-Line Format --thread_handling=nameOption-File Format thread_handlingSystem Variable Name thread_handlingVariable Scope Global Dynamic Variable No Permitted Values Type enumerationValid Values no-threadsone-thread-per-connectiondynamically-loadedThe thread-handling model used by the server for connection threads. The permissible values are
no-threads(the server uses a single thread) andone-thread-per-connection(the server uses one thread to handle each client connection).no-threadsis useful for debugging under Linux; see Section 22.4, “Debugging and Porting MySQL”.This variable has no effect for the embedded server (
libmysqld) and as of MySQL 5.7.2 is no longer visible within the embedded server. -
Command-Line Format --thread_stack=#Option-File Format thread_stackSystem Variable Name thread_stackVariable Scope Global Dynamic Variable No Permitted Values Platform Bit Size 32Type numericDefault 196608Range 131072 .. 4294967295Block Size 1024Permitted Values Platform Bit Size 64Type numericDefault 262144Range 131072 .. 18446744073709547520Block Size 1024The stack size for each thread. Many of the limits detected by the
crash-metest are dependent on this value. See Section 8.12.2, “The MySQL Benchmark Suite”. The default of 192KB (256KB for 64-bit systems) is large enough for normal operation. If the thread stack size is too small, it limits the complexity of the SQL statements that the server can handle, the recursion depth of stored procedures, and other memory-consuming actions. This variable is unused. It is deprecated and will be removed in a future MySQL release.
-
System Variable Name time_zoneVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type stringThe current time zone. This variable is used to initialize the time zone for each client that connects. By default, the initial value of this is
'SYSTEM'(which means, “use the value ofsystem_time_zone”). The value can be specified explicitly at server startup with the--default-time-zoneoption. See Section 10.6, “MySQL Server Time Zone Support”. -
Command-Line Format --timed_mutexesOption-File Format timed_mutexesSystem Variable Name timed_mutexesVariable Scope Global Dynamic Variable Yes Permitted Values Type booleanDefault OFFThis variable controls whether
InnoDBmutexes are timed. If this variable is set to 0 orOFF(the default), mutex timing is disabled. If the variable is set to 1 orON, mutex timing is enabled. With timing enabled, theos_wait_timesvalue in the output fromSHOW ENGINE INNODB MUTEXindicates the amount of time (in ms) spent in operating system waits. Otherwise, the value is 0. timestamp = {timestamp_value| DEFAULT}Set the time for this client. This is used to get the original timestamp if you use the binary log to restore rows.
timestamp_valueshould be a Unix epoch timestamp, not a MySQL timestamp.In MySQL 5.7,
timestampis aDOUBLErather thanBIGINTbecause its value includes a microseconds part.SET timestampaffects the value returned byNOW()but not bySYSDATE(). This means that timestamp settings in the binary log have no effect on invocations ofSYSDATE(). The server can be started with the--sysdate-is-nowoption to causeSYSDATE()to be an alias forNOW(), in which caseSET timestampaffects both functions.-
Command-Line Format --tmp_table_size=#Option-File Format tmp_table_sizeSystem Variable Name tmp_table_sizeVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault system dependentRange 1024 .. 4294967295The maximum size of internal in-memory temporary tables. (The actual limit is determined as the minimum of
tmp_table_sizeandmax_heap_table_size.) If an in-memory temporary table exceeds the limit, MySQL automatically converts it to an on-diskMyISAMtable. Increase the value oftmp_table_size(andmax_heap_table_sizeif necessary) if you do many advancedGROUP BYqueries and you have lots of memory. This variable does not apply to user-createdMEMORYtables.You can compare the number of internal on-disk temporary tables created to the total number of internal temporary tables created by comparing the values of the
Created_tmp_disk_tablesandCreated_tmp_tablesvariables.See also Section 8.4.3.3, “How MySQL Uses Internal Temporary Tables”.
-
Command-Line Format --tmpdir=path-tOption-File Format tmpdirSystem Variable Name tmpdirVariable Scope Global Dynamic Variable No Permitted Values Type file nameThe directory used for temporary files and temporary tables. This variable can be set to a list of several paths that are used in round-robin fashion. Paths should be separated by colon characters (“
:”) on Unix and semicolon characters (“;”) on Windows.The multiple-directory feature can be used to spread the load between several physical disks. If the MySQL server is acting as a replication slave, you should not set
tmpdirto point to a directory on a memory-based file system or to a directory that is cleared when the server host restarts. A replication slave needs some of its temporary files to survive a machine restart so that it can replicate temporary tables orLOAD DATA INFILEoperations. If files in the temporary file directory are lost when the server restarts, replication fails. You can set the slave's temporary directory using theslave_load_tmpdirvariable. In that case, the slave will not use the generaltmpdirvalue and you can settmpdirto a nonpermanent location. -
Command-Line Format --transaction_alloc_block_size=#Option-File Format transaction_alloc_block_sizeSystem Variable Name transaction_alloc_block_sizeVariable Scope Global, Session Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 8192Range 1024 .. 4294967295Block Size 1024Permitted Values Platform Bit Size 64Type numericDefault 8192Range 1024 .. 18446744073709547520Block Size 1024The amount in bytes by which to increase a per-transaction memory pool which needs memory. See the description of
transaction_prealloc_size. -
Command-Line Format --transaction_prealloc_size=#Option-File Format transaction_prealloc_sizeSystem Variable Name transaction_prealloc_sizeVariable Scope Global, Session Dynamic Variable Yes Permitted Values Platform Bit Size 32Type numericDefault 4096Range 1024 .. 4294967295Block Size 1024Permitted Values Platform Bit Size 64Type numericDefault 4096Range 1024 .. 18446744073709547520Block Size 1024There is a per-transaction memory pool from which various transaction-related allocations take memory. The initial size of the pool in bytes is
transaction_prealloc_size. For every allocation that cannot be satisfied from the pool because it has insufficient memory available, the pool is increased bytransaction_alloc_block_sizebytes. When the transaction ends, the pool is truncated totransaction_prealloc_sizebytes.By making
transaction_prealloc_sizesufficiently large to contain all statements within a single transaction, you can avoid manymalloc()calls. -
System Variable Name tx_isolationVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type enumerationDefault REPEATABLE-READValid Values READ-UNCOMMITTEDREAD-COMMITTEDREPEATABLE-READSERIALIZABLEThe default transaction isolation level. Defaults to
REPEATABLE-READ.This variable can be set directly, or indirectly using the
SET TRANSACTIONstatement. See Section 13.3.6, “SET TRANSACTIONSyntax”. If you settx_isolationdirectly to an isolation level name that contains a space, the name should be enclosed within quotation marks, with the space replaced by a dash. For example:SET tx_isolation = 'READ-COMMITTED';
Any unique prefix of a valid value may be used to set the value of this variable.
The default transaction isolation level can also be set at startup using the
--transaction-isolationserver option. -
System Variable Name tx_read_onlyVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type booleanDefault OFFThe default transaction access mode. The value can be
OFF(read/write, the default) orON(read only).This variable can be set directly, or indirectly using the
SET TRANSACTIONstatement. See Section 13.3.6, “SET TRANSACTIONSyntax”.To set the default transaction access mode at startup, use the
--transaction-read-onlyserver option. -
System Variable Name unique_checksVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type booleanDefault 1If set to 1 (the default), uniqueness checks for secondary indexes in
InnoDBtables are performed. If set to 0, storage engines are permitted to assume that duplicate keys are not present in input data. If you know for certain that your data does not contain uniqueness violations, you can set this to 0 to speed up large table imports toInnoDB.Note that setting this variable to 0 does not require storage engines to ignore duplicate keys. An engine is still permitted to check for them and issue duplicate-key errors if it detects them.
-
Command-Line Format --updatable_views_with_limit=#Option-File Format updatable_views_with_limitSystem Variable Name updatable_views_with_limitVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type booleanDefault 1This variable controls whether updates to a view can be made when the view does not contain all columns of the primary key defined in the underlying table, if the update statement contains a
LIMITclause. (Such updates often are generated by GUI tools.) An update is anUPDATEorDELETEstatement. Primary key here means aPRIMARY KEY, or aUNIQUEindex in which no column can containNULL.The variable can have two values:
1orYES: Issue a warning only (not an error message). This is the default value.0orNO: Prohibit the update.
validate_password_xxxThe
validate_passwordplugin implements a set of system variables having names of the formvalidate_password_. These variables affect password testing by that plugin; see Section 6.1.2.6.2, “Password Validation Plugin Options and Variables”.xxx-
System Variable Name validate_user_pluginsVariable Scope Global Dynamic Variable No Permitted Values Type booleanDefault ONIf this variable is enabled (the default), the server checks each user account and produces a warning if conditions are found that would make the account unusable:
The account requires an authentication plugin that is not loaded.
The account requires the
sha256_passwordauthentication plugin but the server was started with neither SSL nor RSA enabled as required by this plugin.
Enabling
validate_user_pluginsslows down server initialization andFLUSH PRIVILEGES. If you do not require the additional checking, you can disable this variable at startup to avoid the performance decrement.This variable was added in MySQL 5.7.1.
The version number for the server. The value might also include a suffix indicating server build or configuration information.
-logindicates that one or more of the general log, slow query log, or binary log are enabled.-debugindicates that the server was built with debugging support enabled.-
System Variable Name version_commentVariable Scope Global Dynamic Variable No Permitted Values Type stringThe CMake configuration program has a
COMPILATION_COMMENToption that permits a comment to be specified when building MySQL. This variable contains the value of that comment. See Section 2.9.4, “MySQL Source-Configuration Options”. -
System Variable Name version_compile_machineVariable Scope Global Dynamic Variable No Permitted Values Type stringThe type of the server binary.
-
System Variable Name version_compile_osVariable Scope Global Dynamic Variable No Permitted Values Type stringThe type of operating system on which MySQL was built.
-
Command-Line Format --wait_timeout=#Option-File Format wait_timeoutSystem Variable Name wait_timeoutVariable Scope Global, Session Dynamic Variable Yes Permitted Values Type numericDefault 28800Range 1 .. 31536000Permitted Values Type (windows) numericDefault 28800Range 1 .. 2147483The number of seconds the server waits for activity on a noninteractive connection before closing it.
On thread startup, the session
wait_timeoutvalue is initialized from the globalwait_timeoutvalue or from the globalinteractive_timeoutvalue, depending on the type of client (as defined by theCLIENT_INTERACTIVEconnect option tomysql_real_connect()). See alsointeractive_timeout. The number of errors, warnings, and notes that resulted from the last statement that generated messages. This variable is read only. See Section 13.7.5.39, “
SHOW WARNINGSSyntax”.
The MySQL server maintains many system variables that indicate how
it is configured. Section 5.1.4, “Server System Variables”,
describes the meaning of these variables. Each system variable has
a default value. System variables can be set at server startup
using options on the command line or in an option file. Most of
them can be changed dynamically while the server is running by
means of the
SET
statement, which enables you to modify operation of the server
without having to stop and restart it. You can refer to system
variable values in expressions.
The server maintains two kinds of system variables. Global variables affect the overall operation of the server. Session variables affect its operation for individual client connections. A given system variable can have both a global and a session value. Global and session system variables are related as follows:
When the server starts, it initializes all global variables to their default values. These defaults can be changed by options specified on the command line or in an option file. (See Section 4.2.3, “Specifying Program Options”.)
The server also maintains a set of session variables for each client that connects. The client's session variables are initialized at connect time using the current values of the corresponding global variables. For example, the client's SQL mode is controlled by the session
sql_modevalue, which is initialized when the client connects to the value of the globalsql_modevalue.
System variable values can be set globally at server startup by
using options on the command line or in an option file. When you
use a startup option to set a variable that takes a numeric value,
the value can be given with a suffix of K,
M, or G (either uppercase or
lowercase) to indicate a multiplier of 1024,
10242 or
10243; that is, units of kilobytes,
megabytes, or gigabytes, respectively. Thus, the following command
starts the server with a query cache size of 16 megabytes and a
maximum packet size of one gigabyte:
mysqld --query_cache_size=16M --max_allowed_packet=1G
Within an option file, those variables are set like this:
[mysqld] query_cache_size=16M max_allowed_packet=1G
The lettercase of suffix letters does not matter;
16M and 16m are equivalent,
as are 1G and 1g.
If you want to restrict the maximum value to which a system
variable can be set at runtime with the
SET
statement, you can specify this maximum by using an option of the
form
--maximum-
at server startup. For example, to prevent the value of
var_name=valuequery_cache_size from being
increased to more than 32MB at runtime, use the option
--maximum-query_cache_size=32M.
Many system variables are dynamic and can be changed while the
server runs by using the
SET
statement. For a list, see
Section 5.1.5.2, “Dynamic System Variables”. To change a system
variable with
SET,
refer to it as var_name, optionally
preceded by a modifier:
To indicate explicitly that a variable is a global variable, precede its name by
GLOBALor@@global.. TheSUPERprivilege is required to set global variables.To indicate explicitly that a variable is a session variable, precede its name by
SESSION,@@session., or@@. Setting a session variable requires no special privilege, but a client can change only its own session variables, not those of any other client.LOCALand@@local.are synonyms forSESSIONand@@session..If no modifier is present,
SETchanges the session variable.
A SET
statement can contain multiple variable assignments, separated by
commas. If you set several system variables, the most recent
GLOBAL or SESSION modifier
in the statement is used for following variables that have no
modifier specified.
Examples:
SET sort_buffer_size=10000; SET @@local.sort_buffer_size=10000; SET GLOBAL sort_buffer_size=1000000, SESSION sort_buffer_size=1000000; SET @@sort_buffer_size=1000000; SET @@global.sort_buffer_size=1000000, @@local.sort_buffer_size=1000000;
The @@
syntax for system variables is supported for compatibility with
some other database systems.
var_name
If you change a session system variable, the value remains in effect until your session ends or until you change the variable to a different value. The change is not visible to other clients.
If you change a global system variable, the value is remembered
and used for new connections until the server restarts. (To make a
global system variable setting permanent, you should set it in an
option file.) The change is visible to any client that accesses
that global variable. However, the change affects the
corresponding session variable only for clients that connect after
the change. The global variable change does not affect the session
variable for any client that is currently connected (not even that
of the client that issues the
SET
GLOBAL statement).
To prevent incorrect usage, MySQL produces an error if you use
SET
GLOBAL with a variable that can only be used with
SET
SESSION or if you do not specify
GLOBAL (or @@global.) when
setting a global variable.
To set a SESSION variable to the
GLOBAL value or a GLOBAL
value to the compiled-in MySQL default value, use the
DEFAULT keyword. For example, the following two
statements are identical in setting the session value of
max_join_size to the global
value:
SET max_join_size=DEFAULT; SET @@session.max_join_size=@@global.max_join_size;
Not all system variables can be set to DEFAULT.
In such cases, use of DEFAULT results in an
error.
You can refer to the values of specific global or session system
variables in expressions by using one of the
@@-modifiers. For example, you can retrieve
values in a SELECT statement like
this:
SELECT @@global.sql_mode, @@session.sql_mode, @@sql_mode;
When you refer to a system variable in an expression as
@@ (that is,
when you do not specify var_name@@global. or
@@session.), MySQL returns the session value if
it exists and the global value otherwise. (This differs from
SET @@, which always refers to
the session value.)
var_name =
value
Some variables displayed by SHOW VARIABLES
may not be available using SELECT
@@ syntax; an
var_nameUnknown system variable occurs. As a
workaround in such cases, you can use SHOW VARIABLES
LIKE '.
var_name'
Suffixes for specifying a value multiplier can be used when
setting a variable at server startup, but not to set the value
with SET
at runtime. On the other hand, with
SET you
can assign a variable's value using an expression, which is not
true when you set a variable at server startup. For example, the
first of the following lines is legal at server startup, but the
second is not:
shell>mysql --max_allowed_packet=16Mshell>mysql --max_allowed_packet=16*1024*1024
Conversely, the second of the following lines is legal at runtime, but the first is not:
mysql>SET GLOBAL max_allowed_packet=16M;mysql>SET GLOBAL max_allowed_packet=16*1024*1024;
Some system variables can be enabled with the
SET
statement by setting them to ON or
1, or disabled by setting them to
OFF or 0. However, to set
such a variable on the command line or in an option file, you
must set it to 1 or 0;
setting it to ON or OFF
will not work. For example, on the command line,
--delay_key_write=1 works but
--delay_key_write=ON does not.
To display system variable names and values, use the
SHOW VARIABLES statement:
mysql> SHOW VARIABLES;
+---------------------------------+-----------------------------------+
| Variable_name | Value |
+---------------------------------+-----------------------------------+
| auto_increment_increment | 1 |
| auto_increment_offset | 1 |
| automatic_sp_privileges | ON |
| back_log | 50 |
| basedir | /home/mysql/ |
| binlog_cache_size | 32768 |
| bulk_insert_buffer_size | 8388608 |
| character_set_client | latin1 |
| character_set_connection | latin1 |
| character_set_database | latin1 |
| character_set_results | latin1 |
| character_set_server | latin1 |
| character_set_system | utf8 |
| character_sets_dir | /home/mysql/share/mysql/charsets/ |
| collation_connection | latin1_swedish_ci |
| collation_database | latin1_swedish_ci |
| collation_server | latin1_swedish_ci |
...
| innodb_additional_mem_pool_size | 1048576 |
| innodb_autoextend_increment | 8 |
| innodb_buffer_pool_size | 8388608 |
| innodb_checksums | ON |
| innodb_commit_concurrency | 0 |
| innodb_concurrency_tickets | 500 |
| innodb_data_file_path | ibdata1:10M:autoextend |
| innodb_data_home_dir | |
...
| version | 5.1.6-alpha-log |
| version_comment | Source distribution |
| version_compile_machine | i686 |
| version_compile_os | suse-linux |
| wait_timeout | 28800 |
+---------------------------------+-----------------------------------+
With a LIKE clause, the statement
displays only those variables that match the pattern. To obtain a
specific variable name, use a LIKE
clause as shown:
SHOW VARIABLES LIKE 'max_join_size'; SHOW SESSION VARIABLES LIKE 'max_join_size';
To get a list of variables whose name match a pattern, use the
“%” wildcard character in a
LIKE clause:
SHOW VARIABLES LIKE '%size%'; SHOW GLOBAL VARIABLES LIKE '%size%';
Wildcard characters can be used in any position within the pattern
to be matched. Strictly speaking, because
“_” is a wildcard that matches any
single character, you should escape it as
“\_” to match it literally. In
practice, this is rarely necessary.
For SHOW VARIABLES, if you specify
neither GLOBAL nor SESSION,
MySQL returns SESSION values.
The reason for requiring the GLOBAL keyword
when setting GLOBAL-only variables but not when
retrieving them is to prevent problems in the future. If we were
to remove a SESSION variable that has the same
name as a GLOBAL variable, a client with the
SUPER privilege might accidentally
change the GLOBAL variable rather than just the
SESSION variable for its own connection. If we
add a SESSION variable with the same name as a
GLOBAL variable, a client that intends to
change the GLOBAL variable might find only its
own SESSION variable changed.
A structured variable differs from a regular system variable in two respects:
Its value is a structure with components that specify server parameters considered to be closely related.
There might be several instances of a given type of structured variable. Each one has a different name and refers to a different resource maintained by the server.
MySQL 5.7 supports one structured variable type, which specifies parameters governing the operation of key caches. A key cache structured variable has these components:
This section describes the syntax for referring to structured
variables. Key cache variables are used for syntax examples, but
specific details about how key caches operate are found
elsewhere, in Section 8.9.2, “The MyISAM Key Cache”.
To refer to a component of a structured variable instance, you
can use a compound name in
instance_name.component_name format.
Examples:
hot_cache.key_buffer_size hot_cache.key_cache_block_size cold_cache.key_cache_block_size
For each structured system variable, an instance with the name
of default is always predefined. If you refer
to a component of a structured variable without any instance
name, the default instance is used. Thus,
default.key_buffer_size and
key_buffer_size both refer to
the same system variable.
Structured variable instances and components follow these naming rules:
For a given type of structured variable, each instance must have a name that is unique within variables of that type. However, instance names need not be unique across structured variable types. For example, each structured variable has an instance named
default, sodefaultis not unique across variable types.The names of the components of each structured variable type must be unique across all system variable names. If this were not true (that is, if two different types of structured variables could share component member names), it would not be clear which default structured variable to use for references to member names that are not qualified by an instance name.
If a structured variable instance name is not legal as an unquoted identifier, refer to it as a quoted identifier using backticks. For example,
hot-cacheis not legal, but`hot-cache`is.global,session, andlocalare not legal instance names. This avoids a conflict with notation such as@@global.for referring to nonstructured system variables.var_name
Currently, the first two rules have no possibility of being violated because the only structured variable type is the one for key caches. These rules will assume greater significance if some other type of structured variable is created in the future.
With one exception, you can refer to structured variable components using compound names in any context where simple variable names can occur. For example, you can assign a value to a structured variable using a command-line option:
shell> mysqld --hot_cache.key_buffer_size=64K
In an option file, use this syntax:
[mysqld] hot_cache.key_buffer_size=64K
If you start the server with this option, it creates a key cache
named hot_cache with a size of 64KB in
addition to the default key cache that has a default size of
8MB.
Suppose that you start the server as follows:
shell>mysqld --key_buffer_size=256K \--extra_cache.key_buffer_size=128K \--extra_cache.key_cache_block_size=2048
In this case, the server sets the size of the default key cache
to 256KB. (You could also have written
--default.key_buffer_size=256K.) In addition,
the server creates a second key cache named
extra_cache that has a size of 128KB, with
the size of block buffers for caching table index blocks set to
2048 bytes.
The following example starts the server with three different key caches having sizes in a 3:1:1 ratio:
shell>mysqld --key_buffer_size=6M \--hot_cache.key_buffer_size=2M \--cold_cache.key_buffer_size=2M
Structured variable values may be set and retrieved at runtime
as well. For example, to set a key cache named
hot_cache to a size of 10MB, use either of
these statements:
mysql>SET GLOBAL hot_cache.key_buffer_size = 10*1024*1024;mysql>SET @@global.hot_cache.key_buffer_size = 10*1024*1024;
To retrieve the cache size, do this:
mysql> SELECT @@global.hot_cache.key_buffer_size;
However, the following statement does not work. The variable is
not interpreted as a compound name, but as a simple string for a
LIKE pattern-matching operation:
mysql> SHOW GLOBAL VARIABLES LIKE 'hot_cache.key_buffer_size';
This is the exception to being able to use structured variable names anywhere a simple variable name may occur.
Many server system variables are dynamic and can be set at
runtime using SET
GLOBAL or
SET
SESSION. You can also obtain their values using
SELECT. See
Section 5.1.5, “Using System Variables”.
The following table shows the full list of all dynamic system
variables. The last column indicates for each variable whether
GLOBAL or SESSION (or
both) apply. The table also lists session options that can be
set with the
SET
statement. Section 5.1.4, “Server System Variables”, discusses
these options.
Variables that have a type of “string” take a
string value. Variables that have a type of
“numeric” take a numeric value. Variables that have
a type of “boolean” can be set to 0, 1,
ON or OFF. (If you set
them on the command line or in an option file, use the numeric
values.) Variables that are marked as “enumeration”
normally should be set to one of the available values for the
variable, but can also be set to the number that corresponds to
the desired enumeration value. For enumerated system variables,
the first enumeration value corresponds to 0. This differs from
ENUM columns, for which the first
enumeration value corresponds to 1.
Table 5.3. Dynamic Variable Summary
The server maintains many status variables that provide
information about its operation. You can view these variables and
their values by using the SHOW [GLOBAL | SESSION]
STATUS statement (see Section 13.7.5.34, “SHOW STATUS Syntax”).
The optional GLOBAL keyword aggregates the
values over all connections, and SESSION shows
the values for the current connection.
mysql> SHOW GLOBAL STATUS;
+-----------------------------------+------------+
| Variable_name | Value |
+-----------------------------------+------------+
| Aborted_clients | 0 |
| Aborted_connects | 0 |
| Bytes_received | 155372598 |
| Bytes_sent | 1176560426 |
...
| Connections | 30023 |
| Created_tmp_disk_tables | 0 |
| Created_tmp_files | 3 |
| Created_tmp_tables | 2 |
...
| Threads_created | 217 |
| Threads_running | 88 |
| Uptime | 1389872 |
+-----------------------------------+------------+
Many status variables are reset to 0 by the
FLUSH STATUS
statement.
The following table lists all available server status variables:
Table 5.4. Status Variable Summary
The status variables have the following meanings.
The number of connections that were aborted because the client died without closing the connection properly. See Section C.5.2.11, “Communication Errors and Aborted Connections”.
The number of failed attempts to connect to the MySQL server. See Section C.5.2.11, “Communication Errors and Aborted Connections”.
For additional connection-related information, check the
Connection_errors_status variables and thexxxhost_cachetable.The number of transactions that used the temporary binary log cache but that exceeded the value of
binlog_cache_sizeand used a temporary file to store statements from the transaction.The number of nontransactional statements that caused the binary log transaction cache to be written to disk is tracked separately in the
Binlog_stmt_cache_disk_usestatus variable.The number of transactions that used the binary log cache.
The number of nontransaction statements that used the binary log statement cache but that exceeded the value of
binlog_stmt_cache_sizeand used a temporary file to store those statements.The number of nontransactional statements that used the binary log statement cache.
The number of bytes received from all clients.
The number of bytes sent to all clients.
The
Com_statement counter variables indicate the number of times eachxxxxxxstatement has been executed. There is one status variable for each type of statement. For example,Com_deleteandCom_updatecountDELETEandUPDATEstatements, respectively.Com_delete_multiandCom_update_multiare similar but apply toDELETEandUPDATEstatements that use multiple-table syntax.If a query result is returned from query cache, the server increments the
Qcache_hitsstatus variable, notCom_select. See Section 8.9.3.4, “Query Cache Status and Maintenance”.All of the
Com_stmt_variables are increased even if a prepared statement argument is unknown or an error occurred during execution. In other words, their values correspond to the number of requests issued, not to the number of requests successfully completed.xxxThe
Com_stmt_status variables are as follows:xxxCom_stmt_prepareCom_stmt_executeCom_stmt_fetchCom_stmt_send_long_dataCom_stmt_resetCom_stmt_close
Those variables stand for prepared statement commands. Their names refer to the
COM_command set used in the network layer. In other words, their values increase whenever prepared statement API calls such as mysql_stmt_prepare(), mysql_stmt_execute(), and so forth are executed. However,xxxCom_stmt_prepare,Com_stmt_executeandCom_stmt_closealso increase forPREPARE,EXECUTE, orDEALLOCATE PREPARE, respectively. Additionally, the values of the older statement counter variablesCom_prepare_sql,Com_execute_sql, andCom_dealloc_sqlincrease for thePREPARE,EXECUTE, andDEALLOCATE PREPAREstatements.Com_stmt_fetchstands for the total number of network round-trips issued when fetching from cursors.Com_stmt_reprepareindicates the number of times statements were automatically reprepared by the server after metadata changes to tables or views referred to by the statement. A reprepare operation incrementsCom_stmt_reprepare, and alsoCom_stmt_prepare.Whether the client connection uses compression in the client/server protocol.
These variables provide information about errors that occur during the client connection process. They are global only and represent error counts aggregated across connections from all hosts. These variables track errors not accounted for by the host cache (see Section 8.11.5.2, “DNS Lookup Optimization and the Host Cache”), such as errors that are not associated with TCP connections, occur very early in the connection process (even before an IP address is known), or are not specific to any particular IP address (such as out-of-memory conditions).
The number of errors that occurred during calls to
accept()on the listening port.The number of connections refused due to internal errors in the server, such as failure to start a new thread or an out-of-memory condition.
Connection_errors_max_connectionsThe number of connections refused because the server
max_connectionslimit was reached.The number of errors that occurred while searching for connecting client IP addresses.
The number of errors that occurred during calls to
select()orpoll()on the listening port. (Failure of this operation does not necessarily means a client connection was rejected.)The number of connections refused by the
libwraplibrary.
The number of connection attempts (successful or not) to the MySQL server.
The number of internal on-disk temporary tables created by the server while executing statements.
If an internal temporary table is created initially as an in-memory table but becomes too large, MySQL automatically converts it to an on-disk table. The maximum size for in-memory temporary tables is the minimum of the
tmp_table_sizeandmax_heap_table_sizevalues. IfCreated_tmp_disk_tablesis large, you may want to increase thetmp_table_sizeormax_heap_table_sizevalue to lessen the likelihood that internal temporary tables in memory will be converted to on-disk tables.You can compare the number of internal on-disk temporary tables created to the total number of internal temporary tables created by comparing the values of the
Created_tmp_disk_tablesandCreated_tmp_tablesvariables.See also Section 8.4.3.3, “How MySQL Uses Internal Temporary Tables”.
How many temporary files mysqld has created.
The number of internal temporary tables created by the server while executing statements.
You can compare the number of internal on-disk temporary tables created to the total number of internal temporary tables created by comparing the values of the
Created_tmp_disk_tablesandCreated_tmp_tablesvariables.See also Section 8.4.3.3, “How MySQL Uses Internal Temporary Tables”.
Each invocation of the
SHOW STATUSstatement uses an internal temporary table and increments the globalCreated_tmp_tablesvalue.In MySQL 5.7, this status variable is deprecated (because
DELAYEDinserts are not supported), and will be removed in a future release.In MySQL 5.7, this status variable is deprecated (because
DELAYEDinserts are not supported), and will be removed in a future release.In MySQL 5.7, this status variable is deprecated (because
DELAYEDinserts are not supported), and will be removed in a future release.The number of times the server flushes tables, whether because a user executed a
FLUSH TABLESstatement or due to internal server operation. It is also incremented by receipt of aCOM_REFRESHpacket. This is in contrast toCom_flush, which indicates how manyFLUSHstatements have been executed, whetherFLUSH TABLES,FLUSH LOGS, and so forth.The number of internal
COMMITstatements.The number of times that rows have been deleted from tables.
The server increments this variable for each call to its
external_lock()function, which generally occurs at the beginning and end of access to a table instance. There might be differences among storage engines. This variable can be used, for example, to discover for a statement that accesses a partitioned table how many partitions were pruned before locking occurred: Check how much the counter increased for the statement, subtract 2 (2 calls for the table itself), then divide by 2 to get the number of partitions locked.The number of times the server uses a storage engine's own Multi-Range Read implementation for table access.
A counter for the prepare phase of two-phase commit operations.
The number of times the first entry in an index was read. If this value is high, it suggests that the server is doing a lot of full index scans; for example,
SELECT col1 FROM foo, assuming thatcol1is indexed.The number of requests to read a row based on a key. If this value is high, it is a good indication that your tables are properly indexed for your queries.
The number of requests to read the last key in an index. With
ORDER BY, the server will issue a first-key request followed by several next-key requests, whereas with WithORDER BY DESC, the server will issue a last-key request followed by several previous-key requests.The number of requests to read the next row in key order. This value is incremented if you are querying an index column with a range constraint or if you are doing an index scan.
The number of requests to read the previous row in key order. This read method is mainly used to optimize
ORDER BY ... DESC.The number of requests to read a row based on a fixed position. This value is high if you are doing a lot of queries that require sorting of the result. You probably have a lot of queries that require MySQL to scan entire tables or you have joins that do not use keys properly.
The number of requests to read the next row in the data file. This value is high if you are doing a lot of table scans. Generally this suggests that your tables are not properly indexed or that your queries are not written to take advantage of the indexes you have.
The number of requests for a storage engine to perform a rollback operation.
The number of requests for a storage engine to place a savepoint.
The number of requests for a storage engine to roll back to a savepoint.
The number of requests to update a row in a table.
The number of requests to insert a row in a table.
The total number of available
InnoDBundo logs. Supplements theinnodb_undo_logssystem variable, which reports the number of active undo logs.Innodb_buffer_pool_dump_statusThe progress of an operation to record the pages held in the
InnoDBbuffer pool, triggered by the setting ofinnodb_buffer_pool_dump_at_shutdownorinnodb_buffer_pool_dump_now.Innodb_buffer_pool_load_statusThe progress of an operation to warm up the
InnoDBbuffer pool by reading in a set of pages corresponding to an earlier point in time, triggered by the setting ofinnodb_buffer_pool_load_at_startuporinnodb_buffer_pool_load_now. If the operation introduces too much overhead, you can cancel it by settinginnodb_buffer_pool_load_abort.The total number of bytes in the
InnoDBbuffer pool containing data. The number includes both dirty and clean pages. For more accurate memory usage calculations than withInnodb_buffer_pool_pages_data, when compressed tables cause the buffer pool to hold pages of different sizes.The number of pages in the
InnoDBbuffer pool containing data. The number includes both dirty and clean pages.Innodb_buffer_pool_bytes_dirtyThe total current number of bytes held in dirty pages in the
InnoDBbuffer pool. For more accurate memory usage calculations than withInnodb_buffer_pool_pages_dirty, when compressed tables cause the buffer pool to hold pages of different sizes.Innodb_buffer_pool_pages_dirtyThe current number of dirty pages in the
InnoDBbuffer pool.Innodb_buffer_pool_pages_flushedThe number of requests to flush pages from the
InnoDBbuffer pool.The number of free pages in the
InnoDBbuffer pool.Innodb_buffer_pool_pages_latchedThe number of latched pages in the
InnoDBbuffer pool. These are pages currently being read or written, or that cannot be flushed or removed for some other reason. Calculation of this variable is expensive, so it is available only when theUNIV_DEBUGsystem is defined at server build time.The number of pages in the
InnoDBbuffer pool that are busy because they have been allocated for administrative overhead, such as row locks or the adaptive hash index. This value can also be calculated asInnodb_buffer_pool_pages_total–Innodb_buffer_pool_pages_free–Innodb_buffer_pool_pages_data.Innodb_buffer_pool_pages_totalThe total size of the
InnoDBbuffer pool, in pages.The number of pages read into the
InnoDBbuffer pool by the read-ahead background thread.Innodb_buffer_pool_read_ahead_evictedThe number of pages read into the
InnoDBbuffer pool by the read-ahead background thread that were subsequently evicted without having been accessed by queries.Innodb_buffer_pool_read_requestsThe number of logical read requests made to the
InnoDBbuffer pool. These requests could be serviced by returning data that was already in memory, or by reading the data from disk into memory first.The number of logical reads that
InnoDBcould not satisfy from the buffer pool, and had to read directly from disk.Normally, writes to the
InnoDBbuffer pool happen in the background. WhenInnoDBneeds to read or create a page and no clean pages are available,InnoDBflushes some dirty pages first and waits for that operation to finish. This counter counts instances of these waits. Ifinnodb_buffer_pool_sizehas been set properly, this value should be small.Innodb_buffer_pool_write_requestsThe number of writes done to the
InnoDBbuffer pool.The number of
fsync()operations so far. The frequency offsync()calls is influenced by the setting of theinnodb_flush_methodconfiguration option.The current number of pending
fsync()operations. The frequency offsync()calls is influenced by the setting of theinnodb_flush_methodconfiguration option.The current number of pending reads.
The current number of pending writes.
The amount of data read since the server was started.
The total number of data reads.
The total number of data writes.
The amount of data written so far, in bytes.
The number of pages that have been written to the doublewrite buffer. See Section 5.3.1, “
InnoDBDisk I/O”.The number of doublewrite operations that have been performed. See Section 5.3.1, “
InnoDBDisk I/O”.Indicates whether the server was built with atomic instructions.
The number of times that the log buffer was too small and a wait was required for it to be flushed before continuing.
The number of write requests for the
InnoDBredo log.The number of physical writes to the
InnoDBredo log file.The number of files
InnoDBcurrently holds open.The number of
fsync()writes done to theInnoDBredo log files.The number of pending
fsync()operations for theInnoDBredo log files.The number of pending writes to the
InnoDBredo log files.The number of bytes written to the
InnoDBredo log files.The compiled-in
InnoDBpage size (default 16KB). Many values are counted in pages; the page size enables them to be easily converted to bytes.The number of pages created by operations on
InnoDBtables.The number of pages read by operations on
InnoDBtables.The number of pages written by operations on
InnoDBtables.The number of row locks currently being waited for by operations on
InnoDBtables.The total time spent in acquiring row locks for
InnoDBtables, in milliseconds.The average time to acquire a row lock for
InnoDBtables, in milliseconds.The maximum time to acquire a row lock for
InnoDBtables, in milliseconds.The number of times operations on
InnoDBtables had to wait for a row lock.The number of rows deleted from
InnoDBtables.The number of rows inserted into
InnoDBtables.The number of rows read from
InnoDBtables.The number of rows updated in
InnoDBtables.Innodb_truncated_status_writesThe number of times output from the
SHOW ENGINE INNODB STATUSstatement has been truncated.The number of key blocks in the
MyISAMkey cache that have changed but have not yet been flushed to disk.The number of unused blocks in the
MyISAMkey cache. You can use this value to determine how much of the key cache is in use; see the discussion ofkey_buffer_sizein Section 5.1.4, “Server System Variables”.The number of used blocks in the
MyISAMkey cache. This value is a high-water mark that indicates the maximum number of blocks that have ever been in use at one time.The number of requests to read a key block from the
MyISAMkey cache.The number of physical reads of a key block from disk into the
MyISAMkey cache. IfKey_readsis large, then yourkey_buffer_sizevalue is probably too small. The cache miss rate can be calculated asKey_reads/Key_read_requests.The number of requests to write a key block to the
MyISAMkey cache.The number of physical writes of a key block from the
MyISAMkey cache to disk.The total cost of the last compiled query as computed by the query optimizer. This is useful for comparing the cost of different query plans for the same query. The default value of 0 means that no query has been compiled yet. The default value is 0.
Last_query_costhas session scope.The
Last_query_costvalue can be computed accurately only for simple “flat” queries, not complex queries such as those with subqueries orUNION. For the latter, the value is set to 0.The number of iterations the query optimizer made in execution plan construction for the previous query.
Last_query_costhas session scope.The maximum number of connections that have been in use simultaneously since the server started.
In MySQL 5.7, this status variable is deprecated (because
DELAYEDinserts are not supported), and will be removed in a future release.The number of files that are open. This count includes regular files opened by the server. It does not include other types of files such as sockets or pipes. Also, the count does not include files that storage engines open using their own internal functions rather than asking the server level to do so.
The number of streams that are open (used mainly for logging).
The number of cached
.frmfiles.The number of tables that are open.
The number of files that have been opened with
my_open()(amysyslibrary function). Parts of the server that open files without using this function do not increment the count.The number of
.frmfiles that have been cached.The number of tables that have been opened. If
Opened_tablesis big, yourtable_open_cachevalue is probably too small.Performance_schema_xxxPerformance Schema status variables are listed in Section 20.13, “Performance Schema Status Variables”. These variables provide information about instrumentation that could not be loaded or created due to memory constraints.
The current number of prepared statements. (The maximum number of statements is given by the
max_prepared_stmt_countsystem variable.)The number of free memory blocks in the query cache.
The amount of free memory for the query cache.
The number of query cache hits.
The number of queries added to the query cache.
The number of queries that were deleted from the query cache because of low memory.
The number of noncached queries (not cacheable, or not cached due to the
query_cache_typesetting).The number of queries registered in the query cache.
The total number of blocks in the query cache.
The number of statements executed by the server. This variable includes statements executed within stored programs, unlike the
Questionsvariable. It does not countCOM_PINGorCOM_STATISTICScommands.The number of statements executed by the server. This includes only statements sent to the server by clients and not statements executed within stored programs, unlike the
Queriesvariable. This variable does not countCOM_PING,COM_STATISTICS,COM_STMT_PREPARE,COM_STMT_CLOSE, orCOM_STMT_RESETcommands.The number of semisynchronous slaves.
This variable is available only if the master-side semisynchronous replication plugin is installed.
Rpl_semi_sync_master_net_avg_wait_timeThe average time in microseconds the master waited for a slave reply.
This variable is available only if the master-side semisynchronous replication plugin is installed.
Rpl_semi_sync_master_net_wait_timeThe total time in microseconds the master waited for slave replies.
This variable is available only if the master-side semisynchronous replication plugin is installed.
Rpl_semi_sync_master_net_waitsThe total number of times the master waited for slave replies.
This variable is available only if the master-side semisynchronous replication plugin is installed.
The number of times the master turned off semisynchronous replication.
This variable is available only if the master-side semisynchronous replication plugin is installed.
The number of commits that were not acknowledged successfully by a slave.
This variable is available only if the master-side semisynchronous replication plugin is installed.
Whether semisynchronous replication currently is operational on the master. The value is
ONif the plugin has been enabled and a commit acknowledgment has occurred. It isOFFif the plugin is not enabled or the master has fallen back to asynchronous replication due to commit acknowledgment timeout.This variable is available only if the master-side semisynchronous replication plugin is installed.
Rpl_semi_sync_master_timefunc_failuresThe number of times the master failed when calling time functions such as
gettimeofday().This variable is available only if the master-side semisynchronous replication plugin is installed.
Rpl_semi_sync_master_tx_avg_wait_timeThe average time in microseconds the master waited for each transaction.
This variable is available only if the master-side semisynchronous replication plugin is installed.
Rpl_semi_sync_master_tx_wait_timeThe total time in microseconds the master waited for transactions.
This variable is available only if the master-side semisynchronous replication plugin is installed.
The total number of times the master waited for transactions.
This variable is available only if the master-side semisynchronous replication plugin is installed.
Rpl_semi_sync_master_wait_pos_backtraverseThe total number of times the master waited for an event with binary coordinates lower than events waited for previously. This can occur when the order in which transactions start waiting for a reply is different from the order in which their binary log events are written.
This variable is available only if the master-side semisynchronous replication plugin is installed.
Rpl_semi_sync_master_wait_sessionsThe number of sessions currently waiting for slave replies.
This variable is available only if the master-side semisynchronous replication plugin is installed.
The number of commits that were acknowledged successfully by a slave.
This variable is available only if the master-side semisynchronous replication plugin is installed.
Whether semisynchronous replication currently is operational on the slave. This is
ONif the plugin has been enabled and the slave I/O thread is running,OFFotherwise.This variable is available only if the slave-side semisynchronous replication plugin is installed.
The RSA public key value used by the
sha256_passwordauthentication plugin. The value is nonempty only if the server successfully initializes the private and public keys in the files named by thesha256_password_private_key_pathandsha256_password_public_key_pathsystem variables. The value ofRsa_public_keycomes from the latter file.For information about
sha256_password, see Section 6.3.7.4, “The SHA-256 Authentication Plugin”.This variable is available only if MySQL was built using OpenSSL.
The number of joins that perform table scans because they do not use indexes. If this value is not 0, you should carefully check the indexes of your tables.
The number of joins that used a range search on a reference table.
The number of joins that used ranges on the first table. This is normally not a critical issue even if the value is quite large.
The number of joins without keys that check for key usage after each row. If this is not 0, you should carefully check the indexes of your tables.
The number of joins that did a full scan of the first table.
Shows the replication heartbeat interval (in seconds) on a replication slave.
Shows when the most recent heartbeat signal was received by a replication slave, as a
TIMESTAMPvalue.The number of temporary tables that the slave SQL thread currently has open. If the value is greater than zero, it is not safe to shut down the slave; see Section 16.4.1.22, “Replication and Temporary Tables”.
This counter increments with each replication heartbeat received by a replication slave since the last time that the slave was restarted or reset, or a
CHANGE MASTER TOstatement was issued.The total number of times since startup that the replication slave SQL thread has retried transactions.
This is
ONif this server is a replication slave that is connected to a replication master, and both the I/O and SQL threads are running; otherwise, it isOFF.The number of threads that have taken more than
slow_launch_timeseconds to create.This variable is not meaningful in the embedded server (
libmysqld) and as of MySQL 5.7.2 is no longer visible within the embedded server.The number of queries that have taken more than
long_query_timeseconds. This counter increments regardless of whether the slow query log is enabled. For information about that log, see Section 5.2.5, “The Slow Query Log”.The number of merge passes that the sort algorithm has had to do. If this value is large, you should consider increasing the value of the
sort_buffer_sizesystem variable.The number of sorts that were done using ranges.
The number of sorted rows.
The number of sorts that were done by scanning the table.
The number of negotiates needed to establish the connection.
The number of accepted SSL connections.
The number of callback cache hits.
The current SSL cipher (empty for non-SSL connections).
The list of possible SSL ciphers.
The number of SSL connection attempts to an SSL-enabled master.
The number of negotiates needed to establish the connection to an SSL-enabled master.
The SSL context verification depth (how many certificates in the chain are tested).
The SSL context verification mode.
The default SSL timeout.
The number of successful SSL connections to the server.
The number of successful slave connections to an SSL-enabled master.
The last date for which the SSL certificate is valid.
The first date for which the SSL certificate is valid.
The number of SSL session cache hits.
The number of SSL session cache misses.
The SSL session cache mode.
The number of SSL session cache overflows.
The SSL session cache size.
The number of SSL session cache timeouts.
How many SSL connections were reused from the cache.
Ssl_used_session_cache_entriesHow many SSL session cache entries were used.
The verification depth for replication SSL connections.
The verification mode for replication SSL connections.
The SSL protocol version of the connection.
The number of times that a request for a table lock could be granted immediately.
The number of times that a request for a table lock could not be granted immediately and a wait was needed. If this is high and you have performance problems, you should first optimize your queries, and then either split your table or tables or use replication.
The number of hits for open tables cache lookups.
The number of misses for open tables cache lookups.
The number of overflows for the open tables cache. This is the number of times, after a table is opened or closed, a cache instance has an unused entry and the size of the instance is larger than
table_open_cache/table_open_cache_instances.For the memory-mapped implementation of the log that is used by mysqld when it acts as the transaction coordinator for recovery of internal XA transactions, this variable indicates the largest number of pages used for the log since the server started. If the product of
Tc_log_max_pages_usedandTc_log_page_sizeis always significantly less than the log size, the size is larger than necessary and can be reduced. (The size is set by the--log-tc-sizeoption. Currently, this variable is unused: It is unneeded for binary log-based recovery, and the memory-mapped recovery log method is not used unless the number of storage engines capable of two-phase commit is greater than one. (InnoDBis the only applicable engine.)The page size used for the memory-mapped implementation of the XA recovery log. The default value is determined using
getpagesize(). Currently, this variable is unused for the same reasons as described forTc_log_max_pages_used.For the memory-mapped implementation of the recovery log, this variable increments each time the server was not able to commit a transaction and had to wait for a free page in the log. If this value is large, you might want to increase the log size (with the
--log-tc-sizeoption). For binary log-based recovery, this variable increments each time the binary log cannot be closed because there are two-phase commits in progress. (The close operation waits until all such transactions are finished.)The number of threads in the thread cache.
This variable is not meaningful in the embedded server (
libmysqld) and as of MySQL 5.7.2 is no longer visible within the embedded server.The number of currently open connections.
The number of threads created to handle connections. If
Threads_createdis big, you may want to increase thethread_cache_sizevalue. The cache miss rate can be calculated asThreads_created/Connections.The number of threads that are not sleeping.
The number of seconds that the server has been up.
The number of seconds since the most recent
FLUSH STATUSstatement.
The MySQL server can operate in different SQL modes, and can apply
these modes differently for different clients, depending on the
value of the sql_mode system variable. This
capability enables each application to tailor the server's
operating mode to its own requirements.
For answers to some questions that are often asked about server SQL modes in MySQL, see Section B.3, “MySQL 5.7 FAQ: Server SQL Mode”.
Modes define what SQL syntax MySQL should support and what kind of data validation checks it should perform. This makes it easier to use MySQL in different environments and to use MySQL together with other database servers.
When working with InnoDB tables, consider also
the innodb_strict_mode
configuration option. It enables additional error checks for
InnoDB tables, as listed in
Section 14.2.5.7, “InnoDB Strict Mode”.
Setting the SQL Mode
The default SQL mode in MySQL 5.7 is
NO_ENGINE_SUBSTITUTION.
To set the SQL mode at server startup, use the
--sql-mode="
option on the command line, or
modes"sql-mode="
in an option file such as modes"my.cnf (Unix
operating systems) or my.ini (Windows).
modes is a list of different modes
separated by commas. To clear the SQL mode explicitly, set it to
an empty string using --sql-mode=""
on the command line, or sql-mode=""
in an option file.
MySQL installation programs may configure the SQL mode during
the installation process. For example,
mysql_install_db creates a default option
file named my.cnf in the base installation
directory. This file contains a line that sets the SQL mode; see
Section 4.4.3, “mysql_install_db — Initialize MySQL Data Directory”.
If the SQL mode differs from the default or from what you expect, check for a setting in an option file that the server reads at startup.
To change the SQL mode at runtime, use a SET
[GLOBAL|SESSION]
sql_mode=' statement to
set the modes'sql_mode system variable.
Setting the GLOBAL variable requires the
SUPER privilege and affects the
operation of all clients that connect from that time on. Setting
the SESSION variable affects only the current
client. Any client can change its own session
sql_mode value at any time.
To determine the current global or session
sql_mode value, use the following
statements:
SELECT @@GLOBAL.sql_mode; SELECT @@SESSION.sql_mode;
SQL mode and user-defined partitioning. Changing the server SQL mode after creating and inserting data into partitioned tables can cause major changes in the behavior of such tables, and could lead to loss or corruption of data. It is strongly recommended that you never change the SQL mode once you have created tables employing user-defined partitioning.
When replicating partitioned tables, differing SQL modes on master and slave can also lead to problems. For best results, you should always use the same server SQL mode on the master and on the slave.
See Section 17.6, “Restrictions and Limitations on Partitioning”, for more information.
Most Important SQL Modes
The most important sql_mode
values are probably these:
This mode changes syntax and behavior to conform more closely to standard SQL. It is one of the special combination modes listed at the end of this section.
If a value could not be inserted as given into a transactional table, abort the statement. For a nontransactional table, abort the statement if the value occurs in a single-row statement or the first row of a multiple-row statement. More detail is given later in this section.
Make MySQL behave like a “traditional” SQL database system. A simple description of this mode is “give an error instead of a warning” when inserting an incorrect value into a column. It is one of the special combination modes listed at the end of this section.
When this manual refers to “strict mode,” it means a
mode where at least one of
STRICT_TRANS_TABLES or
STRICT_ALL_TABLES is enabled.
Full List of SQL Modes
The following list describes all supported modes:
Do not perform full checking of dates. Check only that the month is in the range from 1 to 12 and the day is in the range from 1 to 31. This is very convenient for Web applications where you obtain year, month, and day in three different fields and you want to store exactly what the user inserted (without date validation). This mode applies to
DATEandDATETIMEcolumns. It does not applyTIMESTAMPcolumns, which always require a valid date.The server requires that month and day values be legal, and not merely in the range 1 to 12 and 1 to 31, respectively. With strict mode disabled, invalid dates such as
'2004-04-31'are converted to'0000-00-00'and a warning is generated. With strict mode enabled, invalid dates generate an error. To permit such dates, enableALLOW_INVALID_DATES.Treat “
"” as an identifier quote character (like the “`” quote character) and not as a string quote character. You can still use “`” to quote identifiers with this mode enabled. WithANSI_QUOTESenabled, you cannot use double quotation marks to quote literal strings, because it is interpreted as an identifier.Produce an error in strict mode (otherwise a warning) when a division by zero (or
MOD(X,0)) occurs during anINSERTorUPDATE. If this mode is not enabled, MySQL instead returnsNULLfor divisions by zero. ForINSERT IGNOREorUPDATE IGNORE, MySQL generates a warning for divisions by zero, but the result of the operation isNULL.For
SELECT, division by zero returnsNULL. Enabling this mode causes a warning to be generated as well.The precedence of the
NOToperator is such that expressions such asNOT a BETWEEN b AND care parsed asNOT (a BETWEEN b AND c). In some older versions of MySQL, the expression was parsed as(NOT a) BETWEEN b AND c. The old higher-precedence behavior can be obtained by enabling theHIGH_NOT_PRECEDENCESQL mode.mysql>
SET sql_mode = '';mysql>SELECT NOT 1 BETWEEN -5 AND 5;-> 0 mysql>SET sql_mode = 'HIGH_NOT_PRECEDENCE';mysql>SELECT NOT 1 BETWEEN -5 AND 5;-> 1Permit spaces between a function name and the “
(” character. This causes built-in function names to be treated as reserved words. As a result, identifiers that are the same as function names must be quoted as described in Section 9.2, “Schema Object Names”. For example, because there is aCOUNT()function, the use ofcountas a table name in the following statement causes an error:mysql>
CREATE TABLE count (i INT);ERROR 1064 (42000): You have an error in your SQL syntaxThe table name should be quoted:
mysql>
CREATE TABLE `count` (i INT);Query OK, 0 rows affected (0.00 sec)The
IGNORE_SPACESQL mode applies to built-in functions, not to user-defined functions or stored functions. It is always permissible to have spaces after a UDF or stored function name, regardless of whetherIGNORE_SPACEis enabled.For further discussion of
IGNORE_SPACE, see Section 9.2.4, “Function Name Parsing and Resolution”.Prevent the
GRANTstatement from automatically creating new users if it would otherwise do so, unless authentication information is specified. The statement must specify a nonempty password usingIDENTIFIED BYor an authentication plugin usingIDENTIFIED WITH.NO_AUTO_VALUE_ON_ZEROaffects handling ofAUTO_INCREMENTcolumns. Normally, you generate the next sequence number for the column by inserting eitherNULLor0into it.NO_AUTO_VALUE_ON_ZEROsuppresses this behavior for0so that onlyNULLgenerates the next sequence number.This mode can be useful if
0has been stored in a table'sAUTO_INCREMENTcolumn. (Storing0is not a recommended practice, by the way.) For example, if you dump the table with mysqldump and then reload it, MySQL normally generates new sequence numbers when it encounters the0values, resulting in a table with contents different from the one that was dumped. EnablingNO_AUTO_VALUE_ON_ZERObefore reloading the dump file solves this problem. mysqldump now automatically includes in its output a statement that enablesNO_AUTO_VALUE_ON_ZERO, to avoid this problem.Disable the use of the backslash character (“
\”) as an escape character within strings. With this mode enabled, backslash becomes an ordinary character like any other.When creating a table, ignore all
INDEX DIRECTORYandDATA DIRECTORYdirectives. This option is useful on slave replication servers.Control automatic substitution of the default storage engine when a statement such as
CREATE TABLEorALTER TABLEspecifies a storage engine that is disabled or not compiled in.Because storage engines can be pluggable at runtime, unavailable engines are treated the same way:
With
NO_ENGINE_SUBSTITUTIONdisabled, forCREATE TABLEthe default engine is used and a warning occurs if the desired engine is unavailable. ForALTER TABLE, a warning occurs and the table is not altered.With
NO_ENGINE_SUBSTITUTIONenabled, an error occurs and the table is not created or altered if the desired engine is unavailable.Do not print MySQL-specific column options in the output of
SHOW CREATE TABLE. This mode is used by mysqldump in portability mode.Do not print MySQL-specific index options in the output of
SHOW CREATE TABLE. This mode is used by mysqldump in portability mode.Do not print MySQL-specific table options (such as
ENGINE) in the output ofSHOW CREATE TABLE. This mode is used by mysqldump in portability mode.By default, subtraction between integer operands produces an
UNSIGNEDresult if any operand isUNSIGNED. WhenNO_UNSIGNED_SUBTRACTIONis enabled, the subtraction result is signed, even if any operand is unsigned. For example, compare the type of columnc2in tablet1with that of columnc2in tablet2:mysql>
SET sql_mode='';mysql>CREATE TABLE test (c1 BIGINT UNSIGNED NOT NULL);mysql>CREATE TABLE t1 SELECT c1 - 1 AS c2 FROM test;mysql>DESCRIBE t1;+-------+---------------------+------+-----+---------+-------+ | Field | Type | Null | Key | Default | Extra | +-------+---------------------+------+-----+---------+-------+ | c2 | bigint(21) unsigned | | | 0 | | +-------+---------------------+------+-----+---------+-------+ mysql>SET sql_mode='NO_UNSIGNED_SUBTRACTION';mysql>CREATE TABLE t2 SELECT c1 - 1 AS c2 FROM test;mysql>DESCRIBE t2;+-------+------------+------+-----+---------+-------+ | Field | Type | Null | Key | Default | Extra | +-------+------------+------+-----+---------+-------+ | c2 | bigint(21) | | | 0 | | +-------+------------+------+-----+---------+-------+Note that this means that
BIGINT UNSIGNEDis not 100% usable in all contexts. See Section 12.10, “Cast Functions and Operators”.mysql>
SET sql_mode = '';mysql>SELECT CAST(0 AS UNSIGNED) - 1;+-------------------------+ | CAST(0 AS UNSIGNED) - 1 | +-------------------------+ | 18446744073709551615 | +-------------------------+ mysql>SET sql_mode = 'NO_UNSIGNED_SUBTRACTION';mysql>SELECT CAST(0 AS UNSIGNED) - 1;+-------------------------+ | CAST(0 AS UNSIGNED) - 1 | +-------------------------+ | -1 | +-------------------------+In strict mode, do not permit
'0000-00-00'as a valid date. You can still insert zero dates with theIGNOREoption. When not in strict mode, the date is accepted but a warning is generated.In strict mode, do not accept dates where the year part is nonzero but the month or day part is 0 (for example,
'0000-00-00'is legal but'2010-00-01'and'2010-01-00'are not). If used with theIGNOREoption, MySQL inserts a'0000-00-00'date for any such date. When not in strict mode, the date is accepted but a warning is generated.Do not permit queries for which the select list or
HAVINGlist orORDER BYlist refers to nonaggregated columns that are not named in theGROUP BYclause.The following queries are invalid with
ONLY_FULL_GROUP_BYenabled. The first is invalid becauseaddressin the select list is not named in theGROUP BYclause, and the second becausemax_agein theHAVINGclause is not named in theGROUP BYclause:mysql>
SELECT name, address, MAX(age) FROM t GROUP BY name;ERROR 1055 (42000): 't.address' isn't in GROUP BYmysql>
SELECT name, MAX(age) AS max_age FROM t GROUP BY name->HAVING max_age < 30;Empty set (0.00 sec) ERROR 1463 (42000): Non-grouping field 'max_age' is used in HAVING clauseIn the second example, the query could be rewritten to use
HAVING MAX(age)instead, so that the reference is to a column named in an aggregate function. (max_agefails because it is an aggregate function.)In addition, if a query has aggregate functions and no
GROUP BYclause, it cannot have nonaggregated columns in the select list orORDER BYlist:mysql>
SELECT name, MAX(age) FROM t;ERROR 1140 (42000): Mixing of GROUP columns (MIN(),MAX(),COUNT(),...) with no GROUP columns is illegal if there is no GROUP BY clauseFor more information, see Section 12.17.3, “MySQL Extensions to
GROUP BY”.By default, trailing spaces are trimmed from
CHARcolumn values on retrieval. IfPAD_CHAR_TO_FULL_LENGTHis enabled, trimming does not occur and retrievedCHARvalues are padded to their full length. This mode does not apply toVARCHARcolumns, for which trailing spaces are retained on retrieval.mysql>
CREATE TABLE t1 (c1 CHAR(10));Query OK, 0 rows affected (0.37 sec) mysql>INSERT INTO t1 (c1) VALUES('xy');Query OK, 1 row affected (0.01 sec) mysql>SET sql_mode = '';Query OK, 0 rows affected (0.00 sec) mysql>SELECT c1, CHAR_LENGTH(c1) FROM t1;+------+-----------------+ | c1 | CHAR_LENGTH(c1) | +------+-----------------+ | xy | 2 | +------+-----------------+ 1 row in set (0.00 sec) mysql>SET sql_mode = 'PAD_CHAR_TO_FULL_LENGTH';Query OK, 0 rows affected (0.00 sec) mysql>SELECT c1, CHAR_LENGTH(c1) FROM t1;+------------+-----------------+ | c1 | CHAR_LENGTH(c1) | +------------+-----------------+ | xy | 10 | +------------+-----------------+ 1 row in set (0.00 sec)Treat
||as a string concatenation operator (same asCONCAT()) rather than as a synonym forOR.Treat
REALas a synonym forFLOAT. By default, MySQL treatsREALas a synonym forDOUBLE.Enable strict mode for all storage engines. Invalid data values are rejected. Additional detail follows.
Enable strict mode for transactional storage engines, and when possible for nontransactional storage engines. Additional details follow.
Strict mode controls how MySQL handles invalid or missing values
in data-change statements such as
INSERT or
UPDATE. A value can be invalid for
several reasons. For example, it might have the wrong data type
for the column, or it might be out of range. A value is missing
when a new row to be inserted does not contain a value for a
non-NULL column that has no explicit
DEFAULT clause in its definition. (For a
NULL column, NULL is
inserted if the value is missing.)
For statements that do not change data, such as
SELECT, invalid values generate a
warning in strict mode, not an error.
For transactional tables, an error occurs for invalid or missing
values in a data-change statement when either of the
STRICT_ALL_TABLES or
STRICT_TRANS_TABLES modes are
enabled. The statement is aborted and rolled back.
For nontransactional tables, the behavior is the same for either mode, if the bad value occurs in the first row to be inserted or updated. The statement is aborted and the table remains unchanged. If the statement inserts or modifies multiple rows and the bad value occurs in the second or later row, the result depends on which strict option is enabled:
For
STRICT_ALL_TABLES, MySQL returns an error and ignores the rest of the rows. However, in this case, the earlier rows still have been inserted or updated. This means that you might get a partial update, which might not be what you want. To avoid this, it is best to use single-row statements because these can be aborted without changing the table.For
STRICT_TRANS_TABLES, MySQL converts an invalid value to the closest valid value for the column and insert the adjusted value. If a value is missing, MySQL inserts the implicit default value for the column data type. In either case, MySQL generates a warning rather than an error and continues processing the statement. Implicit defaults are described in Section 11.5, “Data Type Default Values”.
Strict mode disallows invalid date values such as
'2004-04-31'. It does not disallow dates with
zero month or day parts such as '2004-04-00' or
“zero” dates. To disallow these as well, enable the
NO_ZERO_IN_DATE and
NO_ZERO_DATE SQL modes in
addition to strict mode.
If you are not using strict mode (that is, neither
STRICT_TRANS_TABLES nor
STRICT_ALL_TABLES is enabled),
MySQL inserts adjusted values for invalid or missing values and
produces warnings. In strict mode, you can produce this behavior
by using INSERT
IGNORE or UPDATE
IGNORE. See Section 13.7.5.39, “SHOW WARNINGS Syntax”.
Strict mode does not affect whether foreign key constraints are
checked. foreign_key_checks can
be used for that. (See Section 5.1.4, “Server System Variables”.)
Combination Modes
The following special modes are provided as shorthand for combinations of mode values from the preceding list.
The descriptions include all mode values that are available in the most recent version of MySQL. For older versions, a combination mode does not include individual mode values that are not available except in newer versions.
Equivalent to
REAL_AS_FLOAT,PIPES_AS_CONCAT,ANSI_QUOTES,IGNORE_SPACE.ANSImode also causes the server to return an error for queries where a set functionSwith an outer referenceS(outer_ref)SELECT * FROM t1 WHERE t1.a IN (SELECT MAX(t1.b) FROM t2 WHERE ...);
Here,
MAX(t1.b)cannot aggregated in the outer query because it appears in theWHEREclause of that query. Standard SQL requires an error in this situation. IfANSImode is not enabled, the server treatsS(outer_ref)S(const)Equivalent to
PIPES_AS_CONCAT,ANSI_QUOTES,IGNORE_SPACE,NO_KEY_OPTIONS,NO_TABLE_OPTIONS,NO_FIELD_OPTIONS.Equivalent to
PIPES_AS_CONCAT,ANSI_QUOTES,IGNORE_SPACE,NO_KEY_OPTIONS,NO_TABLE_OPTIONS,NO_FIELD_OPTIONS,NO_AUTO_CREATE_USER.Equivalent to
PIPES_AS_CONCAT,ANSI_QUOTES,IGNORE_SPACE,NO_KEY_OPTIONS,NO_TABLE_OPTIONS,NO_FIELD_OPTIONS.Equivalent to
NO_FIELD_OPTIONS,HIGH_NOT_PRECEDENCE.Equivalent to
NO_FIELD_OPTIONS,HIGH_NOT_PRECEDENCE.Equivalent to
PIPES_AS_CONCAT,ANSI_QUOTES,IGNORE_SPACE,NO_KEY_OPTIONS,NO_TABLE_OPTIONS,NO_FIELD_OPTIONS,NO_AUTO_CREATE_USER.Equivalent to
PIPES_AS_CONCAT,ANSI_QUOTES,IGNORE_SPACE,NO_KEY_OPTIONS,NO_TABLE_OPTIONS,NO_FIELD_OPTIONS.Equivalent to
STRICT_TRANS_TABLES,STRICT_ALL_TABLES,NO_ZERO_IN_DATE,NO_ZERO_DATE,ERROR_FOR_DIVISION_BY_ZERO,NO_AUTO_CREATE_USER, andNO_ENGINE_SUBSTITUTION.
MySQL supports a plugin API that enables creation of server components. Plugins can be loaded at server startup, or loaded and unloaded at runtime without restarting the server. The components supported by this interface include, but are not limited to, storage engines, full-text parser plugins, partitioning support, and server extensions.
Server plugins must be loaded in to the server before they can be used. MySQL enables you to load a plugin at server startup or at runtime. It is also possible to control the activation of loaded plugins at startup, and to unload them at runtime.
Installing Plugins
Server plugins must be known to the server before they can be
used. A plugin can be made known several ways, as described
here. In the following descriptions,
plugin_name stands for a plugin name
such as innodb or csv.
Built-in plugins:
A plugin that is built in to the server is known by the server
automatically. Normally, the server enables the plugin at
startup, although this can be changed with the
--
option.
plugin_name
Plugins registered in the
mysql.plugin table:
The mysql.plugin table serves as a registry
of plugins. The server normally enables each plugin listed in
the table at startup, although whether a given plugin is enabled
can be changed with the
--
option. If the server is started with the
plugin_name--skip-grant-tables option, it
does not consult this table and does not load the plugins listed
there.
Plugins named with command-line options:
A plugin that is located in a plugin library file can be loaded
at server startup with the
--plugin-load option. Normally,
the server enables the plugin at startup, although this can be
changed with the
--
option.
plugin_name
The option value is a semicolon-separated list of
name=plugin_libraryname is the name of the
plugin, and plugin_library is the
name of the shared library that contains the plugin code. If a
plugin library is named without any preceding plugin name, the
server loads all plugins in the library. Each library file must
be located in the directory named by the
plugin_dir system variable.
This option does not register any plugin in the
mysql.plugin table. For subsequent restarts,
the server loads the plugin again only if
--plugin-load is given again.
That is, this option effects a one-time installation that
persists only for one server invocation.
--plugin-load enables plugins to
be loaded even when
--skip-grant-tables is given
(which causes the server to ignore the
mysql.plugin table).
--plugin-load also enables
plugins to be loaded at startup under configurations when
plugins cannot be loaded at runtime.
The --plugin-load-add option
complements the --plugin-load
option. --plugin-load-add adds a
plugin or plugins to the set of plugins to be loaded at startup.
The argument format is the same as for
--plugin-load.
--plugin-load-add can be used to
avoid specifying a large set of plugins as a single long
unwieldy --plugin-load. argument.
--plugin-load-add can be given in
the absence of --plugin-load, but
any instance of --plugin-load-add
that appears before
--plugin-load. has no effect
because --plugin-load resets the
set of plugins to load. In other words, these options:
--plugin-load=x --plugin-load-add=y
are equivalent to this option:
--plugin-load="x;y"
But these options:
--plugin-load-add=y --plugin-load=x
are equivalent to this option:
--plugin-load=x
Plugins installed with the
INSTALL PLUGIN
statement:
A plugin that is located in a plugin library file can be loaded
at runtime with the INSTALL
PLUGIN statement. The statement also registers the
plugin in the mysql.plugin table to cause the
server to load it on subsequent restarts. For this reason,
INSTALL PLUGIN requires the
INSERT privilege for the
mysql.plugin table.
If a plugin is named both using a
--plugin-load option and in the
mysql.plugin table, the server starts but
writes these messages to the error log:
100310 19:15:44 [ERROR] Function 'plugin_name' already exists 100310 19:15:44 [Warning] Couldn't load plugin named 'plugin_name' with soname 'plugin_object_file'.
Example: The --plugin-load option
installs a plugin at server startup. To install a plugin named
myplugin in a plugin library file named
somepluglib.so, use these lines in a
my.cnf file:
[mysqld] plugin-load=myplugin=somepluglib.so
In this case, the plugin is not registered in
mysql.plugin. Restarting the server without
the --plugin-load option causes
the plugin not to be loaded at startup.
Alternatively, the INSTALL PLUGIN
statement causes the server to load the plugin code from the
library file at runtime:
mysql> INSTALL PLUGIN myplugin SONAME 'somepluglib.so';
INSTALL PLUGIN also causes
“permanent” plugin registration: The server lists
the plugin in the mysql.plugin table to
ensure that it is loaded on subsequent server restarts.
Many plugins can be loaded either at server startup or at
runtime. However, if a plugin is designed such that it must be
loaded and initialized during server startup, use
--plugin-load rather than
INSTALL PLUGIN.
While a plugin is loaded, information about it is available at
runtime from several sources, such as the
INFORMATION_SCHEMA.PLUGINS table
and the SHOW PLUGINS statement.
For more information, see
Section 5.1.8.2, “Obtaining Server Plugin Information”.
Controlling Plugin Activation
If the server knows about a plugin when it starts (for example,
because the plugin is named using a
--plugin-load option or
registered in the mysql.plugin table), the
server loads and enables the plugin by default. It is possible
to control activation for such a plugin using a
--
startup option named after the plugin. In the following
descriptions, plugin_name[=value]plugin_name stands for
a plugin name such as innodb or
csv. As with other options, dashes and
underscores are interchangeable in option names. For example,
--my_plugin=ON and
--my-plugin=ON are equivalent.
--plugin_name=OFFTells the server to disable the plugin.
--plugin_name[=ON]Tells the server to enable the plugin. (Specifying the option as
--without a value has the same effect.) If the plugin fails to initialize, the server runs with the plugin disabled.plugin_name--plugin_name=FORCETells the server to enable the plugin, but if plugin initialization fails, the server does not start. In other words, this option forces the server to run with the plugin enabled or not at all.
--plugin_name=FORCE_PLUS_PERMANENTLike
FORCE, but in addition prevents the plugin from being unloaded at runtime. If a user attempts to do so withUNINSTALL PLUGIN, an error occurs.
The values OFF, ON,
FORCE, and
FORCE_PLUS_PERMANENT are not case sensitive.
The activation state for plugins is visible in the
LOAD_OPTION column of the
INFORMATION_SCHEMA.PLUGINS table.
Suppose that CSV,
BLACKHOLE, and ARCHIVE are
built-in pluggable storage engines and that you want the server
to load them at startup, subject to these conditions: The server
is permitted to run if CSV initialization
fails, but must require that BLACKHOLE
initialization succeeds, and ARCHIVE should
be disabled. To accomplish that, use these lines in an option
file:
[mysqld] csv=ON blackhole=FORCE archive=OFF
The
--enable-
option format is supported as a synonym for
plugin_name--.
The
plugin_name=ON--disable-
and
plugin_name--skip-
option formats are supported as synonyms for
plugin_name--.
plugin_name=OFF
If a plugin is disabled, either explicitly with
OFF or implicitly because it was enabled with
ON but failed to initialize, aspects of
server operation that require the plugin will change. For
example, if the plugin implements a storage engine, existing
tables for the storage engine become inaccessible, and attempts
to create new tables for the storage engine result in tables
that use the default storage engine unless the
NO_ENGINE_SUBSTITUTION SQL
mode has been enabled to cause an error to occur instead.
Disabling a plugin may require adjustment to other options. For
example, if you start the server using
--skip-innodb
to disable InnoDB, other
innodb_
options likely will need to be omitted from the startup command.
In addition, because xxxInnoDB is the
default storage engine, it will not start unless you specify
another available storage engine with
--default_storage_engine. You
must also set
--default_tmp_storage_engine.
Uninstalling Plugins
A plugin known to the server can be uninstalled to disable it at
runtime with the UNINSTALL PLUGIN
statement. The statement unloads the plugin and removes it from
the mysql.plugin table if it is registered
there. For this reason, UNINSTALL
PLUGIN statement requires the
DELETE privilege for the
mysql.plugin table. With the plugin no longer
registered in the table, the server will not load the plugin
automatically for subsequent restarts.
UNINSTALL PLUGIN can unload
plugins regardless of whether they were loaded with
INSTALL PLUGIN or
--plugin-load.
UNINSTALL PLUGIN is subject to
these exceptions:
It cannot unload plugins that are built in to the server. These can be identified as those that have a library name of
NULLin the output fromINFORMATION_SCHEMA.PLUGINSorSHOW PLUGINS.It cannot unload plugins for which the server was started with
--, which prevents plugin unloading at runtime. These can be identified from theplugin_name=FORCE_PLUS_PERMANENTLOAD_OPTIONcolumn of theINFORMATION_SCHEMA.PLUGINStable.
There are several ways to determine which plugins are installed in the server:
The
INFORMATION_SCHEMA.PLUGINStable contains a row for each loaded plugin. Any that have aPLUGIN_LIBRARYvalue ofNULLare built in and cannot be unloaded.mysql>
SELECT * FROM information_schema.PLUGINS\G*************************** 1. row *************************** PLUGIN_NAME: binlog PLUGIN_VERSION: 1.0 PLUGIN_STATUS: ACTIVE PLUGIN_TYPE: STORAGE ENGINE PLUGIN_TYPE_VERSION: 50158.0 PLUGIN_LIBRARY: NULL PLUGIN_LIBRARY_VERSION: NULL PLUGIN_AUTHOR: MySQL AB PLUGIN_DESCRIPTION: This is a pseudo storage engine to represent the binlog in a transaction PLUGIN_LICENSE: GPL LOAD_OPTION: FORCE ... *************************** 10. row *************************** PLUGIN_NAME: InnoDB PLUGIN_VERSION: 1.0 PLUGIN_STATUS: ACTIVE PLUGIN_TYPE: STORAGE ENGINE PLUGIN_TYPE_VERSION: 50158.0 PLUGIN_LIBRARY: ha_innodb_plugin.so PLUGIN_LIBRARY_VERSION: 1.0 PLUGIN_AUTHOR: Innobase Oy PLUGIN_DESCRIPTION: Supports transactions, row-level locking, and foreign keys PLUGIN_LICENSE: GPL LOAD_OPTION: ON ...The
SHOW PLUGINSstatement displays a row for each loaded plugin. Any that have aLibraryvalue ofNULLare built in and cannot be unloaded.mysql>
SHOW PLUGINS\G*************************** 1. row *************************** Name: binlog Status: ACTIVE Type: STORAGE ENGINE Library: NULL License: GPL ... *************************** 10. row *************************** Name: InnoDB Status: ACTIVE Type: STORAGE ENGINE Library: ha_innodb_plugin.so License: GPL ...The
mysql.plugintable shows which plugins have been registered withINSTALL PLUGIN. The table contains only plugin names and library file names, so it does not provide as much information as thePLUGINStable or theSHOW PLUGINSstatement.
Support for IPv6 in MySQL includes these capabilities:
MySQL Server can accept TCP/IP connections from clients connecting over IPv6. For example, this command connects over IPv6 to the MySQL server on the local host:
shell>
mysql -h ::1To use this capability, two things must be true:
Your system must be configured to support IPv6. See Section 5.1.9.1, “Verifying System Support for IPv6”.
The default MySQL server configuration permits only IPv4 connections, so the server must be configured for IPv6 connections. To permit IPv6 connections in addition to or instead of IPv4 connections, start the server with an appropriate
--bind-addressoption. See Section 5.1.4, “Server System Variables”.
MySQL account names permit IPv6 addresses to enable DBAs to specify privileges for clients that connect to the server over IPv6. See Section 6.2.3, “Specifying Account Names”. IPv6 addresses can be specified in account names in statements such as
CREATE USER,GRANT, andREVOKE. For example:mysql>
CREATE USER 'bill'@'::1' IDENTIFIED BY 'secret';mysql>GRANT SELECT ON mydb.* TO 'bill'@'::1';IPv6 functions enable conversion between string and internal format IPv6 address formats, and checking whether values represent valid IPv6 addresses. For example,
INET6_ATON()andINET6_NTOA()are similar toINET_ATON()andINET_NTOA(), but handle IPv6 addresses in addition to IPv4 addresses. See Section 12.16, “Miscellaneous Functions”.
The following sections describe how to set up MySQL so that clients can connect to the server over IPv6.
Before MySQL Server can accept IPv6 connections, the operating system on your server host must support IPv6. As a simple test to determine whether that is true, try this command:
shell> ping6 ::1
16 bytes from ::1, icmp_seq=0 hlim=64 time=0.171 ms
16 bytes from ::1, icmp_seq=1 hlim=64 time=0.077 ms
...
To produce a description of your system's network interfaces, invoke ifconfig -a and look for IPv6 addresses in the output.
If your host does not support IPv6, consult your system documentation for instructions on enabling it. It might be that you need only reconfigure an existing network interface to add an IPv6 address. Or a more extensive change might be needed, such as rebuilding the kernel with IPv6 options enabled.
These links may be helpful in setting up IPv6 on various platforms:
The MySQL server listens on a single network socket for TCP/IP
connections. This socket is bound to a single address, but it is
possible for an address to map onto multiple network interfaces.
To specify an address, use the
--bind-address=
option at server startup, where addraddr
is an IPv4 or IPv6 address or a host name. (IPv6 addresses are
not supported before MySQL 5.5.3.) If
addr is a host name, the server
resolves the name to an IP address and binds to that address.
The server treats different types of addresses as follows:
If the address is
*, the server accepts TCP/IP connections on all server host IPv6 and IPv4 interfaces if the server host supports IPv6, or accepts TCP/IP connections on all IPv4 addresses otherwise. Use this address to permit both IPv4 and IPv6 connections on all server interfaces. This value is the default.If the address is
0.0.0.0, the server accepts TCP/IP connections on all server host IPv4 interfaces.If the address is
::, the server accepts TCP/IP connections on all server host IPv4 and IPv6 interfaces. Use this address to permit both IPv4 and IPv6 connections on all server interfaces.If the address is an IPv4-mapped address, the server accepts TCP/IP connections for that address, in either IPv4 or IPv6 format. For example, if the server is bound to
::ffff:127.0.0.1, clients can connect using--host=127.0.0.1or--host=::ffff:127.0.0.1.If the address is a “regular” IPv4 or IPv6 address (such as
127.0.0.1or::1), the server accepts TCP/IP connections only for that IPv4 or IPv6 address.
If you intend to bind the server to a specific address, be sure
that the mysql.user grant table contains an
account with administrative privileges that you can use to
connect to that address. Otherwise, you will not be able to shut
down the server. For example, if you bind the server to
*, you can connect to it using all existing
accounts. But if you bind the server to ::1,
it accepts connections only on that address. In that case, first
make sure that the 'root'@'::1' account is
present in the mysql.user table so you can
still connect to the server to shut it down.
The following procedure shows how to configure MySQL to permit
IPv6 connections by clients that connect to the local server
using the ::1 local host address. The
instructions given here assume that your system supports IPv6.
Start the MySQL server with an appropriate
--bind-addressoption to permit it to accept IPv6 connections. For example, put the following lines in your server option file and restart the server:[mysqld] bind-address = *
Alternatively, you can bind the server to
::1, but that makes the server more restrictive for TCP/IP connections. It accepts only IPv6 connections for that single address and rejects IPv4 connections. For more information, see Section 5.1.9.2, “Configuring the MySQL Server to Permit IPv6 Connections”.As an administrator, connect to the server and create an account for a local user who will connect from the
::1local IPv6 host address:mysql>
CREATE USER 'ipv6user'@'::1' IDENTIFIED BY 'ipv6pass';For the permitted syntax of IPv6 addresses in account names, see Section 6.2.3, “Specifying Account Names”. In addition to the
CREATE USERstatement, you can issueGRANTstatements that give specific privileges to the account, although that is not necessary for the remaining steps in this procedure.Invoke the mysql client to connect to the server using the new account:
shell>
mysql -h ::1 -u ipv6user -pipv6passTry some simple statements that show connection information:
mysql>
STATUS... Connection: ::1 via TCP/IP ... mysql>SELECT CURRENT_USER(), @@bind_address;+----------------+----------------+ | CURRENT_USER() | @@bind_address | +----------------+----------------+ | ipv6user@::1 | :: | +----------------+----------------+
The following procedure shows how to configure MySQL to permit IPv6 connections by remote clients. It is similar to the preceding procedure for local clients, but the server and client hosts are distinct and each has its own nonlocal IPv6 address. The example uses these addresses:
Server host: 2001:db8:0:f101::1 Client host: 2001:db8:0:f101::2
These addresses are chosen from the nonroutable address range recommended by IANA for documentation purposes and suffice for testing on your local network. To accept IPv6 connections from clients outside the local network, the server host must have a public address. If your network provider assigns you an IPv6 address, you can use that. Otherwise, another way to obtain an address is to use an IPv6 broker; see Section 5.1.9.5, “Obtaining an IPv6 Address from a Broker”.
Start the MySQL server with an appropriate
--bind-addressoption to permit it to accept IPv6 connections. For example, put the following lines in your server option file and restart the server:[mysqld] bind-address = *
Alternatively, you can bind the server to
2001:db8:0:f101::1, but that makes the server more restrictive for TCP/IP connections. It accepts only IPv6 connections for that single address and rejects IPv4 connections. For more information, see Section 5.1.9.2, “Configuring the MySQL Server to Permit IPv6 Connections”.On the server host (
2001:db8:0:f101::1), create an account for a user who will connect from the client host (2001:db8:0:f101::2):mysql>
CREATE USER 'remoteipv6user'@'2001:db8:0:f101::2' IDENTIFIED BY 'remoteipv6pass';On the client host (
2001:db8:0:f101::2), invoke the mysql client to connect to the server using the new account:shell>
mysql -h 2001:db8:0:f101::1 -u remoteipv6user -premoteipv6passTry some simple statements that show connection information:
mysql>
STATUS... Connection: 2001:db8:0:f101::1 via TCP/IP ... mysql>SELECT CURRENT_USER(), @@bind_address;+-----------------------------------+----------------+ | CURRENT_USER() | @@bind_address | +-----------------------------------+----------------+ | remoteipv6user@2001:db8:0:f101::2 | :: | +-----------------------------------+----------------+
If you do not have a public IPv6 address that enables your system to communicate over IPv6 outside your local network, you can obtain one from an IPv6 broker. The Wikipedia IPv6 Tunnel Broker page lists several brokers and their features, such as whether they provide static addresses and the supported routing protocols.
After configuring your server host to use a broker-supplied IPv6
address, start the MySQL server with an appropriate
--bind-address option to permit
the server to accept IPv6 connections. For example, put the
following lines in the server option file and restart the
server:
[mysqld] bind-address = *
Alternatively, you can bind the server to to the specific IPv6 address provided by the broker, but that makes the server more restrictive for TCP/IP connections. It accepts only IPv6 connections for that single address and rejects IPv4 connections. For more information, see Section 5.1.9.2, “Configuring the MySQL Server to Permit IPv6 Connections”. In addition, if the broker allocates dynamic addresses, the address provided for your system might change the next time you connect to the broker. If so, any accounts you create that name the original address become invalid. To bind to a specific address but avoid this change-of-address problem, you may be able to arrange with the broker for a static IPv6 address.
The following example shows how to use Freenet6 as the broker and the gogoc IPv6 client package on Gentoo Linux.
Create a account at Freenet6 by visiting this URL and signing up:
http://gogonet.gogo6.com
After creating the account, go to this URL, sign in, and create a user ID and password for the IPv6 broker:
http://gogonet.gogo6.com/page/freenet6-registration
As
root, install gogoc:shell>
emerge gogocEdit
/etc/gogoc/gogoc.confto set theuseridandpasswordvalues. For example:userid=gogouser passwd=gogopass
Start gogoc:
shell>
/etc/init.d/gogoc startTo start gogoc each time your system boots, execute this command:
shell>
rc-update add gogoc defaultUse ping6 to try to ping a host:
shell>
ping6 ipv6.google.comTo see your IPv6 address:
shell>
ifconfig tun
MySQL Server supports a HELP
statement that returns online information from the MySQL Reference
manual (see Section 13.8.3, “HELP Syntax”). The proper operation of this
statement requires that the help tables in the
mysql database be initialized with help topic
information, which is done by processing the contents of the
fill_help_tables.sql script.
If you install MySQL using a binary or source distribution on Unix, help table setup occurs when you run mysql_install_db. For an RPM distribution on Linux or binary distribution on Windows, help table setup occurs as part of the MySQL installation process.
If you upgrade MySQL using a binary distribution, the help tables
are not upgraded automatically, but you can upgrade them manually.
Locate the fill_help_tables.sql file in the
share or share/mysql
directory. Change location into that directory and process the
file with the mysql client as follows:
shell> mysql -u root mysql < fill_help_tables.sql
You can also obtain the latest
fill_help_tables.sql at any time to upgrade
your help tables. Download the proper file for your version of
MySQL from http://dev.mysql.com/doc/index-other.html. After
downloading and uncompressing the file, process it with
mysql as described previously.
If you are working with Bazaar and a MySQL development source
tree, you will need to download the
fill_help_tables.sql file because the tree
contains only a “stub” version.
On Unix, signals can be sent to processes. mysqld responds to signals sent to it as follows:
SIGTERMcauses the server to shut down.SIGHUPcauses the server to reload the grant tables and to flush tables, logs, the thread cache, and the host cache. These actions are like various forms of theFLUSHstatement. The server also writes a status report to the error log that has this format:Status information: Current dir: /var/mysql/data/ Running threads: 0 Stack size: 196608 Current locks: Key caches: default Buffer_size: 8388600 Block_size: 1024 Division_limit: 100 Age_limit: 300 blocks used: 0 not flushed: 0 w_requests: 0 writes: 0 r_requests: 0 reads: 0 handler status: read_key: 0 read_next: 0 read_rnd 0 read_first: 1 write: 0 delete 0 update: 0 Table status: Opened tables: 5 Open tables: 0 Open files: 7 Open streams: 0 Alarm status: Active alarms: 1 Max used alarms: 2 Next alarm time: 67
On some Mac OS X 10.3 versions, mysqld ignores
SIGHUP and SIGQUIT.
The server shutdown process takes place as follows:
The shutdown process is initiated.
This can occur initiated several ways. For example, a user with the
SHUTDOWNprivilege can execute a mysqladmin shutdown command. mysqladmin can be used on any platform supported by MySQL. Other operating system-specific shutdown initiation methods are possible as well: The server shuts down on Unix when it receives aSIGTERMsignal. A server running as a service on Windows shuts down when the services manager tells it to.The server creates a shutdown thread if necessary.
Depending on how shutdown was initiated, the server might create a thread to handle the shutdown process. If shutdown was requested by a client, a shutdown thread is created. If shutdown is the result of receiving a
SIGTERMsignal, the signal thread might handle shutdown itself, or it might create a separate thread to do so. If the server tries to create a shutdown thread and cannot (for example, if memory is exhausted), it issues a diagnostic message that appears in the error log:Error: Can't create thread to kill server
The server stops accepting new connections.
To prevent new activity from being initiated during shutdown, the server stops accepting new client connections by closing the handlers for the network interfaces to which it normally listens for connections: the TCP/IP port, the Unix socket file, the Windows named pipe, and shared memory on Windows.
The server terminates current activity.
For each thread associated with a client connection, the server breaks the connection to the client and marks the thread as killed. Threads die when they notice that they are so marked. Threads for idle connections die quickly. Threads that currently are processing statements check their state periodically and take longer to die. For additional information about thread termination, see Section 13.7.6.4, “
KILLSyntax”, in particular for the instructions about killedREPAIR TABLEorOPTIMIZE TABLEoperations onMyISAMtables.For threads that have an open transaction, the transaction is rolled back. Note that if a thread is updating a nontransactional table, an operation such as a multiple-row
UPDATEorINSERTmay leave the table partially updated because the operation can terminate before completion.If the server is a master replication server, it treats threads associated with currently connected slaves like other client threads. That is, each one is marked as killed and exits when it next checks its state.
If the server is a slave replication server, it stops the I/O and SQL threads, if they are active, before marking client threads as killed. The SQL thread is permitted to finish its current statement (to avoid causing replication problems), and then stops. If the SQL thread is in the middle of a transaction at this point, the server waits until the current replication event group (if any) has finished executing, or until the user issues a
KILL QUERYorKILL CONNECTIONstatement. See also Section 13.4.2.6, “STOP SLAVESyntax”. Since nontransactional statements cannot be rolled back, in order to guarantee crash-safe replication, only transactional tables should be used.NoteIn order to guarantee crash safety on the slave, you must also run the slave with
--relay-log-recoveryenabled.See also Section 16.2.2, “Replication Relay and Status Logs”).
The server shuts down or closes storage engines.
At this stage, the server flushes the table cache and closes all open tables.
Each storage engine performs any actions necessary for tables that it manages.
InnoDBflushes its buffer pool to disk (unlessinnodb_fast_shutdownis 2), writes the current LSN to the tablespace, and terminates its own internal threads.MyISAMflushes any pending index writes for a table.The server exits.
MySQL Server has several logs that can help you find out what activity is taking place.
| Log Type | Information Written to Log |
|---|---|
| Error log | Problems encountered starting, running, or stopping mysqld |
| General query log | Established client connections and statements received from clients |
| Binary log | Statements that change data (also used for replication) |
| Relay log | Data changes received from a replication master server |
| Slow query log | Queries that took more than
long_query_time seconds to
execute |
By default, no logs are enabled (except the error log on Windows). The following log-specific sections provide information about the server options that enable logging.
By default, the server writes files for all enabled logs in the data
directory. You can force the server to close and reopen the log
files (or in some cases switch to a new log file) by flushing the
logs. Log flushing occurs when you issue a
FLUSH LOGS
statement; execute mysqladmin with a
flush-logs or refresh
argument; or execute mysqldump with a
--flush-logs or
--master-data option. See
Section 13.7.6.3, “FLUSH Syntax”, Section 4.5.2, “mysqladmin — Client for Administering a MySQL Server”, and
Section 4.5.4, “mysqldump — A Database Backup Program”. In addition, the binary log is flushed
when its size reaches the value of the
max_binlog_size system variable.
You can control the general query and slow query logs during runtime. You can enable or disable logging, or change the log file name. You can tell the server to write general query and slow query entries to log tables, log files, or both. For details, see Section 5.2.1, “Selecting General Query and Slow Query Log Output Destinations”, Section 5.2.3, “The General Query Log”, and Section 5.2.5, “The Slow Query Log”.
The relay log is used only on slave replication servers, to hold data changes from the master server that must also be made on the slave. For discussion of relay log contents and configuration, see Section 16.2.2.1, “The Slave Relay Log”.
For information about log maintenance operations such as expiration of old log files, see Section 5.2.6, “Server Log Maintenance”.
For information about keeping logs secure, see Section 6.1.2.3, “Passwords and Logging”.
MySQL Server provides flexible control over the destination of
output to the general query log and the slow query log, if those
logs are enabled. Possible destinations for log entries are log
files or the general_log and
slow_log tables in the mysql
database. Either or both destinations can be selected.
Log control at server startup.
The --log-output option specifies
the destination for log output. This option does not in itself
enable the logs. Its syntax is
--log-output[=:
value,...]
If
--log-outputis given with a value, the value should be a comma-separated list of one or more of the wordsTABLE(log to tables),FILE(log to files), orNONE(do not log to tables or files).NONE, if present, takes precedence over any other specifiers.If
--log-outputis omitted, the default logging destination isFILE.
The general_log system variable
controls logging to the general query log for the selected log
destinations. If specified at server startup,
general_log takes an optional
argument of 1 or 0 to enable or disable the log. To specify a file
name other than the default for file logging, set the
general_log_file variable.
Similarly, the slow_query_log
variable controls logging to the slow query log for the selected
destinations and setting
slow_query_log_file specifies a
file name for file logging. If either log is enabled, the server
opens the corresponding log file and writes startup messages to
it. However, further logging of queries to the file does not occur
unless the FILE log destination is selected.
Examples:
To write general query log entries to the log table and the log file, use
--log-output=TABLE,FILEto select both log destinations and--general_logto enable the general query log.To write general and slow query log entries only to the log tables, use
--log-output=TABLEto select tables as the log destination and--general_logand--slow_query_logto enable both logs.To write slow query log entries only to the log file, use
--log-output=FILEto select files as the log destination and--slow_query_logto enable the slow query log. (In this case, because the default log destination isFILE, you could omit the--log-outputoption.)
Log control at runtime. The system variables associated with log tables and files enable runtime control over logging:
The global
log_outputsystem variable indicates the current logging destination. It can be modified at runtime to change the destination.The global
general_logandslow_query_logvariables indicate whether the general query log and slow query log are enabled (ON) or disabled (OFF). You can set these variables at runtime to control whether the logs are enabled.The global
general_log_fileandslow_query_log_filevariables indicate the names of the general query log and slow query log files. You can set these variables at server startup or at runtime to change the names of the log files.To disable or enable general query logging for the current connection, set the session
sql_log_offvariable toONorOFF.
The use of tables for log output offers the following benefits:
Log entries have a standard format. To display the current structure of the log tables, use these statements:
SHOW CREATE TABLE mysql.general_log; SHOW CREATE TABLE mysql.slow_log;
Log contents are accessible through SQL statements. This enables the use of queries that select only those log entries that satisfy specific criteria. For example, to select log contents associated with a particular client (which can be useful for identifying problematic queries from that client), it is easier to do this using a log table than a log file.
Logs are accessible remotely through any client that can connect to the server and issue queries (if the client has the appropriate log table privileges). It is not necessary to log in to the server host and directly access the file system.
The log table implementation has the following characteristics:
In general, the primary purpose of log tables is to provide an interface for users to observe the runtime execution of the server, not to interfere with its runtime execution.
CREATE TABLE,ALTER TABLE, andDROP TABLEare valid operations on a log table. ForALTER TABLEandDROP TABLE, the log table cannot be in use and must be disabled, as described later.By default, the log tables use the
CSVstorage engine that writes data in comma-separated values format. For users who have access to the.CSVfiles that contain log table data, the files are easy to import into other programs such as spreadsheets that can process CSV input.The log tables can be altered to use the
MyISAMstorage engine. You cannot useALTER TABLEto alter a log table that is in use. The log must be disabled first. No engines other thanCSVorMyISAMare legal for the log tables.To disable logging so that you can alter (or drop) a log table, you can use the following strategy. The example uses the general query log; the procedure for the slow query log is similar but uses the
slow_logtable andslow_query_logsystem variable.SET @old_log_state = @@global.general_log; SET GLOBAL general_log = 'OFF'; ALTER TABLE mysql.general_log ENGINE = MyISAM; SET GLOBAL general_log = @old_log_state;
TRUNCATE TABLEis a valid operation on a log table. It can be used to expire log entries.RENAME TABLEis a valid operation on a log table. You can atomically rename a log table (to perform log rotation, for example) using the following strategy:USE mysql; DROP TABLE IF EXISTS general_log2; CREATE TABLE general_log2 LIKE general_log; RENAME TABLE general_log TO general_log_backup, general_log2 TO general_log;
CHECK TABLEis a valid operation on a log table.LOCK TABLEScannot be used on a log table.INSERT,DELETE, andUPDATEcannot be used on a log table. These operations are permitted only internally to the server itself.FLUSH TABLES WITH READ LOCKand the state of the globalread_onlysystem variable have no effect on log tables. The server can always write to the log tables.Entries written to the log tables are not written to the binary log and thus are not replicated to slave servers.
To flush the log tables or log files, use
FLUSH TABLESorFLUSH LOGS, respectively.Partitioning of log tables is not permitted.
A mysqldump dump includes statements to recreate those tables so that they are not missing after reloading the dump file. Log table contents are not dumped.
The error log contains information indicating when mysqld was started and stopped and also any critical errors that occur while the server is running. If mysqld notices a table that needs to be automatically checked or repaired, it writes a message to the error log.
On some operating systems, the error log contains a stack trace if mysqld dies. The trace can be used to determine where mysqld died. See Section 22.4, “Debugging and Porting MySQL”.
In the following discussion, “console” means
stderr, the standard error output; this is your
terminal or console window unless the standard error output has
been redirected. (For example, if invoked with the
--syslog option,
mysqld_safe arranges for the server's
stderr to be sent to the
syslog facility, as described later.)
On Windows, the --log-error and
--console options both affect error
logging:
Without
--log-error, mysqld writes error messages tohost_name.errWith
--log-error[=, mysqld writes error messages to an error log file. The server uses the named file if present, creating it in in the data directory unless an absolute path name is given to specify a different directory. If no file is named, the default name isfile_name]host_name.errIf
--consoleis given, mysqld writes error messages to the console.--log-error, if given, is ignored and has no effect. If both options are present, their order does not matter:--consoletakes precedence and error messages go to the console. (In MySQL 5.5 and 5.6, the precedence is reversed:--log-errorcauses--consoleto be ignored.)
In addition, on Windows, events and error messages are written to
the Windows Event Log within the Application log. Entries marked
as Warning and Note are
written to the Event Log, but not informational messages such as
information statements from individual storage engines. These log
entries have a source of MySQL. You cannot
disable writing information to the Windows Event Log.
On Unix and Unix-like systems, mysqld writes error log messages as follows:
Without
--log-error, mysqld writes error messages to the console.With
--log-error[=, mysqld writes error messages to an error log file. The server uses the named file if present, creating it in the data directory unless an absolute path name is given to specify a different directory. If no file is named, the default name isfile_name]host_name.err
At runtime, if the server writes error messages to the console, it
sets the log_error system
variable to stderr. Otherwise,
log_error indicates the error log
file name. In particular, on Windows,
--console overrides use of an error
log file and sends error messages to the console, so
log_error is set to
stderr. This occurs even if
--log-error is also given.
If you flush the logs using
FLUSH LOGS or
mysqladmin flush-logs and
mysqld is writing the error log to a file (for
example, if it was started with the
--log-error option), the server
closes and reopens the log file. To rename the file, do so
manually before flushing. Then flushing the logs reopens a new
file with the original file name. For example, you can rename the
file and create a new one using the following commands:
shell>mvshell>host_name.errhost_name.err-oldmysqladmin flush-logsshell>mvhost_name.err-oldbackup-directory
On Windows, use rename rather than mv.
No error log renaming occurs when the logs are flushed if the server is not writing to a named file.
If you use mysqld_safe to start
mysqld, mysqld_safe arranges
for mysqld to write error messages to a log
file or to syslog.
mysqld_safe has three error-logging options,
--syslog,
--skip-syslog,
and --log-error. The default
with no logging options or with
--skip-syslog
is to use the default log file. To explicitly specify use of an
error log file, specify
--log-error=
to mysqld_safe, and
mysqld_safe will arrange for
mysqld to write messages to a log file. To use
file_namesyslog instead, specify the
--syslog option.
If you specify --log-error in an
option file in a [mysqld],
[server], or [mysqld_safe]
section, mysqld_safe will find and use the
option.
If mysqld_safe is used to start
mysqld and mysqld dies
unexpectedly, mysqld_safe notices that it needs
to restart mysqld and writes a
restarted mysqld message to the error log.
The --log-warnings option or
log_warnings system variable can
be used to control warning logging to the error log. The default
value is enabled (1). Warning logging can be disabled using a
value of 0. If the value is greater than 1, aborted connections
are written to the error log, and access-denied errors for new
connection attempts are written. See
Section C.5.2.11, “Communication Errors and Aborted Connections”.
The general query log is a general record of what mysqld is doing. The server writes information to this log when clients connect or disconnect, and it logs each SQL statement received from clients. The general query log can be very useful when you suspect an error in a client and want to know exactly what the client sent to mysqld.
mysqld writes statements to the query log in the order that it receives them, which might differ from the order in which they are executed. This logging order is in contrast with that of the binary log, for which statements are written after they are executed but before any locks are released. In addition, the query log may contain statements that only select data while such statements are never written to the binary log.
When using statement-based logging all statements are written to
the query log, but when using row-based logging, updates are sent
as row changes rather than SQL statements, and thus these
statements are never written to the query log when
binlog_format is
ROW. A given update also might not be written
to the query log when this variable is set to
MIXED, depending on the statement used. See
Section 16.1.2.1, “Advantages and Disadvantages of Statement-Based and Row-Based
Replication”, for more information.
By default, the general query log is disabled. To specify the
initial general query log state explicitly, use
--general_log[={0|1}]. With no
argument or an argument of 1,
--general_log enables the log. With
an argument of 0, this option disables the log. To specify a log
file name, use
--general_log_file=.
To specify the log destination, use
file_name--log-output (as described in
Section 5.2.1, “Selecting General Query and Slow Query Log Output Destinations”).
If you specify no name for the general query log file, the default
name is
host_name.log
To disable or enable the general query log or change the log file
name at runtime, use the global
general_log and
general_log_file system
variables. Set general_log to 0
(or OFF) to disable the log or to 1 (or
ON) to enable it. Set
general_log_file to specify the
name of the log file. If a log file already is open, it is closed
and the new file is opened.
When the general query log is enabled, the server writes output to
any destinations specified by the
--log-output option or
log_output system variable. If
you enable the log, the server opens the log file and writes
startup messages to it. However, further logging of queries to the
file does not occur unless the FILE log
destination is selected. If the destination is
NONE, the server writes no queries even if the
general log is enabled. Setting the log file name has no effect on
logging if the log destination value does not contain
FILE.
Server restarts and log flushing do not cause a new general query log file to be generated (although flushing closes and reopens it). To rename the file and create a new one, use the following commands:
shell>mvshell>host_name.loghost_name-old.logmysqladmin flush-logsshell>mvhost_name-old.logbackup-directory
On Windows, use rename rather than mv.
You can also rename the general query log file at runtime by disabling the log:
SET GLOBAL general_log = 'OFF';
With the log disabled, rename the log file externally; for example, from the command line. Then enable the log again:
SET GLOBAL general_log = 'ON';
This method works on any platform and does not require a server restart.
The session sql_log_off variable
can be set to ON or OFF to
disable or enable general query logging for the current
connection.
Passwords in statements written to the general query log are
rewritten by the server not to occur literally in plain text.
Password rewriting can be suppressed for the general query log by
starting the server with the
--log-raw option. This option may
be useful for diagnostic purposes, to see the exact text of
statements as received by the server, but for security reasons is
not recommended for production use. See also
Section 6.1.2.3, “Passwords and Logging”.
The binary log contains “events” that describe
database changes such as table creation operations or changes to
table data. It also contains events for statements that
potentially could have made changes (for example, a
DELETE which matched no rows),
unless row-based logging is used. The binary log also contains
information about how long each statement took that updated data.
The binary log has two important purposes:
For replication, the binary log on a master replication server provides a record of the data changes to be sent to slave servers. The master server sends the events contained in its binary log to its slaves, which execute those events to make the same data changes that were made on the master. See Section 16.2, “Replication Implementation”.
Certain data recovery operations require use of the binary log. After a backup has been restored, the events in the binary log that were recorded after the backup was made are re-executed. These events bring databases up to date from the point of the backup. See Section 7.5, “Point-in-Time (Incremental) Recovery Using the Binary Log”.
The binary log is not used for statements such as
SELECT or
SHOW that do not modify data. To
log all statements (for example, to identify a problem query), use
the general query log. See Section 5.2.3, “The General Query Log”.
Running a server with binary logging enabled makes performance slightly slower. However, the benefits of the binary log in enabling you to set up replication and for restore operations generally outweigh this minor performance decrement.
The binary log is crash-safe. Only complete events or transactions are logged or read back.
Passwords in statements written to the binary log are rewritten by the server not to occur literally in plain text. See also Section 6.1.2.3, “Passwords and Logging”.
The following discussion describes some of the server options and variables that affect the operation of binary logging. For a complete list, see Section 16.1.4.4, “Binary Log Options and Variables”.
To enable the binary log, start the server with the
--log-bin[=
option. If no base_name]base_name value is given,
the default name is the value of the pid-file
option (which by default is the name of host machine) followed by
-bin. If the basename is given, the server
writes the file in the data directory unless the basename is given
with a leading absolute path name to specify a different
directory. It is recommended that you specify a basename
explicitly rather than using the default of the host name; see
Section C.5.8, “Known Issues in MySQL”, for the reason.
If you supply an extension in the log name (for example,
--log-bin=),
the extension is silently removed and ignored.
base_name.extension
mysqld appends a numeric extension to the
binary log basename to generate binary log file names. The number
increases each time the server creates a new log file, thus
creating an ordered series of files. The server creates a new file
in the series each time it starts or flushes the logs. The server
also creates a new binary log file automatically after the current
log's size reaches
max_binlog_size. A binary log
file may become larger than
max_binlog_size if you are using
large transactions because a transaction is written to the file in
one piece, never split between files.
To keep track of which binary log files have been used,
mysqld also creates a binary log index file
that contains the names of all used binary log files. By default,
this has the same basename as the binary log file, with the
extension '.index'. You can change the name of
the binary log index file with the
--log-bin-index[=
option. You should not manually edit this file while
mysqld is running; doing so would confuse
mysqld.
file_name]
The term “binary log file” generally denotes an individual numbered file containing database events. The term “binary log” collectively denotes the set of numbered binary log files plus the index file.
A client that has the SUPER
privilege can disable binary logging of its own statements by
using a SET sql_log_bin=0 statement. See
Section 5.1.4, “Server System Variables”.
By default, the server logs the length of the event as well as the
event itself and uses this to verify that the event was written
correctly. You can also cause the server to write checksums for
the events by setting the
binlog_checksum system variable.
When reading back from the binary log, the master uses the event
length by default, but can be made to use checksums if available
by enabling the
master_verify_checksum system
variable. The slave I/O thread also verifies events received from
the master. You can cause the slave SQL thread to use checksums if
available when reading from the relay log by enabling the
slave_sql_verify_checksum system
variable.
The format of the events recorded in the binary log is dependent on the binary logging format. Three format types are supported, row-based logging, statement-based logging and mixed-base logging. The binary logging format used depends on the MySQL version. For general descriptions of the logging formats, see Section 5.2.4.1, “Binary Logging Formats”. For detailed information about the format of the binary log, see MySQL Internals: The Binary Log.
The server evaluates the
--binlog-do-db and
--binlog-ignore-db options in the
same way as it does the
--replicate-do-db and
--replicate-ignore-db options. For
information about how this is done, see
Section 16.2.3.1, “Evaluation of Database-Level Replication and Binary Logging Options”.
A replication slave server by default does not write to its own
binary log any data modifications that are received from the
replication master. To log these modifications, start the slave
with the --log-slave-updates option
in addition to the --log-bin option
(see Section 16.1.4.3, “Replication Slave Options and Variables”). This is done
when a slave is also to act as a master to other slaves in chained
replication.
You can delete all binary log files with the
RESET MASTER statement, or a subset
of them with PURGE BINARY LOGS. See
Section 13.7.6.6, “RESET Syntax”, and Section 13.4.1.1, “PURGE BINARY LOGS Syntax”.
If you are using replication, you should not delete old binary log
files on the master until you are sure that no slave still needs
to use them. For example, if your slaves never run more than three
days behind, once a day you can execute mysqladmin
flush-logs on the master and then remove any logs that
are more than three days old. You can remove the files manually,
but it is preferable to use PURGE BINARY
LOGS, which also safely updates the binary log index
file for you (and which can take a date argument). See
Section 13.4.1.1, “PURGE BINARY LOGS Syntax”.
You can display the contents of binary log files with the mysqlbinlog utility. This can be useful when you want to reprocess statements in the log for a recovery operation. For example, you can update a MySQL server from the binary log as follows:
shell> mysqlbinlog log_file | mysql -h server_name
mysqlbinlog also can be used to display replication slave relay log file contents because they are written using the same format as binary log files. For more information on the mysqlbinlog utility and how to use it, see Section 4.6.8, “mysqlbinlog — Utility for Processing Binary Log Files”. For more information about the binary log and recovery operations, see Section 7.5, “Point-in-Time (Incremental) Recovery Using the Binary Log”.
Binary logging is done immediately after a statement or transaction completes but before any locks are released or any commit is done. This ensures that the log is logged in commit order.
Updates to nontransactional tables are stored in the binary log immediately after execution.
Within an uncommitted transaction, all updates
(UPDATE,
DELETE, or
INSERT) that change transactional
tables such as InnoDB tables are cached until a
COMMIT statement is received by the
server. At that point, mysqld writes the entire
transaction to the binary log before the
COMMIT is executed.
Modifications to nontransactional tables cannot be rolled back. If
a transaction that is rolled back includes modifications to
nontransactional tables, the entire transaction is logged with a
ROLLBACK
statement at the end to ensure that the modifications to those
tables are replicated.
When a thread that handles the transaction starts, it allocates a
buffer of binlog_cache_size to
buffer statements. If a statement is bigger than this, the thread
opens a temporary file to store the transaction. The temporary
file is deleted when the thread ends.
The Binlog_cache_use status
variable shows the number of transactions that used this buffer
(and possibly a temporary file) for storing statements. The
Binlog_cache_disk_use status
variable shows how many of those transactions actually had to use
a temporary file. These two variables can be used for tuning
binlog_cache_size to a large
enough value that avoids the use of temporary files.
The max_binlog_cache_size system
variable (default 4GB, which is also the maximum) can be used to
restrict the total size used to cache a multiple-statement
transaction. If a transaction is larger than this many bytes, it
fails and rolls back. The minimum value is 4096.
If you are using the binary log and row based logging, concurrent
inserts are converted to normal inserts for CREATE ...
SELECT or
INSERT ...
SELECT statements. This is done to ensure that you can
re-create an exact copy of your tables by applying the log during
a backup operation. If you are using statement-based logging, the
original statement is written to the log.
The binary log format has some known limitations that can affect recovery from backups. See Section 16.4.1, “Replication Features and Issues”.
Binary logging for stored programs is done as described in Section 18.7, “Binary Logging of Stored Programs”.
Note that the binary log format differs in MySQL 5.7 from previous versions of MySQL, due to enhancements in replication. See Section 16.4.2, “Replication Compatibility Between MySQL Versions”.
Writes to the binary log file and binary log index file are
handled in the same way as writes to MyISAM
tables. See Section C.5.4.3, “How MySQL Handles a Full Disk”.
By default, the binary log is not synchronized to disk at each
write. So if the operating system or machine (not only the MySQL
server) crashes, there is a chance that the last statements of the
binary log are lost. To prevent this, you can make the binary log
be synchronized to disk after every N
writes to the binary log, with the
sync_binlog system variable. See
Section 5.1.4, “Server System Variables”. 1 is the safest value
for sync_binlog, but also the
slowest. Even with sync_binlog
set to 1, there is still the chance of an inconsistency between
the table content and binary log content in case of a crash. For
example, if you are using InnoDB tables and the
MySQL server processes a COMMIT
statement, it writes the whole transaction to the binary log and
then commits this transaction into InnoDB. If
the server crashes between those two operations, the transaction
is rolled back by InnoDB at restart but still
exists in the binary log. To resolve this, you should set
--innodb_support_xa to 1. Although
this option is related to the support of XA transactions in
InnoDB, it also ensures that the binary log and InnoDB data files
are synchronized.
For this option to provide a greater degree of safety, the MySQL
server should also be configured to synchronize the binary log and
the InnoDB logs to disk before committing the
transaction. The InnoDB logs are synchronized
by default, and sync_binlog=1 can be used to
synchronize the binary log. The effect of this option is that at
restart after a crash, after doing a rollback of transactions, the
MySQL server cuts rolled back InnoDB
transactions from the binary log. This ensures that the binary log
reflects the exact data of InnoDB tables, and
so, that the slave remains in synchrony with the master (not
receiving a statement which has been rolled back).
If the MySQL server discovers at crash recovery that the binary
log is shorter than it should have been, it lacks at least one
successfully committed InnoDB transaction. This
should not happen if sync_binlog=1 and the
disk/file system do an actual sync when they are requested to
(some do not), so the server prints an error message The
binary log . In this case, this binary log is not
correct and replication should be restarted from a fresh snapshot
of the master's data.
file_name is shorter than
its expected size
The session values of the following system variables are written to the binary log and honored by the replication slave when parsing the binary log:
sql_mode(except that theNO_DIR_IN_CREATEmode is not replicated; see Section 16.4.1.33, “Replication and Variables”)
The server uses several logging formats to record information in the binary log. The exact format employed depends on the version of MySQL being used. There are three logging formats:
Replication capabilities in MySQL originally were based on propagation of SQL statements from master to slave. This is called statement-based logging. You can cause this format to be used by starting the server with
--binlog-format=STATEMENT.In row-based logging, the master writes events to the binary log that indicate how individual table rows are affected. You can cause the server to use row-based logging by starting it with
--binlog-format=ROW.A third option is also available: mixed logging. With mixed logging, statement-based logging is used by default, but the logging mode switches automatically to row-based in certain cases as described below. You can cause MySQL to use mixed logging explicitly by starting mysqld with the option
--binlog-format=MIXED.
In MySQL 5.7, the default binary logging format is
STATEMENT.
The logging format can also be set or limited by the storage engine being used. This helps to eliminate issues when replicating certain statements between a master and slave which are using different storage engines.
With statement-based replication, there may be issues with replicating nondeterministic statements. In deciding whether or not a given statement is safe for statement-based replication, MySQL determines whether it can guarantee that the statement can be replicated using statement-based logging. If MySQL cannot make this guarantee, it marks the statement as potentially unreliable and issues the warning, Statement may not be safe to log in statement format.
You can avoid these issues by using MySQL's row-based replication instead.
You can select the binary logging format explicitly by starting
the MySQL server with
--binlog-format=.
The supported values for typetype are:
STATEMENTcauses logging to be statement based.ROWcauses logging to be row based.MIXEDcauses logging to use mixed format.
In MySQL 5.7, the default binary logging format is
STATEMENT.
The logging format also can be switched at runtime. To specify
the format globally for all clients, set the global value of the
binlog_format system variable:
mysql>SET GLOBAL binlog_format = 'STATEMENT';mysql>SET GLOBAL binlog_format = 'ROW';mysql>SET GLOBAL binlog_format = 'MIXED';
An individual client can control the logging format for its own
statements by setting the session value of
binlog_format:
mysql>SET SESSION binlog_format = 'STATEMENT';mysql>SET SESSION binlog_format = 'ROW';mysql>SET SESSION binlog_format = 'MIXED';
Each MySQL Server can set its own and only its own binary
logging format (true whether
binlog_format is set with
global or session scope). This means that changing the logging
format on a replication master does not cause a slave to
change its logging format to match. (When using
STATEMENT mode, the
binlog_format system variable
is not replicated; when using MIXED or
ROW logging mode, it is replicated but is
ignored by the slave.) Changing the binary logging format on
the master while replication is ongoing, or without also
changing it on the slave can thus cause unexpected results, or
even cause replication to fail altogether.
To change the global or session
binlog_format value, you must
have the SUPER privilege.
In addition to switching the logging format manually, a slave
server may switch the format automatically.
This happens when the server is running in either
STATEMENT or MIXED format
and encounters an event in the binary log that is written in
ROW logging format. In that case, the slave
switches to row-based replication temporarily for that event,
and switches back to the previous format afterward.
There are several reasons why a client might want to set binary logging on a per-session basis:
A session that makes many small changes to the database might want to use row-based logging.
A session that performs updates that match many rows in the
WHEREclause might want to use statement-based logging because it will be more efficient to log a few statements than many rows.Some statements require a lot of execution time on the master, but result in just a few rows being modified. It might therefore be beneficial to replicate them using row-based logging.
There are exceptions when you cannot switch the replication format at runtime:
From within a stored function or a trigger
If the
NDBCLUSTERstorage engine is enabledIf the session is currently in row-based replication mode and has open temporary tables
Trying to switch the format in any of these cases results in an error.
If you are using InnoDB tables and
the transaction isolation level is READ
COMMITTED or READ
UNCOMMITTED, only row-based logging can be used. It is
possible to change the logging format to
STATEMENT, but doing so at runtime leads very
rapidly to errors because InnoDB can no
longer perform inserts.
Switching the replication format at runtime is not recommended
when any temporary tables exist, because temporary tables are
logged only when using statement-based replication, whereas with
row-based replication they are not logged. With mixed
replication, temporary tables are usually logged; exceptions
happen with user-defined functions (UDFs) and with the
UUID() function.
With the binary log format set to ROW, many
changes are written to the binary log using the row-based
format. Some changes, however, still use the statement-based
format. Examples include all DDL (data definition language)
statements such as CREATE TABLE,
ALTER TABLE, or
DROP TABLE.
The --binlog-row-event-max-size
option is available for servers that are capable of row-based
replication. Rows are stored into the binary log in chunks
having a size in bytes not exceeding the value of this option.
The value must be a multiple of 256. The default value is 1024.
When using statement-based logging for replication, it is possible for the data on the master and slave to become different if a statement is designed in such a way that the data modification is nondeterministic; that is, it is left to the will of the query optimizer. In general, this is not a good practice even outside of replication. For a detailed explanation of this issue, see Section C.5.8, “Known Issues in MySQL”.
For information about logs kept by replication slaves, see Section 16.2.2, “Replication Relay and Status Logs”.
When running in MIXED logging format, the
server automatically switches from statement-based to row-based
logging under the following conditions:
When a function contains
UUID().When one or more tables with
AUTO_INCREMENTcolumns are updated and a trigger or stored function is invoked. Like all other unsafe statements, this generates a warning ifbinlog_format = STATEMENT.For more information, see Section 16.4.1.1, “Replication and
AUTO_INCREMENT”.When the body of a view requires row-based replication, the statement creating the view also uses it. For example, this occurs when the statement creating a view uses the
UUID()function.When a call to a UDF is involved.
If a statement is logged by row and the session that executed the statement has any temporary tables, logging by row is used for all subsequent statements (except for those accessing temporary tables) until all temporary tables in use by that session are dropped.
This is true whether or not any temporary tables are actually logged.
Temporary tables cannot be logged using row-based format; thus, once row-based logging is used, all subsequent statements using that table are unsafe. The server approximates this condition by treating all statements executed during the session as unsafe until the session no longer holds any temporary tables.
When
FOUND_ROWS()orROW_COUNT()is used. (Bug #12092, Bug #30244)When
USER(),CURRENT_USER(), orCURRENT_USERis used. (Bug #28086)When a statement refers to one or more system variables. (Bug #31168)
Exception. The following system variables, when used with session scope (only), do not cause the logging format to switch:
For information about determining system variable scope, see Section 5.1.5, “Using System Variables”.
For information about how replication treats
sql_mode, see Section 16.4.1.33, “Replication and Variables”.When one of the tables involved is a log table in the
mysqldatabase.When the
LOAD_FILE()function is used. (Bug #39701)
A warning is generated if you try to execute a statement using
statement-based logging that should be written using row-based
logging. The warning is shown both in the client (in the
output of SHOW WARNINGS) and
through the mysqld error log. A warning is
added to the SHOW WARNINGS
table each time such a statement is executed. However, only
the first statement that generated the warning for each client
session is written to the error log to prevent flooding the
log.
In addition to the decisions above, individual engines can also determine the logging format used when information in a table is updated. The logging capabilities of an individual engine can be defined as follows:
If an engine supports row-based logging, the engine is said to be row-logging capable.
If an engine supports statement-based logging, the engine is said to be statement-logging capable.
A given storage engine can support either or both logging formats. The following table lists the formats supported by each engine.
| Storage Engine | Row Logging Supported | Statement Logging Supported |
|---|---|---|
ARCHIVE | Yes | Yes |
BLACKHOLE | Yes | Yes |
CSV | Yes | Yes |
EXAMPLE | Yes | No |
FEDERATED | Yes | Yes |
HEAP | Yes | Yes |
InnoDB | Yes | Yes when the transaction isolation level is
REPEATABLE READ or
SERIALIZABLE; No
otherwise. |
MyISAM | Yes | Yes |
MERGE | Yes | Yes |
NDBCLUSTER | Yes | No |
In MySQL 5.7, whether a statement is to be logged
and the logging mode to be used is determined according to the
type of statement (safe, unsafe, or binary injected), the binary
logging format (STATEMENT,
ROW, or MIXED), and the
logging capabilities of the storage engine (statement capable,
row capable, both, or neither). Statements may be logged with or
without a warning; failed statements are not logged, but
generate errors in the log. This is shown in the following
decision table, where SLC
stands for “statement-logging capable” and
RLC stands for
“row-logging capable”.
| Condition | Action | ||||
|---|---|---|---|---|---|
| Type | binlog_format | SLC | RLC | Error / Warning | Logged as |
| * | * | No | No | Error: Cannot execute statement: Binary logging is impossible since at least one engine is involved that is both row-incapable and statement-incapable. | - |
| Safe | STATEMENT | Yes | No | - | STATEMENT |
| Safe | MIXED | Yes | No | - | STATEMENT |
| Safe | ROW | Yes | No | Error: Cannot execute statement: Binary logging
is impossible since BINLOG_FORMAT =
ROW and at least one table uses a storage
engine that is not capable of row-based logging. | - |
| Unsafe | STATEMENT | Yes | No | Warning: Unsafe statement binlogged in statement
format, since BINLOG_FORMAT =
STATEMENT | STATEMENT |
| Unsafe | MIXED | Yes | No | Error: Cannot execute statement: Binary logging
of an unsafe statement is impossible when the storage
engine is limited to statement-based logging, even if
BINLOG_FORMAT = MIXED. | - |
| Unsafe | ROW | Yes | No | Error: Cannot execute statement: Binary logging
is impossible since BINLOG_FORMAT =
ROW and at least one table uses a storage
engine that is not capable of row-based logging. | - |
| Row Injection | STATEMENT | Yes | No | Error: Cannot execute row injection: Binary logging is not possible since at least one table uses a storage engine that is not capable of row-based logging. | - |
| Row Injection | MIXED | Yes | No | Error: Cannot execute row injection: Binary logging is not possible since at least one table uses a storage engine that is not capable of row-based logging. | - |
| Row Injection | ROW | Yes | No | Error: Cannot execute row injection: Binary logging is not possible since at least one table uses a storage engine that is not capable of row-based logging. | - |
| Safe | STATEMENT | No | Yes | Error: Cannot execute statement: Binary logging
is impossible since BINLOG_FORMAT =
STATEMENT and at least one table uses a
storage engine that is not capable of statement-based
logging. | - |
| Safe | MIXED | No | Yes | - | ROW |
| Safe | ROW | No | Yes | - | ROW |
| Unsafe | STATEMENT | No | Yes | Error: Cannot execute statement: Binary logging
is impossible since BINLOG_FORMAT =
STATEMENT and at least one table uses a
storage engine that is not capable of statement-based
logging. | - |
| Unsafe | MIXED | No | Yes | - | ROW |
| Unsafe | ROW | No | Yes | - | ROW |
| Row Injection | STATEMENT | No | Yes | Error: Cannot execute row injection: Binary
logging is not possible since BINLOG_FORMAT =
STATEMENT. | - |
| Row Injection | MIXED | No | Yes | - | ROW |
| Row Injection | ROW | No | Yes | - | ROW |
| Safe | STATEMENT | Yes | Yes | - | STATEMENT |
| Safe | MIXED | Yes | Yes | - | ROW |
| Safe | ROW | Yes | Yes | - | ROW |
| Unsafe | STATEMENT | Yes | Yes | Warning: Unsafe statement binlogged in statement
format since BINLOG_FORMAT =
STATEMENT. | STATEMENT |
| Unsafe | MIXED | Yes | Yes | - | ROW |
| Unsafe | ROW | Yes | Yes | - | ROW |
| Row Injection | STATEMENT | Yes | Yes | Error: Cannot execute row injection: Binary
logging is not possible because BINLOG_FORMAT =
STATEMENT. | - |
| Row Injection | MIXED | Yes | Yes | - | ROW |
| Row Injection | ROW | Yes | Yes | - | ROW |
When a warning is produced by the determination, a standard
MySQL warning is produced (and is available using
SHOW WARNINGS). The information
is also written to the mysqld error log. Only
one error for each error instance per client connection is
logged to prevent flooding the log. The log message includes the
SQL statement that was attempted.
If a slave server was started with
--log-warnings enabled, the slave
prints messages to the error log to provide information about
its status, such as the binary log and relay log coordinates
where it starts its job, when it is switching to another relay
log, when it reconnects after a disconnect, and so forth.
The contents of the grant tables in the mysql
database can be modified directly (for example, with
INSERT or
DELETE) or indirectly (for
example, with GRANT or
CREATE USER). Statements that
affect mysql database tables are written to
the binary log using the following rules:
Data manipulation statements that change data in
mysqldatabase tables directly are logged according to the setting of thebinlog_formatsystem variable. This pertains to statements such asINSERT,UPDATE,DELETE,REPLACE,DO,LOAD DATA INFILE,SELECT, andTRUNCATE TABLE.Statements that change the
mysqldatabase indirectly are logged as statements regardless of the value ofbinlog_format. This pertains to statements such asGRANT,REVOKE,SET PASSWORD,RENAME USER,CREATE(all forms exceptCREATE TABLE ... SELECT),ALTER(all forms), andDROP(all forms).
CREATE TABLE ...
SELECT is a combination of data definition and data
manipulation. The CREATE TABLE
part is logged using statement format and the
SELECT part is logged according
to the value of binlog_format.
The slow query log consists of SQL statements that took more than
long_query_time seconds to
execute and required at least
min_examined_row_limit rows to be
examined. The minimum and default values of
long_query_time are 0 and 10,
respectively. The value can be specified to a resolution of
microseconds. For logging to a file, times are written including
the microseconds part. For logging to tables, only integer times
are written; the microseconds part is ignored.
By default, administrative statements are not logged, nor are
queries that do not use indexes for lookups. This behavior can be
changed using
--log-slow-admin-statements and
log_queries_not_using_indexes, as
described later.
The time to acquire the initial locks is not counted as execution time. mysqld writes a statement to the slow query log after it has been executed and after all locks have been released, so log order might differ from execution order.
By default, the slow query log is disabled. To specify the initial
slow query log state explicitly, use
--slow_query_log[={0|1}]. With no
argument or an argument of 1,
--slow_query_log enables the log.
With an argument of 0, this option disables the log. To specify a
log file name, use
--slow_query_log_file=.
To specify the log destination, use
file_name--log-output (as described in
Section 5.2.1, “Selecting General Query and Slow Query Log Output Destinations”).
If you specify no name for the slow query log file, the default
name is
host_name-slow.log
To disable or enable the slow query log or change the log file
name at runtime, use the global
slow_query_log and
slow_query_log_file system
variables. Set slow_query_log to
0 (or OFF) to disable the log or to 1 (or
ON) to enable it. Set
slow_query_log_file to specify
the name of the log file. If a log file already is open, it is
closed and the new file is opened.
When the slow query log is enabled, the server writes output to
any destinations specified by the
--log-output option or
log_output system variable. If
you enable the log, the server opens the log file and writes
startup messages to it. However, further logging of queries to the
file does not occur unless the FILE log
destination is selected. If the destination is
NONE, the server writes no queries even if the
slow query log is enabled. Setting the log file name has no effect
on logging if the log destination value does not contain
FILE.
The server writes less information to the slow query log (and
binary log) if you use the
--log-short-format option.
To include slow administrative statements in the statements
written to the slow query log, use the
--log-slow-admin-statements server
option. Administrative statements include
ALTER TABLE,
ANALYZE TABLE,
CHECK TABLE,
CREATE INDEX,
DROP INDEX,
OPTIMIZE TABLE, and
REPAIR TABLE.
To include queries that do not use indexes for row lookups in the
statements written to the slow query log, enable the
log_queries_not_using_indexes
system variable. When such queries are logged, the slow query log
may grow quickly. It is possible to put a rate limit on these
queries by setting the
log_throttle_queries_not_using_indexes
system variable. By default, this variable is 0, which means there
is no limit. Positive values impose a per-minute limit on logging
of queries that do not use indexes. The first such query opens a
60-second window within which the server logs queries up to the
given limit, then suppresses additional queries. If there are
suppressed queries when the window ends, the server logs a summary
that indicates how many there were and the aggregate time spent in
them. The next 60-second window begins when the server logs the
next query that does not use indexes.
The server uses the controlling parameters in the following order to determine whether to write a query to the slow query log:
The query must either not be an administrative statement, or
--log-slow-admin-statementsmust have been specified.The query must have taken at least
long_query_timeseconds, orlog_queries_not_using_indexesmust be enabled and the query used no indexes for row lookups.The query must have examined at least
min_examined_row_limitrows.The query must not be suppressed according to the
log_throttle_queries_not_using_indexessetting.
The server does not write queries handled by the query cache to the slow query log, nor queries that would not benefit from the presence of an index because the table has zero rows or one row.
By default, a replication slave does not write replicated queries
to the slow query log. To change this, use the
--log-slow-slave-statements server
option.
Passwords in statements written to the slow query log are rewritten by the server not to occur literally in plain text. See also Section 6.1.2.3, “Passwords and Logging”.
The slow query log can be used to find queries that take a long time to execute and are therefore candidates for optimization. However, examining a long slow query log can become a difficult task. To make this easier, you can process a slow query log file using the mysqldumpslow command to summarize the queries that appear in the log. See Section 4.6.9, “mysqldumpslow — Summarize Slow Query Log Files”.
As described in Section 5.2, “MySQL Server Logs”, MySQL Server can create several different log files to help you see what activity is taking place. However, you must clean up these files regularly to ensure that the logs do not take up too much disk space.
When using MySQL with logging enabled, you may want to back up and remove old log files from time to time and tell MySQL to start logging to new files. See Section 7.2, “Database Backup Methods”.
On a Linux (Red Hat) installation, you can use the
mysql-log-rotate script for this. If you
installed MySQL from an RPM distribution, this script should have
been installed automatically. Be careful with this script if you
are using the binary log for replication. You should not remove
binary logs until you are certain that their contents have been
processed by all slaves.
On other systems, you must install a short script yourself that you start from cron (or its equivalent) for handling log files.
For the binary log, you can set the
expire_logs_days system variable
to expire binary log files automatically after a given number of
days (see Section 5.1.4, “Server System Variables”). If you are
using replication, you should set the variable no lower than the
maximum number of days your slaves might lag behind the master. To
remove binary logs on demand, use the PURGE
BINARY LOGS statement (see
Section 13.4.1.1, “PURGE BINARY LOGS Syntax”).
You can force MySQL to start using new log files by flushing the
logs. Log flushing occurs when you issue a
FLUSH LOGS
statement or execute a mysqladmin flush-logs,
mysqladmin refresh, mysqldump
--flush-logs, or mysqldump
--master-data command. See Section 13.7.6.3, “FLUSH Syntax”,
Section 4.5.2, “mysqladmin — Client for Administering a MySQL Server”, and Section 4.5.4, “mysqldump — A Database Backup Program”. In
addition, the binary log is flushed when its size reaches the
value of the max_binlog_size
system variable.
FLUSH LOGS
supports optional modifiers to enable selective flushing of
individual logs (for example,
FLUSH BINARY
LOGS).
A log-flushing operation does the following:
If general query logging or slow query logging to a log file is enabled, the server closes and reopens the general query log file or slow query log file.
If binary logging is enabled, the server closes the current binary log file and opens a new log file with the next sequence number.
If the server was started with the
--log-erroroption to cause the error log to be written to a file, the server closes and reopens the log file.
The server creates a new binary log file when you flush the logs.
However, it just closes and reopens the general and slow query log
files. To cause new files to be created on Unix, rename the
current log files before flushing them. At flush time, the server
opens new log files with the original names. For example, if the
general and slow query log files are named
mysql.log and
mysql-slow.log, you can use a series of
commands like this:
shell>cdshell>mysql-data-directorymv mysql.log mysql.oldshell>mv mysql-slow.log mysql-slow.oldshell>mysqladmin flush-logs
On Windows, use rename rather than mv.
At this point, you can make a backup of
mysql.old and
mysql-slow.old and then remove them from
disk.
A similar strategy can be used to back up the error log file, if there is one.
You can rename the general query log or slow query log at runtime by disabling the log:
SET GLOBAL general_log = 'OFF'; SET GLOBAL slow_query_log = 'OFF';
With the logs disabled, rename the log files externally; for example, from the command line. Then enable the logs again:
SET GLOBAL general_log = 'ON'; SET GLOBAL slow_query_log = 'ON';
This method works on any platform and does not require a server restart.
As a DBA, you must manage disk I/O to keep the I/O subsystem from
becoming saturated, and manage disk space to avoid filling up
storage devices. The ACID design
model requires a certain amount of I/O that might seem redundant,
but helps to ensure data reliability. Within these constraints,
InnoDB tries to optimize the database work and
the organization of disk files to minimize the amount of disk I/O.
Sometimes, I/O is postponed until the database is not busy, or until
everything needs to be brought to a consistent state, such as during
a database restart after a fast
shutdown.
This section discusses the main considerations for I/O and disk
space with the default kind of MySQL tables (also known as
InnoDB tables):
Controlling the amount of background I/O used to improve query performance.
Enabling or disabling features that provide extra durability at the expense of additional I/O.
Organizing tables into many small files, a few larger files, or a combination of both.
Balancing the size of redo log files against the I/O activity that occurs when the log files become full.
How to reorganize a table for optimal query performance.
InnoDB uses asynchronous disk I/O where
possible, by creating a number of threads to handle I/O
operations, while permitting other database operations to proceed
while the I/O is still in progress. On Linux and Windows
platforms, InnoDB uses the available OS and library functions to
perform “native” asynchronous I/O. On other
platforms, InnoDB still uses I/O threads, but the threads may
actually wait for I/O requests to complete; this technique is
known as “simulated” asynchronous I/O.
Read-Ahead
If InnoDB can determine there is a high probability that data might be needed soon, it performs read-ahead operations to bring that data into the buffer pool so that it is available in memory. Making a few large read requests for contiguous data can be more efficient than making several small, spread-out requests. There are two read-ahead heuristics in InnoDB:
In sequential read-ahead, if
InnoDBnotices that the access pattern to a segment in the tablespace is sequential, it posts in advance a batch of reads of database pages to the I/O system.In random read-ahead, if
InnoDBnotices that some area in a tablespace seems to be in the process of being fully read into the buffer pool, it posts the remaining reads to the I/O system.
Doublewrite Buffer
InnoDB uses a novel file flush technique
involving a structure called the
doublewrite buffer.
It adds safety to recovery following an operating system crash or
a power outage, and improves performance on most varieties of Unix
by reducing the need for fsync() operations.
Before writing pages to a data file, InnoDB
first writes them to a contiguous tablespace area called the
doublewrite buffer. Only after the write and the flush to the
doublewrite buffer has completed does InnoDB
write the pages to their proper positions in the data file. If the
operating system crashes in the middle of a page write (causing a
torn page condition),
InnoDB can later find a good copy of the page
from the doublewrite buffer during recovery.
The data files that you define in the configuration file form the
InnoDB
system tablespace.
The files are logically concatenated to form the tablespace. There
is no striping in use. Currently, you cannot define where within
the tablespace your tables are allocated. In a newly created
tablespace, InnoDB allocates space starting
from the first data file.
To avoid the issues that come with storing all tables and indexes
inside the system tablespace, you can turn on the
innodb_file_per_table configuration option,
which stores each newly created table in a separate tablespace
file (with extension .ibd). For tables stored
this way, there is less fragmentation within the disk file, and
when the table is truncated, the space is returned to the
operating system rather than still being reserved by InnoDB within
the system tablespace.
Pages, Extents, Segments, and Tablespaces
Each tablespace consists of database
pages. Every tablespace in a
MySQL instance has the same page
size. By default, all tablespaces have a page size of 16KB;
you can reduce the page size to 8KB or 4KB by specifying the
innodb_page_size option when you
create the MySQL instance.
The pages are grouped into
extents of size 1MB (64
consecutive 16KB pages, or 128 8KB pages, or 256 4KB pages). The
“files” inside a tablespace are called
segments in
InnoDB. (These segments are different from the
rollback segment,
which actually contains many tablespace segments.)
When a segment grows inside the tablespace,
InnoDB allocates the first 32 pages to it one
at a time. After that, InnoDB starts to
allocate whole extents to the segment. InnoDB
can add up to 4 extents at a time to a large segment to ensure
good sequentiality of data.
Two segments are allocated for each index in
InnoDB. One is for nonleaf nodes of the B-tree,
the other is for the leaf nodes. Keeping the leaf nodes contiguous
on disk enables better sequential I/O operations, because these
leaf nodes contain the actual table data.
Some pages in the tablespace contain bitmaps of other pages, and
therefore a few extents in an InnoDB tablespace
cannot be allocated to segments as a whole, but only as individual
pages.
When you ask for available free space in the tablespace by issuing
a SHOW TABLE STATUS statement,
InnoDB reports the extents that are definitely
free in the tablespace. InnoDB always reserves
some extents for cleanup and other internal purposes; these
reserved extents are not included in the free space.
When you delete data from a table, InnoDB
contracts the corresponding B-tree indexes. Whether the freed
space becomes available for other users depends on whether the
pattern of deletes frees individual pages or extents to the
tablespace. Dropping a table or deleting all rows from it is
guaranteed to release the space to other users, but remember that
deleted rows are physically removed only by the
purge operation, which happens
automatically some time after they are no longer needed for
transaction rollbacks or consistent reads. (See
Section 14.2.3.11, “InnoDB Multi-Versioning”.)
To see information about the tablespace, use the Tablespace
Monitor. See Section 14.2.4.4, “SHOW ENGINE INNODB
STATUS and the InnoDB Monitors”.
How Pages Relate to Table Rows
The maximum row length, except for variable-length columns
(VARBINARY,
VARCHAR,
BLOB and
TEXT), is slightly less than half
of a database page. That is, the maximum row length is about 8000
bytes. LONGBLOB and
LONGTEXT columns
must be less than 4GB, and the total row length, including
BLOB and
TEXT columns, must be less than
4GB.
If a row is less than half a page long, all of it is stored
locally within the page. If it exceeds half a page,
variable-length columns are chosen for external off-page storage
until the row fits within half a page. For a column chosen for
off-page storage, InnoDB stores the first 768
bytes locally in the row, and the rest externally into overflow
pages. Each such column has its own list of overflow pages. The
768-byte prefix is accompanied by a 20-byte value that stores the
true length of the column and points into the overflow list where
the rest of the value is stored.
Making your log files very large may reduce disk I/O during checkpointing. It often makes sense to set the total size of the log files as large as the buffer pool or even larger. Although in the past large log files could make crash recovery take excessive time, starting with MySQL 5.5, performance enhancements to crash recovery make it possible to use large log files with fast startup after a crash. (Strictly speaking, this performance improvement is available for MySQL 5.1 with the InnoDB Plugin 1.0.7 and higher. It is with MySQL 5.5 that this improvement is available in the default InnoDB storage engine.)
How Checkpoint Processing Works
InnoDB implements a
checkpoint mechanism known
as fuzzy
checkpointing. InnoDB flushes modified
database pages from the buffer pool in small batches. There is no
need to flush the buffer pool in one single batch, which would
disrupt processing of user SQL statements during the checkpointing
process.
During crash recovery,
InnoDB looks for a checkpoint label written to
the log files. It knows that all modifications to the database
before the label are present in the disk image of the database.
Then InnoDB scans the log files forward from
the checkpoint, applying the logged modifications to the database.
Random insertions into or deletions from a secondary index can cause the index to become fragmented. Fragmentation means that the physical ordering of the index pages on the disk is not close to the index ordering of the records on the pages, or that there are many unused pages in the 64-page blocks that were allocated to the index.
One symptom of fragmentation is that a table takes more space than
it “should” take. How much that is exactly, is
difficult to determine. All InnoDB data and
indexes are stored in B-trees, and their
fill factor may vary from
50% to 100%. Another symptom of fragmentation is that a table scan
such as this takes more time than it “should” take:
SELECT COUNT(*) FROM t WHERE non_indexed_column <> 12345;
The preceding query requires MySQL to perform a full table scan, the slowest type of query for a large table.
To speed up index scans, you can periodically perform a
“null” ALTER TABLE
operation, which causes MySQL to rebuild the table:
ALTER TABLE tbl_name ENGINE=INNODB
Another way to perform a defragmentation operation is to use mysqldump to dump the table to a text file, drop the table, and reload it from the dump file.
If the insertions into an index are always ascending and records
are deleted only from the end, the InnoDB
filespace management algorithm guarantees that fragmentation in
the index does not occur.
- 5.4.1. Managing InnoDB Tablespaces
- 5.4.2. Grouping DML Operations with Transactions
- 5.4.3. Converting Tables from
MyISAMtoInnoDB - 5.4.4.
AUTO_INCREMENTHandling inInnoDB - 5.4.5.
InnoDBandFOREIGN KEYConstraints - 5.4.6. Working with
InnoDBCompressed Tables - 5.4.7.
InnoDBFile-Format Management - 5.4.8. How
InnoDBStores Variable-Length Columns
To create an InnoDB table, use the
CREATE TABLE statement without any
special clauses. Formerly, you needed the
ENGINE=InnoDB clause, but not anymore now that
InnoDB is the default storage engine. (You might
still use that clause if you plan to use
mysqldump or replication to replay the
CREATE TABLE statement on a server
running MySQL 5.1 or earlier, where the default storage engine is
MyISAM.)
-- Default storage engine = InnoDB. CREATE TABLE t1 (a INT, b CHAR (20), PRIMARY KEY (a)); -- Backwards-compatible with older MySQL. CREATE TABLE t2 (a INT, b CHAR (20), PRIMARY KEY (a)) ENGINE=InnoDB;
Depending on the file-per-table setting, InnoDB
creates each table and associated primary key index either in the
system tablespace, or
in a separate tablespace (represented by a
.ibd file) for each table.
MySQL creates t1.frm and
t2.frm files in the test
directory under the MySQL database directory. Internally,
InnoDB adds an entry for the table to its own
data dictionary. The entry includes the database name. For example,
if test is the database in which the
t1 table is created, the entry is for
'test/t1'. This means you can create a table of
the same name t1 in some other database, and the
table names do not collide inside InnoDB.
To see the properties of these tables, issue a
SHOW TABLE STATUS statement:
SHOW TABLE STATUS FROM test LIKE 't%';
In the status output, you see the
row format property of these
first tables is Compact. Although that setting is
fine for basic experimentation, to take advantage of the most
powerful InnoDB performance features, you will
quickly graduate to using other row formats such as
Dynamic
and
Compressed.
Using those values requires a little bit of setup first:
set global innodb_file_per_table=1; set global innodb_file_format=barracuda; CREATE TABLE t3 (a INT, b CHAR (20), PRIMARY KEY (a)) row_format=dynamic; CREATE TABLE t4 (a INT, b CHAR (20), PRIMARY KEY (a)) row_format=compressed;
Always set up a primary key
for each InnoDB table, specifying the column or
columns that:
Are referenced by the most important queries.
Are never left blank.
Never have duplicate values.
Rarely if ever change value once inserted.
For example, in a table containing information about people, you
would not create a primary key on (firstname,
lastname) because more than one person can have the same
name, some people have blank last names, and sometimes people change
their names. With so many constraints, often there is not an obvious
set of columns to use as a primary key, so you create a new column
with a numeric ID to serve as all or part of the primary key. You
can declare an
auto-increment column so
that ascending values are filled in automatically as rows are
inserted:
-- The value of ID can act like a pointer between related items in different tables. CREATE TABLE t5 (id INT AUTO_INCREMENT, b CHAR (20), PRIMARY KEY (id)); -- The primary key can consist of more than one column. Any autoinc column must come first. CREATE TABLE t6 (id INT AUTO_INCREMENT, a INT, b CHAR (20), PRIMARY KEY (id,a));
Although the table works correctly without you
defining a primary key, the primary key is involved with many
aspects of performance and is a crucial design aspect for any large
or frequently used table. Make a habit of always specifying one in
the CREATE TABLE statement. (If you
create the table, load data, and then do ALTER
TABLE to add a primary key later, that operation is much
slower than defining the primary key when creating the table.)
Historically, all InnoDB tables and indexes
were stored in the system
tablespace. This monolithic approach was targeted at
machines dedicated entirely to database processing, with carefully
planned data growth, where any disk storage allocated to MySQL
would never be needed for other purposes.
InnoDB's
file-per-table mode is
a more flexible alternative, where you store each
InnoDB table and its indexes in a separate
file. Each such
.ibd file
represents a separate
tablespace. This mode is
controlled by the
innodb_file_per_table
configuration option, and is the default in MySQL 5.6.6 and
higher.
Advantages of File-Per-Table Mode
You can reclaim disk space when truncating or dropping a table. For tables created when file-per-table mode is turned off, truncating or dropping them creates free space internally in the ibdata files. That free space can only be used for new
InnoDBdata.The
TRUNCATE TABLEoperation is faster when run on individual.ibdfiles.You can store specific tables on separate storage devices, for I/O optimization, space management, or backup purposes. In previous releases, you had to move entire database directories to other drives and create symbolic links in the MySQL data directory, as described in Section 8.11.3.1, “Using Symbolic Links”. In MySQL 5.6 and higher, you can specify the location of each table using the syntax
CREATE TABLE ... DATA DIRECTORY =, as explained in Section 5.4.1.2, “Specifying the Location of a Tablespace”.absolute_path_to_directoryYou can run
OPTIMIZE TABLEto compact or recreate a tablespace. When you run anOPTIMIZE TABLE,InnoDBwill create a new.ibdfile with a temporary name, using only the space required to store actual data. When the optimization is complete,InnoDBremoves the old.ibdfile and replaces it with the new.ibdfile. If the previous.ibdfile had grown significantly but actual data only accounted for a portion of its size, runningOPTIMIZE TABLEallows you to reclaim the unused space.You can move individual
InnoDBtables rather than entire databases.You can copy individual
InnoDBtables from one MySQL instance to another (known as the transportable tablespace feature).You can enable compression for table and index data, using the compressed row format.
You can enable more efficient storage for tables with large BLOB or text columns using the dynamic row format.
Using
innodb_file_per_tablemay improve chances for a successful recovery and save time if a corruption occurs, a server cannot be restarted, or backup and binary logs are unavailable.You can back up or restore a single table quickly, without interrupting the use of other
InnoDBtables, using the MySQL Enterprise Backup product. See Backing Up and Restoring a Single.ibdFile for the procedure and restrictions.File-per-table mode allows you to excluded tables from a backup. This is beneficial if you have tables that require backup less frequently or on a different schedule.
File-per-table mode is convenient for per-table status reporting when copying or backing up tables.
File-per-table mode allows you to monitor table size at a file system level, without accessing MySQL.
Common Linux file systems do not permit concurrent writes to a single file when
innodb_flush_methodis set toO_DIRECT. As a result, there are possible performance improvements when usinginnodb_file_per_tablein conjunction withinnodb_flush_method.If
innodb_file_per_tableis disabled, there is one shared tablespace (the system tablespace) for tables, the data dictionary, and undo logs. This single tablespace has a 64TB size limit. Ifinnodb_file_per_tableis enabled, each table has its own tablespace, each with a 64TB size limit. See Section D.10.3, “Limits on Table Size” for related information.
Potential Disadvantages of File-Per-Table Mode
With
innodb_file_per_table, each table may have unused table space, which can only be utilized by rows of the same table. This could lead to more rather than less wasted table space if not properly managed.fsyncoperations must run on each open table rather than on a single file. Because there is a separatefsyncoperation for each file, write operations on multiple tables cannot be combined into a single I/O operation. This may requireInnoDBto perform a higher total number offsyncoperations.mysqld must keep 1 open file handle per table, which may impact performance if you have numerous tables.
More file descriptors are used.
innodb_file_per_tableis on by default in MySQL 5.6.6 and higher. You may want to consider disabling it if backward compatibility with MySQL 5.5 or 5.1 is a concern. Disablinginnodb_file_per_tablepreventsALTER TABLEfrom movingInnoDBtables from the system tablespace to individual.ibdfiles.If many tables are growing there is potential for more fragmentation which can impede
DROP TABLEand table scan performance. However, when fragmentation is managed, having files in their own tablespace can improve performance.The buffer pool is scanned when dropping a per-table tablespace, which can take several seconds for buffer pools that are tens of gigabytes in size. The scan is performed with a broad internal lock, which may delay other operations. Tables in the shared tablespace are not affected.
The
innodb_autoextend_incrementvariable, which defines increment size (in MB) for extending the size of an auto-extending shared tablespace file when it becomes full, does not apply to per-table tablespace files. Per-table tablespace files are auto-extending regardless of the value ofinnodb_autoextend_increment. The initial extensions are by small amounts, after which extensions occur in increments of 4MB.
To make file-per-table mode the default for a MySQL server,
start the server with the
--innodb_file_per_table
command-line option, or add this line to the
[mysqld] section of
my.cnf:
[mysqld] innodb_file_per_table
You can also issue the command while the server is running:
SET GLOBAL innodb_file_per_table=1;
With file-per-table mode enabled, InnoDB
stores each newly created table in its own
tbl_name.ibdMyISAM storage engine, with its separate
tbl_name.MYDtbl_name.MYIInnoDB stores the
data and the indexes together in a single
.ibd file. The
tbl_name.frm
If you remove
innodb_file_per_table from your
startup options and restart the server, or turn it off with the
SET GLOBAL command, InnoDB
creates any new tables inside the system tablespace.
You can always read and write any InnoDB
tables, regardless of the file-per-table setting.
To move a table from the system tablespace to its own
tablespace, or vice versa, change the
innodb_file_per_table setting
and rebuild the table:
-- Move table from system tablespace to its own tablespace. SET GLOBAL innodb_file_per_table=1; ALTER TABLEtable_nameENGINE=InnoDB; -- Move table from its own tablespace to system tablespace. SET GLOBAL innodb_file_per_table=0; ALTER TABLEtable_nameENGINE=InnoDB;
InnoDB always needs the system tablespace
because it puts its internal
data dictionary
and undo logs there. The
.ibd files are not sufficient for
InnoDB to operate.
When a table is moved out of the system tablespace into its
own .ibd file, the data files that make
up the system tablespace remain the same size. The space
formerly occupied by the table can be reused for new
InnoDB data, but is not reclaimed for use
by the operating system. When moving large
InnoDB tables out of the system tablespace,
where disk space is limited, you might prefer to turn on
innodb_file_per_table and then recreate the
entire instance using the mysqldump
command.
To create a new InnoDB table in a specific
location outside the MySQL data directory, use the DATA
DIRECTORY =
clause of the absolute_path_to_directoryCREATE TABLE
statement. (Plan the location in advance, because you cannot use
this clause with the ALTER TABLE
statement.) The directory you specify could be on another
storage device with particular performance or capacity
characteristics, such as a fast
SSD or a high-capacity
HDD.
Within the destination directory, MySQL creates a subdirectory
corresponding to the database name, and within that a
.ibd file for the new
table. In the database directory underneath the MySQL
DATADIR directory, MySQL
creates a
table_name.islInnoDB tables.)
The following example shows how you might run a small
development or test instance of MySQL on a laptop with a primary
hard drive that is 95% full, and place a new table
EXTERNAL on a different storage device with
more free space. The shell commands show the different paths to
the LOCAL table in its default location under
the DATADIR directory, and the
EXTERNAL table in the location you specified:
mysql> \! df -k . Filesystem 1024-blocks Used Available Capacity iused ifree %iused Mounted on /dev/disk0s2 244277768 231603532 12418236 95% 57964881 3104559 95% / mysql> use test; Database changed mysql> show variables like 'innodb_file_per_table'; +-----------------------+-------+ | Variable_name | Value | +-----------------------+-------+ | innodb_file_per_table | ON | +-----------------------+-------+ 1 row in set (0.00 sec) mysql> \! pwd /usr/local/mysql mysql> create table local (x int unsigned not null primary key); Query OK, 0 rows affected (0.03 sec) mysql> \! ls -l data/test/local.ibd -rw-rw---- 1 cirrus staff 98304 Nov 13 15:24 data/test/local.ibd mysql> create table external (x int unsigned not null primary key) data directory = '/volumes/external1/data'; Query OK, 0 rows affected (0.03 sec) mysql> \! ls -l /volumes/external1/data/test/external.ibd -rwxrwxrwx 1 cirrus staff 98304 Nov 13 15:34 /volumes/external1/data/test/external.ibd mysql> select count(*) from local; +----------+ | count(*) | +----------+ | 0 | +----------+ 1 row in set (0.01 sec) mysql> select count(*) from external; +----------+ | count(*) | +----------+ | 0 | +----------+ 1 row in set (0.01 sec)
MySQL initially holds the
.ibdfile open, preventing you from dismounting the device, but might eventually close the table if the server is busy. Be careful not to accidentally dismount the external device while MySQL is running, or to start MySQL while the device is disconnected. Attempting to access a table when the associated.ibdfile is missing causes a serious error that requires a server restart.The server restart might fail if the
.ibdfile is still not at the expected path. In this case, manually remove thetable_name.islDROP TABLEto delete the.frmfile and remove the information about the table from the data dictionary.Do not put MySQL tables on an NFS-mounted volume. NFS uses a message-passing protocol to write to files, which could cause data inconsistency if network messages are lost or received out of order.
If you use an LVM snapshot, file copy, or other file-based mechanism to back up the
.ibdfile, always use theFLUSH TABLES ... FOR EXPORTstatement first to make sure all changes that were buffered in memory are flushed to disk before the backup occurs.The
DATA DIRECTORYclause is a supported alternative to using symbolic links, which has always been problematic and was never supported for individualInnoDBtables.
There are many reasons why you might copy an
InnoDB table to a different database server:
To run reports without putting extra load on a production server.
To set up identical data for a table on a new slave server.
To restore a backed-up version of a table after a problem or mistake.
As a faster way of moving data around than importing the results of a mysqldump command. The data is available immediately, rather than having to be re-inserted and the indexes rebuilt.
To copy an InnoDB table to another server
instance or to perform a full table restore on the same
instance, you can use the
FLUSH
TABLES statement with the FOR
EXPORT clause.
FLUSH TABLES ... FOR
EXPORT places .ibd files into a
consistent state so that they can be copied. It also and creates
a .cfg binary metadata file that is used by
ALTER
TABLE ... IMPORT TABLESPACE for schema verification
during the import process. See Section 13.7.6.3, “FLUSH Syntax” for
additional information about
FLUSH TABLES ... FOR
EXPORT.
As of MySQL 5.6.8,
ALTER TABLE ...
IMPORT TABLESPACE does not require a
.cfg metadata file to import a tablespace.
However, metadata checks are not performed when importing
without a .cfg file, and the following
warning will be issued:
Message: InnoDB: IO Read error: (2, No such file or directory) Error opening '.\
test\t.cfg', will attempt to import without schema verification
1 row in set (0.00 sec)
The ability to import without a .cfg file
may be more convenient when no schema mismatches are expected.
Additionally, the ability to import without a
.cfg file could be useful in crash and
recovery scenarios in which metadata cannot be collected from an
.ibd file.
Tablespace Copying Limitations (Transportable Tablespaces)
The tablespace copy procedure is only possible when
innodb_file_per_tableis set toON. Tables residing in the shared system tablespace cannot be quiesced.When a table is quiesced, only read-only transactions are allowed on the affected table.
When importing a tablespace, the page size must match the page size of the importing instance.
DISCARD TABLESPACEis not supported for partitioned tables meaning that transportable tablespaces is also unsupported. If you runALTER TABLE ... DISCARD TABLESPACEon a partitioned table, the following error is returned: ERROR 1031 (HY000): Table storage engine for 'part' doesn't have this option.DISCARD TABLESPACEis not supported for tablespaces with a parent-child (primary key-foreign key) relationship whenforeign_key_checksis set to1. Before discarding a tablespace for parent-child tables, setforeign_key_checks=0.ALTER TABLE ... IMPORT TABLESPACEdoes not enforce foreign key constraints on imported data. If there are foreign key constraints between tables, all tables should be exported at the same (logical) point in time.In replication scenarios,
innodb_file_per_tablemust be set toONon both the master and slave.
Example Procedure: Copying a Tablespace From One Server To Another (Transportable Tablespaces)
This procedure describes how to copy a table from a running MySQL server instance to another running instance. The same procedure, with minor adjustments, can also be used to perform a full table restore on the same instance.
On the source server, create a table if one does not already exist:
mysql> use test; mysql> CREATE TABLE t(c1 INT) engine=InnoDB;
On the destination server, create a table if one does not already exist:
mysql> use test; mysql> CREATE TABLE t(c1 INT) engine=InnoDB;
On the destination server, discard the existing tablespace. Before a tablespace can be imported,
InnoDBmust discard the tablespace that is attached to the receiving table.mysql> ALTER TABLE t DISCARD TABLESPACE;
NoteThe tablespace file need not necessarily have been created on the server into which the tablespace file is being imported. In MySQL 5.6 or later, importing a tablespace file from another server works if the both servers have GA (General Availability) status and their versions are within the same series. Otherwise, the file must have been created on the server into which it is imported.
On the source server, quiesce the table and create the
.cfgmetadata file by runningFLUSH TABLES ... FOR EXPORT:mysql> use test; mysql> FLUSH TABLES t FOR EXPORT;
The metadata file (
t.cfg) is created in theInnoDBdata directory.NoteFLUSH TABLES ... FOR EXPORTapplies toInnoDBtables. It is available as of MySQL 5.6.6. The statement ensures that changes to the named tables have been flushed to disk so that binary table copies can be made while the server is running. WhenFLUSH TABLES ... FOR EXPORTis run,InnoDBproduces a.cfgfile in the same database directory as the table. The.cfgfile contains metadata used for schema verification when importing the tablespace file.Copy the
.cfgmetadata file and.ibdfile from the source server to the destination server. For example:shell> scp /innodb_data_dir/test/t.{ibd,cfg} destination-server:/innodb_data_dir/testNoteThe
.ibdfile and.cfgfile must be copied before releasing the shared locks, as described in the next step.On the source server, use
UNLOCK TABLESto release the locks acquired byFLUSH TABLES ... FOR EXPORT:mysql> use test; mysql> UNLOCK TABLES;
On the destination server, import the tablespace:
mysql> use test; mysql> ALTER TABLE t IMPORT TABLESPACE;
NoteThe
ALTER TABLE ... IMPORT TABLESPACEfeature does not enforce foreign key constraints on imported data. If there are foreign key constraints between tables, all tables should be exported at the same (logical) point in time. In this case you would stop updating the tables, commit all transactions, acquire shared locks on the tables, and then perform the export operation.
Tablespace Copying Internals (Transportable Tablespaces)
The following information describes internals and error log messaging for the transportable tablespaces copy procedure.
When ALTER TABLE
... DISCARD TABLESPACE is run on the destination
instance:
The table is locked in X mode.
The tablespace is detached from the table.
When FLUSH TABLES ... FOR
EXPORT is run on the source instance:
The table being flushed for export is locked in shared mode.
The purge coordinator thread is stopped.
Dirty pages are synchronized to disk.
Table metadata is written to the binary
.cfgfile.
Expected error log messages for this operation:
2013-07-18 14:47:31 34471 [Note] InnoDB: Sync to disk of '"test"."t"' started. 2013-07-18 14:47:31 34471 [Note] InnoDB: Stopping purge 2013-07-18 14:47:31 34471 [Note] InnoDB: Writing table metadata to './test/t.cfg' 2013-07-18 14:47:31 34471 [Note] InnoDB: Table '"test"."t"' flushed to disk
When UNLOCK
TABLES is run on the source instance:
The binary .cfg file is deleted.
The shared lock on the table or tables being imported is released and the purge coordinator thread is restarted.
Expected error log messages for this operation:
2013-07-18 15:01:40 34471 [Note] InnoDB: Deleting the meta-data file './test/t.cfg' 2013-07-18 15:01:40 34471 [Note] InnoDB: Resuming purge
When ALTER TABLE
... IMPORT TABLESPACE is run on the destination
instance, the import algorithm performs the following operations
for each tablespace being imported:
Each tablespace page is checked for corruption.
The space ID and log sequence numbers (LSNs) on each page are updated
Flags are validated and LSN updated for the header page.
Btree pages are updated.
The page state is set to dirty so that it will be written to disk.
Expected error log messages for this operation:
2013-07-18 15:15:01 34960 [Note] InnoDB: Importing tablespace for table 'test/t' that was exported from host 'ubuntu' 2013-07-18 15:15:01 34960 [Note] InnoDB: Phase I - Update all pages 2013-07-18 15:15:01 34960 [Note] InnoDB: Sync to disk 2013-07-18 15:15:01 34960 [Note] InnoDB: Sync to disk - done! 2013-07-18 15:15:01 34960 [Note] InnoDB: Phase III - Flush changes to disk 2013-07-18 15:15:01 34960 [Note] InnoDB: Phase IV - Flush complete
You may also receive a warning that a tablespace is discarded
(if you discarded the tablespace for the destination table)
and a message stating that statistics could not be calculated
due to a missing .ibd file:
2013-07-18 15:14:38 34960 [Warning] InnoDB: Table "test"."t" tablespace is set as discarded. 2013-07-18 15:14:38 7f34d9a37700 InnoDB: cannot calculate statistics for table "test"."t" because the .ibd file is missing. For help, please refer to http://dev.mysql.com/doc/refman/5.7/en/innodb-troubleshooting.html
Although tablespace management typically involves files holding tables and indexes, you can also divide the undo log into separate undo tablespace files. This layout is different from the default configuration where the undo log is part of the system tablespace. See Section 14.2.4.2.4, “Separate Tablespaces for InnoDB Undo Logs” for details.
By default, connection to the MySQL server begins with autocommit mode enabled, which automatically commits every SQL statement as you execute it. This mode of operation might be unfamiliar if you have experience with other database systems, where it is standard practice to issue a sequence of DML statements and commit them or roll them back all together.
To use multiple-statement
transactions, switch
autocommit off with the SQL statement SET autocommit =
0 and end each transaction with
COMMIT or
ROLLBACK as
appropriate. To leave autocommit on, begin each transaction with
START
TRANSACTION and end it with
COMMIT or
ROLLBACK. The
following example shows two transactions. The first is committed;
the second is rolled back.
shell>mysql testmysql>CREATE TABLE customer (a INT, b CHAR (20), INDEX (a));Query OK, 0 rows affected (0.00 sec) mysql>-- Do a transaction with autocommit turned on.mysql>START TRANSACTION;Query OK, 0 rows affected (0.00 sec) mysql>INSERT INTO customer VALUES (10, 'Heikki');Query OK, 1 row affected (0.00 sec) mysql>COMMIT;Query OK, 0 rows affected (0.00 sec) mysql>-- Do another transaction with autocommit turned off.mysql>SET autocommit=0;Query OK, 0 rows affected (0.00 sec) mysql>INSERT INTO customer VALUES (15, 'John');Query OK, 1 row affected (0.00 sec) mysql>INSERT INTO customer VALUES (20, 'Paul');Query OK, 1 row affected (0.00 sec) mysql>DELETE FROM customer WHERE b = 'Heikki';Query OK, 1 row affected (0.00 sec) mysql>-- Now we undo those last 2 inserts and the delete.mysql>ROLLBACK;Query OK, 0 rows affected (0.00 sec) mysql>SELECT * FROM customer;+------+--------+ | a | b | +------+--------+ | 10 | Heikki | +------+--------+ 1 row in set (0.00 sec) mysql>
Transactions in Client-Side Languages
In APIs such as PHP, Perl DBI, JDBC, ODBC, or the standard C call
interface of MySQL, you can send transaction control statements
such as COMMIT to the MySQL server
as strings just like any other SQL statements such as
SELECT or
INSERT. Some APIs also offer
separate special transaction commit and rollback functions or
methods.
If you have existing tables, and applications that use them, that
you want to convert to InnoDB for better
reliability and scalability, use the following guidelines and tips.
This section assumes most such tables were originally
MyISAM, which was formerly the default.
Reduce Memory Usage for MyISAM, Increase Memory Usage for InnoDB
As you transition away from MyISAM tables, lower
the value of the key_buffer_size
configuration option to free memory no longer needed for caching
results. Increase the value of the
innodb_buffer_pool_size
configuration option, which performs a similar role of allocating
cache memory for InnoDB tables. The
InnoDB buffer
pool caches both table data and index data, so it does double
duty in speeding up lookups for queries and keeping query results in
memory for reuse.
Allocate as much memory to this option as you can afford, often up to 80% of physical memory on the server.
If the operating system runs short of memory for other processes and begins to swap, reduce the
innodb_buffer_pool_sizevalue. Swapping is such an expensive operation that it drastically reduces the benefit of the cache memory.If the
innodb_buffer_pool_sizevalue is several gigabytes or higher, consider increasing the values ofinnodb_buffer_pool_instances. Doing so helps on busy servers where many connections are reading data into the cache at the same time.On a busy server, run benchmarks with the Query Cache turned off. The
InnoDBbuffer pool provides similar benefits, so the Query Cache might be tying up memory unnecessarily.
Watch Out for Too-Long Or Too-Short Transactions
Because MyISAM tables do not support
transactions, you might not
have paid much attention to the
autocommit configuration option and
the COMMIT and
ROLLBACK
statements. These keywords are important to allow multiple sessions
to read and write InnoDB tables concurrently,
providing substantial scalability benefits in write-heavy workloads.
While a transaction is open, the system keeps a snapshot of the data as seen at the beginning of the transaction, which can cause substantial overhead if the system inserts, updates, and deletes millions of rows while a stray transaction keeps running. Thus, take care to avoid transactions that run for too long:
If you are using a mysql session for interactive experiments, always
COMMIT(to finalize the changes) orROLLBACK(to undo the changes) when finished. Close down interactive sessions rather than leaving them open for long periods, to avoid keeping transactions open for long periods by accident.Make sure that any error handlers in your application also
ROLLBACKincomplete changes orCOMMITcompleted changes.ROLLBACKis a relatively expensive operation, becauseINSERT,UPDATE, andDELETEoperations are written toInnoDBtables prior to theCOMMIT, with the expectation that most changes will be committed successfully and rollbacks will be rare. When experimenting with large volumes of data, avoid making changes to large numbers of rows and then rolling back those changes.When loading large volumes of data with a sequence of
INSERTstatements, periodicallyCOMMITthe results to avoid having transactions that last for hours. In typical load operations for data warehousing, if something goes wrong, youTRUNCATE TABLEand start over from the beginning rather than doing aROLLBACK.
The preceding tips save memory and disk space that can be wasted
during too-long transactions. When transactions are shorter than
they should be, the problem is excessive I/O. With each
COMMIT, MySQL makes sure each change
is safely recorded to disk, which involves some I/O.
For most operations on
InnoDBtables, you should use the settingautocommit=0. From an efficiency perspective, this avoids unnecessary I/O when you issue large numbers of consecutiveINSERT,UPDATE, orDELETEstatements. From a safety perspective, this allows you to issue aROLLBACKstatement to recover lost or garbled data if you make a mistake on the mysql command line, or in an exception handler in your application.The time when
autocommit=1is suitable forInnoDBtables is when running a sequence of queries for generating reports or analyzing statistics. In this situation, there is no I/O penalty related toCOMMITorROLLBACK, andInnoDBcan automatically optimize the read-only workload.If you make a series of related changes, finalize all those changes at once with a single
COMMITat the end. For example, if you insert related pieces of information into several tables, do a singleCOMMITafter making all the changes. Or if you run many consecutiveINSERTstatements, do a singleCOMMITafter all the data is loaded; if you are doing millions ofINSERTstatements, perhaps split up the huge transaction by issuing aCOMMITevery ten thousand or hundred thousand records, so the transaction does not grow too large.Remember that even a
SELECTstatement opens a transaction, so after running some report or debugging queries in an interactive mysql session, either issue aCOMMITor close the mysql session.
Don't Worry Too Much About Deadlocks
You might see warning messages referring to “deadlocks”
in the MySQL error log, or the output of
SHOW ENGINE INNODB
STATUS. Despite the scary-sounding name, a
deadlock is not a serious issue
for InnoDB tables, and often does not require any
corrective action. When two transactions start modifying multiple
tables, accessing the tables in a different order, they can reach a
state where each transaction is waiting for the other and neither
can proceed. MySQL immediately detects this condition and cancels
(rolls back) the
“smaller” transaction, allowing the other to proceed.
Your applications do need error-handling logic to restart a transaction that is forcibly cancelled like this. When you re-issue the same SQL statements as before, the original timing issue no longer applies: either the other transaction has already finished and yours can proceed, or the other transaction is still in progress and your transaction waits until it finishes.
If deadlock warnings occur constantly, you might review the
application code to reorder the SQL operations in a consistent way,
or to shorten the transactions. You can test with the
innodb_print_all_deadlocks option
enabled to see all deadlock warnings in the MySQL error log, rather
than only the last warning in the
SHOW ENGINE INNODB
STATUS output.
Plan the Storage Layout
To get the best performance from InnoDB tables,
you can adjust a number of parameters related to storage layout.
When you convert MyISAM tables that are large,
frequently accessed, and hold vital data, investigate and consider
the innodb_file_per_table,
innodb_file_format, and
innodb_page_size configuration
options, and the
ROW_FORMAT and
KEY_BLOCK_SIZE clauses of the
CREATE TABLE statement.
During your initial experiments, the most important setting is
innodb_file_per_table. Enabling
this option before creating new InnoDB tables
ensures that the InnoDB
system tablespace
files do not allocate disk space permanently for all the
InnoDB data. With
innodb_file_per_table enabled,
DROP TABLE and
TRUNCATE TABLE free disk space as you
would expect.
Converting an Existing Table
To convert a non-InnoDB table to use
InnoDB use ALTER
TABLE:
ALTER TABLE table_name ENGINE=InnoDB;
Do not convert MySQL system tables in the mysql
database (such as user or
host) to the InnoDB type.
This is an unsupported operation. The system tables must always be
of the MyISAM type.
Cloning the Structure of a Table
You might make an InnoDB table that is a clone of a MyISAM table,
rather than doing the ALTER TABLE
conversion, to test the old and new table side-by-side before
switching.
Create an empty InnoDB table with identical
column and index definitions. Use show create table
to see the full
table_name\GCREATE TABLE statement to use. Change
the ENGINE clause to
ENGINE=INNODB.
Transferring Existing Data
To transfer a large volume of data into an empty
InnoDB table created as shown in the previous
section, insert the rows with INSERT INTO
.
innodb_table SELECT * FROM
myisam_table ORDER BY
primary_key_columns
You can also create the indexes for the InnoDB
table after inserting the data. Historically, creating new secondary
indexes was a slow operation for InnoDB, but now you can create the
indexes after the data is loaded with relatively little overhead
from the index creation step.
If you have UNIQUE constraints on secondary keys,
you can speed up a table import by turning off the uniqueness checks
temporarily during the import operation:
SET unique_checks=0;... import operation ...
SET unique_checks=1;
For big tables, this saves disk I/O because
InnoDB can use its
insert buffer to write
secondary index records as a batch. Be certain that the data
contains no duplicate keys.
unique_checks permits but does not
require storage engines to ignore duplicate keys.
To get better control over the insertion process, you might insert big tables in pieces:
INSERT INTO newtable SELECT * FROM oldtable WHERE yourkey >somethingAND yourkey <=somethingelse;
After all records have been inserted, you can rename the tables.
During the conversion of big tables, increase the size of the
InnoDB buffer pool to reduce disk I/O, to a
maximum of 80% of physical memory. You can also increase the sizes
of the InnoDB log files.
Storage Requirements
By this point, as already mentioned, you should already have the
innodb_file_per_table option
enabled, so that if you temporarily make several copies of your data
in InnoDB tables, you can recover all that disk
space by dropping unneeded tables afterward.
Whether you convert the MyISAM table directly or
create a cloned InnoDB table, make sure that you
have sufficient disk space to hold both the old and new tables
during the process. InnoDB tables require more
disk space than MyISAM tables. If an
ALTER TABLE operation runs out of
space, it starts a rollback, and that can take hours if it is
disk-bound. For inserts, InnoDB uses the insert
buffer to merge secondary index records to indexes in batches. That
saves a lot of disk I/O. For rollback, no such mechanism is used,
and the rollback can take 30 times longer than the insertion.
In the case of a runaway rollback, if you do not have valuable data
in your database, it may be advisable to kill the database process
rather than wait for millions of disk I/O operations to complete.
For the complete procedure, see
Section 14.2.4.6, “Starting InnoDB on a Corrupted Database”.
Carefully Choose a PRIMARY KEY for Each Table
The PRIMARY KEY clause is a critical factor
affecting the performance of MySQL queries and the space usage for
tables and indexes. Perhaps you have phoned a financial institution
where you are asked for an account number. If you do not have the
number, you are asked for a dozen different pieces of information to
“uniquely identify” yourself. The primary key is like
that unique account number that lets you get straight down to
business when querying or modifying the information in a table.
Every row in the table must have a primary key value, and no two
rows can have the same primary key value.
Here are some guidelines for the primary key, followed by more detailed explanations.
Declare a
PRIMARY KEYfor each table. Typically, it is the most important column that you refer to inWHEREclauses when looking up a single row.Declare the
PRIMARY KEYclause in the originalCREATE TABLEstatement, rather than adding it later through anALTER TABLEstatement.Choose the column and its data type carefully. Prefer numeric columns over character or string ones.
Consider using an auto-increment column if there is not another stable, unique, non-null, numeric column to use.
An auto-increment column is also a good choice if there is any doubt whether the value of the primary key column could ever change. Changing the value of a primary key column is an expensive operation, possibly involving rearranging data within the table and within each secondary index.
Consider adding a primary key to any table that does not already have one. Use the smallest practical numeric type based on the maximum projected size of the table. This can make each row slightly more compact, which can yield substantial space savings for large tables. The space savings are multiplied if the table has any secondary indexes, because the primary key value is repeated in each secondary index entry. In addition to reducing data size on disk, a small primary key also lets more data fit into the buffer pool, speeding up all kinds of operations and improving concurrency.
If the table already has a primary key on some longer column, such
as a VARCHAR, consider adding a new unsigned
AUTO_INCREMENT column and switching the primary
key to that, even if that column is not referenced in queries. This
design change can produce substantial space savings in the secondary
indexes. You can designate the former primary key columns as
UNIQUE NOT NULL to enforce the same constraints
as the PRIMARY KEY clause, that is, to prevent
duplicate or null values across all those columns.
If you spread related information across multiple tables, typically each table uses the same column for its primary key. For example, a personnel database might have several tables, each with a primary key of employee number. A sales database might have some tables with a primary key of customer number, and other tables with a primary key of order number. Because lookups using the primary key are very fast, you can construct efficient join queries for such tables.
If you leave the PRIMARY KEY clause out entirely,
MySQL creates an invisible one for you. It is a 6-byte value that
might be longer than you need, thus wasting space. Because it is
hidden, you cannot refer to it in queries.
Application Performance Considerations
The extra reliability and scalability features of
InnoDB do require more disk storage than
equivalent MyISAM tables. You might change the
column and index definitions slightly, for better space utilization,
reduced I/O and memory consumption when processing result sets, and
better query optimization plans making efficient use of index
lookups.
If you do set up a numeric ID column for the primary key, use that
value to cross-reference with related values in any other tables,
particularly for join queries. For
example, rather than accepting a country name as input and doing
queries searching for the same name, do one lookup to determine the
country ID, then do other queries (or a single join query) to look
up relevant information across several tables. Rather than storing a
customer or catalog item number as a string of digits, potentially
using up several bytes, convert it to a numeric ID for storing and
querying. A 4-byte unsigned INT
column can index over 4 billion items (with the US meaning of
billion: 1000 million). For the ranges of the different integer
types, see Section 11.2.1, “Integer Types (Exact Value) - INTEGER,
INT, SMALLINT,
TINYINT, MEDIUMINT,
BIGINT”.
Understand Files Associated with InnoDB Tables
InnoDB files require more care and planning than
MyISAM files do:
You must not delete the ibdata files that represent the
InnoDBsystem tablespace.Copying InnoDB tables from one server to another requires issuing the
FLUSH TABLES ... FOR EXPORTstatement first, and copying thetable_name.cfgtable_name.ibd
InnoDB provides an optimization that
significantly improves scalability and performance of SQL
statements that insert rows into tables with
AUTO_INCREMENT columns. To use the
AUTO_INCREMENT mechanism with an
InnoDB table, an
AUTO_INCREMENT column
ai_col must be defined as part of an
index such that it is possible to perform the equivalent of an
indexed SELECT
MAX( lookup on the
table to obtain the maximum column value. Typically, this is
achieved by making the column the first column of some table
index.
ai_col)
This section provides background information on the original
(“traditional”) implementation of auto-increment
locking in InnoDB, explains the configurable
locking mechanism, documents the parameter for configuring the
mechanism, and describes its behavior and interaction with
replication.
The original implementation of auto-increment handling in
InnoDB uses the following strategy to prevent
problems when using the binary log for statement-based
replication or for certain recovery scenarios.
If you specify an AUTO_INCREMENT column for
an InnoDB table, the table handle in the
InnoDB data dictionary contains a special
counter called the auto-increment counter that is used in
assigning new values for the column. This counter is stored only
in main memory, not on disk.
InnoDB uses the following algorithm to
initialize the auto-increment counter for a table
t that contains an
AUTO_INCREMENT column named
ai_col: After a server startup, for the first
insert into a table t,
InnoDB executes the equivalent of this
statement:
SELECT MAX(ai_col) FROM t FOR UPDATE;
InnoDB increments the value retrieved by the
statement and assigns it to the column and to the auto-increment
counter for the table. By default, the value is incremented by
one. This default can be overridden by the
auto_increment_increment
configuration setting.
If the table is empty, InnoDB uses the value
1. This default can be overridden by the
auto_increment_offset
configuration setting.
If a SHOW TABLE STATUS statement
examines the table t before the
auto-increment counter is initialized, InnoDB
initializes but does not increment the value and stores it for
use by later inserts. This initialization uses a normal
exclusive-locking read on the table and the lock lasts to the
end of the transaction.
InnoDB follows the same procedure for
initializing the auto-increment counter for a freshly created
table.
After the auto-increment counter has been initialized, if you do
not explicitly specify a value for an
AUTO_INCREMENT column,
InnoDB increments the counter and assigns the
new value to the column. If you insert a row that explicitly
specifies the column value, and the value is bigger than the
current counter value, the counter is set to the specified
column value.
If a user specifies NULL or
0 for the AUTO_INCREMENT
column in an INSERT,
InnoDB treats the row as if the value was not
specified and generates a new value for it.
The behavior of the auto-increment mechanism is not defined if you assign a negative value to the column, or if the value becomes bigger than the maximum integer that can be stored in the specified integer type.
When accessing the auto-increment counter,
InnoDB uses a special table-level
AUTO-INC lock that it keeps to the end of the
current SQL statement, not to the end of the transaction. The
special lock release strategy was introduced to improve
concurrency for inserts into a table containing an
AUTO_INCREMENT column. Nevertheless, two
transactions cannot have the AUTO-INC lock on
the same table simultaneously, which can have a performance
impact if the AUTO-INC lock is held for a
long time. That might be the case for a statement such as
INSERT INTO t1 ... SELECT ... FROM t2 that
inserts all rows from one table into another.
InnoDB uses the in-memory auto-increment
counter as long as the server runs. When the server is stopped
and restarted, InnoDB reinitializes the
counter for each table for the first
INSERT to the table, as described
earlier.
A server restart also cancels the effect of the
AUTO_INCREMENT =
table option in NCREATE TABLE and
ALTER TABLE statements, which you
can use with InnoDB tables to set the initial
counter value or alter the current counter value.
You may see gaps in the sequence of values assigned to the
AUTO_INCREMENT column if you roll back
transactions that have generated numbers using the counter.
As described in the previous section, InnoDB
uses a special lock called the table-level
AUTO-INC lock for inserts into tables with
AUTO_INCREMENT columns. This lock is normally
held to the end of the statement (not to the end of the
transaction), to ensure that auto-increment numbers are assigned
in a predictable and repeatable order for a given sequence of
INSERT statements.
In the case of statement-based replication, this means that when
an SQL statement is replicated on a slave server, the same
values are used for the auto-increment column as on the master
server. The result of execution of multiple
INSERT statements is
deterministic, and the slave reproduces the same data as on the
master. If auto-increment values generated by multiple
INSERT statements were
interleaved, the result of two concurrent
INSERT statements would be
nondeterministic, and could not reliably be propagated to a
slave server using statement-based replication.
To make this clear, consider an example that uses this table:
CREATE TABLE t1 ( c1 INT(11) NOT NULL AUTO_INCREMENT, c2 VARCHAR(10) DEFAULT NULL, PRIMARY KEY (c1) ) ENGINE=InnoDB;
Suppose that there are two transactions running, each inserting
rows into a table with an AUTO_INCREMENT
column. One transaction is using an
INSERT ...
SELECT statement that inserts 1000 rows, and another
is using a simple INSERT
statement that inserts one row:
Tx1: INSERT INTO t1 (c2) SELECT 1000 rows from another table ...
Tx2: INSERT INTO t1 (c2) VALUES ('xxx');
InnoDB cannot tell in advance how many rows
will be retrieved from the SELECT
in the INSERT statement in Tx1,
and it assigns the auto-increment values one at a time as the
statement proceeds. With a table-level lock, held to the end of
the statement, only one INSERT
statement referring to table t1 can execute
at a time, and the generation of auto-increment numbers by
different statements is not interleaved. The auto-increment
value generated by the Tx1
INSERT ...
SELECT statement will be consecutive, and the (single)
auto-increment value used by the
INSERT statement in Tx2 will
either be smaller or larger than all those used for Tx1,
depending on which statement executes first.
As long as the SQL statements execute in the same order when
replayed from the binary log (when using statement-based
replication, or in recovery scenarios), the results will be the
same as they were when Tx1 and Tx2 first ran. Thus, table-level
locks held until the end of a statement make
INSERT statements using
auto-increment safe for use with statement-based replication.
However, those locks limit concurrency and scalability when
multiple transactions are executing insert statements at the
same time.
In the preceding example, if there were no table-level lock, the
value of the auto-increment column used for the
INSERT in Tx2 depends on
precisely when the statement executes. If the
INSERT of Tx2 executes while the
INSERT of Tx1 is running (rather
than before it starts or after it completes), the specific
auto-increment values assigned by the two
INSERT statements are
nondeterministic, and may vary from run to run.
InnoDB can avoid using the table-level
AUTO-INC lock for a class of
INSERT statements where the
number of rows is known in advance, and still preserve
deterministic execution and safety for statement-based
replication. Further, if you are not using the binary log to
replay SQL statements as part of recovery or replication, you
can entirely eliminate use of the table-level
AUTO-INC lock for even greater concurrency
and performance, at the cost of permitting gaps in
auto-increment numbers assigned by a statement and potentially
having the numbers assigned by concurrently executing statements
interleaved.
For INSERT statements where the
number of rows to be inserted is known at the beginning of
processing the statement, InnoDB quickly
allocates the required number of auto-increment values without
taking any lock, but only if there is no concurrent session
already holding the table-level AUTO-INC lock
(because that other statement will be allocating auto-increment
values one-by-one as it proceeds). More precisely, such an
INSERT statement obtains
auto-increment values under the control of a mutex (a
light-weight lock) that is not held until
the statement completes, but only for the duration of the
allocation process.
This new locking scheme enables much greater scalability, but it
does introduce some subtle differences in how auto-increment
values are assigned compared to the original mechanism. To
describe the way auto-increment works in
InnoDB, the following discussion defines some
terms, and explains how InnoDB behaves using
different settings of the
innodb_autoinc_lock_mode
configuration parameter, which you can set at server startup.
Additional considerations are described following the
explanation of auto-increment locking behavior.
First, some definitions:
“
INSERT-like” statementsAll statements that generate new rows in a table, including
INSERT,INSERT ... SELECT,REPLACE,REPLACE ... SELECT, andLOAD DATA.“Simple inserts”
Statements for which the number of rows to be inserted can be determined in advance (when the statement is initially processed). This includes single-row and multiple-row
INSERTandREPLACEstatements that do not have a nested subquery, but notINSERT ... ON DUPLICATE KEY UPDATE.“Bulk inserts”
Statements for which the number of rows to be inserted (and the number of required auto-increment values) is not known in advance. This includes
INSERT ... SELECT,REPLACE ... SELECT, andLOAD DATAstatements, but not plainINSERT.InnoDBwill assign new values for theAUTO_INCREMENTcolumn one at a time as each row is processed.“Mixed-mode inserts”
These are “simple insert” statements that specify the auto-increment value for some (but not all) of the new rows. An example follows, where
c1is anAUTO_INCREMENTcolumn of tablet1:INSERT INTO t1 (c1,c2) VALUES (1,'a'), (NULL,'b'), (5,'c'), (NULL,'d');
Another type of “mixed-mode insert” is
INSERT ... ON DUPLICATE KEY UPDATE, which in the worst case is in effect anINSERTfollowed by aUPDATE, where the allocated value for theAUTO_INCREMENTcolumn may or may not be used during the update phase.
There are three possible settings for the
innodb_autoinc_lock_mode
parameter:
innodb_autoinc_lock_mode = 0(“traditional” lock mode)This lock mode provides the same behavior as before
innodb_autoinc_lock_modeexisted. For all “INSERT-like” statements, a special table-levelAUTO-INClock is obtained and held to the end of the statement. This assures that the auto-increment values assigned by any given statement are consecutive.This lock mode is provided for:
Backward compatibility.
Performance testing.
Working around issues with “mixed-mode inserts”, due to the possible differences in semantics described later.
innodb_autoinc_lock_mode = 1(“consecutive” lock mode)This is the default lock mode. In this mode, “bulk inserts” use the special
AUTO-INCtable-level lock and hold it until the end of the statement. This applies to allINSERT ... SELECT,REPLACE ... SELECT, andLOAD DATAstatements. Only one statement holding theAUTO-INClock can execute at a time.With this lock mode, “simple inserts” (only) use a new locking model where a light-weight mutex is used during the allocation of auto-increment values, and no table-level
AUTO-INClock is used, unless anAUTO-INClock is held by another transaction. If another transaction does hold anAUTO-INClock, a “simple insert” waits for theAUTO-INClock, as if it too were a “bulk insert”.This lock mode ensures that, in the presence of
INSERTstatements where the number of rows is not known in advance (and where auto-increment numbers are assigned as the statement progresses), all auto-increment values assigned by any “INSERT-like” statement are consecutive, and operations are safe for statement-based replication.Simply put, the important impact of this lock mode is significantly better scalability. This mode is safe for use with statement-based replication. Further, as with “traditional” lock mode, auto-increment numbers assigned by any given statement are consecutive. In this mode, there is no change in semantics compared to “traditional” mode for any statement that uses auto-increment, with one important exception.
The exception is for “mixed-mode inserts”, where the user provides explicit values for an
AUTO_INCREMENTcolumn for some, but not all, rows in a multiple-row “simple insert”. For such inserts,InnoDBwill allocate more auto-increment values than the number of rows to be inserted. However, all values automatically assigned are consecutively generated (and thus higher than) the auto-increment value generated by the most recently executed previous statement. “Excess” numbers are lost.innodb_autoinc_lock_mode = 2(“interleaved” lock mode)In this lock mode, no “
INSERT-like” statements use the table-levelAUTO-INClock, and multiple statements can execute at the same time. This is the fastest and most scalable lock mode, but it is not safe when using statement-based replication or recovery scenarios when SQL statements are replayed from the binary log.In this lock mode, auto-increment values are guaranteed to be unique and monotonically increasing across all concurrently executing “
INSERT-like” statements. However, because multiple statements can be generating numbers at the same time (that is, allocation of numbers is interleaved across statements), the values generated for the rows inserted by any given statement may not be consecutive.If the only statements executing are “simple inserts” where the number of rows to be inserted is known ahead of time, there will be no gaps in the numbers generated for a single statement, except for “mixed-mode inserts”. However, when “bulk inserts” are executed, there may be gaps in the auto-increment values assigned by any given statement.
The auto-increment locking modes provided by
innodb_autoinc_lock_mode have
several usage implications:
Using auto-increment with replication
If you are using statement-based replication, set
innodb_autoinc_lock_modeto 0 or 1 and use the same value on the master and its slaves. Auto-increment values are not ensured to be the same on the slaves as on the master if you useinnodb_autoinc_lock_mode= 2 (“interleaved”) or configurations where the master and slaves do not use the same lock mode.If you are using row-based replication, all of the auto-increment lock modes are safe. Row-based replication is not sensitive to the order of execution of the SQL statements.
“Lost” auto-increment values and sequence gaps
In all lock modes (0, 1, and 2), if a transaction that generated auto-increment values rolls back, those auto-increment values are “lost”. Once a value is generated for an auto-increment column, it cannot be rolled back, whether or not the “
INSERT-like” statement is completed, and whether or not the containing transaction is rolled back. Such lost values are not reused. Thus, there may be gaps in the values stored in anAUTO_INCREMENTcolumn of a table.Gaps in auto-increment values for “bulk inserts”
With
innodb_autoinc_lock_modeset to 0 (“traditional”) or 1 (“consecutive”), the auto-increment values generated by any given statement will be consecutive, without gaps, because the table-levelAUTO-INClock is held until the end of the statement, and only one such statement can execute at a time.With
innodb_autoinc_lock_modeset to 2 (“interleaved”), there may be gaps in the auto-increment values generated by “bulk inserts,” but only if there are concurrently executing “INSERT-like” statements.For lock modes 1 or 2, gaps may occur between successive statements because for bulk inserts the exact number of auto-increment values required by each statement may not be known and overestimation is possible.
Auto-increment values assigned by “mixed-mode inserts”
Consider a “mixed-mode insert,” where a “simple insert” specifies the auto-increment value for some (but not all) resulting rows. Such a statement will behave differently in lock modes 0, 1, and 2. For example, assume
c1is anAUTO_INCREMENTcolumn of tablet1, and that the most recent automatically generated sequence number is 100. Consider the following “mixed-mode insert” statement:INSERT INTO t1 (c1,c2) VALUES (1,'a'), (NULL,'b'), (5,'c'), (NULL,'d');
With
innodb_autoinc_lock_modeset to 0 (“traditional”), the four new rows will be:+-----+------+ | c1 | c2 | +-----+------+ | 1 | a | | 101 | b | | 5 | c | | 102 | d | +-----+------+
The next available auto-increment value will be 103 because the auto-increment values are allocated one at a time, not all at once at the beginning of statement execution. This result is true whether or not there are concurrently executing “
INSERT-like” statements (of any type).With
innodb_autoinc_lock_modeset to 1 (“consecutive”), the four new rows will also be:+-----+------+ | c1 | c2 | +-----+------+ | 1 | a | | 101 | b | | 5 | c | | 102 | d | +-----+------+
However, in this case, the next available auto-increment value will be 105, not 103 because four auto-increment values are allocated at the time the statement is processed, but only two are used. This result is true whether or not there are concurrently executing “
INSERT-like” statements (of any type).With
innodb_autoinc_lock_modeset to mode 2 (“interleaved”), the four new rows will be:+-----+------+ | c1 | c2 | +-----+------+ | 1 | a | |
x| b | | 5 | c | |y| d | +-----+------+The values of
xandywill be unique and larger than any previously generated rows. However, the specific values ofxandywill depend on the number of auto-increment values generated by concurrently executing statements.Finally, consider the following statement, issued when the most-recently generated sequence number was the value 4:
INSERT INTO t1 (c1,c2) VALUES (1,'a'), (NULL,'b'), (5,'c'), (NULL,'d');
With any
innodb_autoinc_lock_modesetting, this statement will generate a duplicate-key error 23000 (Can't write; duplicate key in table) because 5 will be allocated for the row(NULL, 'b')and insertion of the row(5, 'c')will fail.
This section describes differences in the InnoDB storage engine' handling of foreign keys as compared with that of the MySQL Server.
Foreign Key Definitions
Foreign key definitions for InnoDB tables are
subject to the following conditions:
InnoDBpermits a foreign key to reference any index column or group of columns. However, in the referenced table, there must be an index where the referenced columns are listed as the first columns in the same order.InnoDBdoes not currently support foreign keys for tables with user-defined partitioning. This means that no user-partitionedInnoDBtable may contain foreign key references or columns referenced by foreign keys.InnoDBallows a foreign key constraint to reference a non-unique key. This is anInnoDBextension to standard SQL.
Referential Actions
Referential actions for foreign keys of InnoDB
tables are subject to the following conditions:
While
SET DEFAULTis allowed by the MySQL Server, it is rejected as invalid byInnoDB.CREATE TABLEandALTER TABLEstatements using this clause are not allowed for InnoDB tables.If there are several rows in the parent table that have the same referenced key value,
InnoDBacts in foreign key checks as if the other parent rows with the same key value do not exist. For example, if you have defined aRESTRICTtype constraint, and there is a child row with several parent rows,InnoDBdoes not permit the deletion of any of those parent rows.InnoDBperforms cascading operations through a depth-first algorithm, based on records in the indexes corresponding to the foreign key constraints.If
ON UPDATE CASCADEorON UPDATE SET NULLrecurses to update the same table it has previously updated during the cascade, it acts likeRESTRICT. This means that you cannot use self-referentialON UPDATE CASCADEorON UPDATE SET NULLoperations. This is to prevent infinite loops resulting from cascaded updates. A self-referentialON DELETE SET NULL, on the other hand, is possible, as is a self-referentialON DELETE CASCADE. Cascading operations may not be nested more than 15 levels deep.Like MySQL in general, in an SQL statement that inserts, deletes, or updates many rows,
InnoDBchecksUNIQUEandFOREIGN KEYconstraints row-by-row. When performing foreign key checks,InnoDBsets shared row-level locks on child or parent records it has to look at.InnoDBchecks foreign key constraints immediately; the check is not deferred to transaction commit. According to the SQL standard, the default behavior should be deferred checking. That is, constraints are only checked after the entire SQL statement has been processed. UntilInnoDBimplements deferred constraint checking, some things will be impossible, such as deleting a record that refers to itself using a foreign key.
Foreign Key Usage and Error Information
You can obtain general information about foreign keys and their
usage from querying the
INFORMATION_SCHEMA.KEY_COLUMN_USAGE
table, and more information more specific to
InnoDB tables can be found in the
INNODB_SYS_FOREIGN and
INNODB_SYS_FOREIGN_COLS tables, also
in the INFORMATION_SCHEMA database. See also
Section 13.1.14.2, “Using FOREIGN KEY Constraints”.
In addition to SHOW ERRORS, in the
event of a foreign key error involving InnoDB
tables (usually Error 150 in the MySQL Server), you can obtain a
detailed explanation of the most recent InnoDB
foreign key error by checking the output of
SHOW ENGINE INNODB
STATUS.
By using the SQL syntax and MySQL configuration options for compression, you can create tables where the data is stored in compressed form. Compression can help to improve both raw performance and scalability. The compression means less data is transferred between disk and memory, and takes up less space on disk and in memory. The benefits are amplified for tables with secondary indexes, because index data is compressed also. Compression can be especially important for SSD storage devices, because they tend to have lower capacity than HDD devices.
Because processors and cache memories have increased in speed more than disk storage devices, many workloads are disk-bound. Data compression enables smaller database size, reduced I/O, and improved throughput, at the small cost of increased CPU utilization. Compression is especially valuable for read-intensive applications, on systems with enough RAM to keep frequently used data in memory.
An InnoDB table created with
ROW_FORMAT=COMPRESSED can use a smaller
page size on disk than the
usual 16KB default. Smaller pages require less I/O to read from
and write to disk, which is especially valuable for
SSD devices.
The page size is specified through the
KEY_BLOCK_SIZE parameter. The different page
size means the table must be in its own .ibd
file rather than in the
system tablespace,
which requires enabling the
innodb_file_per_table option. The level of
compression is the same regardless of the
KEY_BLOCK_SIZE value. As you specify smaller
values for KEY_BLOCK_SIZE, you get the I/O
benefits of increasingly smaller pages. But if you specify a value
that is too small, there is additional overhead to reorganize the
pages when data values cannot be compressed enough to fit multiple
rows in each page. There is a hard limit on how small
KEY_BLOCK_SIZE can be for a table, based on the
lengths of the key columns for each of its indexes. Specify a
value that is too small, and the CREATE
TABLE or ALTER TABLE
statement fails.
In the buffer pool, the compressed data is held in small pages,
with a page size based on the KEY_BLOCK_SIZE
value. For extracting or updating the column values, MySQL also
creates a 16KB page in the buffer pool with the uncompressed data.
Within the buffer pool, any updates to the uncompressed page are
also re-written back to the equivalent compressed page. You might
need to size your buffer pool to accommodate the additional data
of both compressed and uncompressed pages, although the
uncompressed pages are
evicted from the buffer pool
when space is needed, and then uncompressed again on the next
access.
Before creating a compressed table, make sure the
innodb_file_per_table
configuration option is enabled, and
innodb_file_format is set to
Barracuda. You can set these parameters in the
MySQL configuration
file my.cnf or
my.ini, or with the SET
statement without shutting down the MySQL server.
To enable compression for a table, you use the clauses
ROW_FORMAT=COMPRESSED,
KEY_BLOCK_SIZE, or both in a
CREATE TABLE or
ALTER TABLE statement.
To create a compressed table, you might use statements like these:
SET GLOBAL innodb_file_per_table=1; SET GLOBAL innodb_file_format=Barracuda; CREATE TABLE t1 (c1 INT PRIMARY KEY) ROW_FORMAT=COMPRESSED KEY_BLOCK_SIZE=8;
If you specify
ROW_FORMAT=COMPRESSED, you can omitKEY_BLOCK_SIZE; the default compressed page size of 8KB is used.If you specify
KEY_BLOCK_SIZE, you can omitROW_FORMAT=COMPRESSED; compression is enabled automatically.To determine the best value for
KEY_BLOCK_SIZE, typically you create several copies of the same table with different values for this clause, then measure the size of the resulting.ibdfiles and see how well each performs with a realistic workload.For additional performance-related configuration options, see Section 5.4.6.3, “Tuning Compression for InnoDB Tables”.
The default uncompressed size of InnoDB data
pages is 16KB. Depending on the
combination of option values, MySQL uses a page size of 1KB, 2KB,
4KB, 8KB, or 16KB for the .ibd file of the
table. The actual compression algorithm is not affected by the
KEY_BLOCK_SIZE value; the value determines how
large each compressed chunk is, which in turn affects how many
rows can be packed into each compressed page.
Setting KEY_BLOCK_SIZE=16 typically does not
result in much compression, since the normal InnoDB
page size is 16KB. This
setting may still be useful for tables with many long
BLOB,
VARCHAR or
TEXT columns, because such values
often do compress well, and might therefore require fewer
overflow pages as
described in Section 5.4.6.5, “How Compression Works for InnoDB Tables”.
All indexes of a table (including the
clustered index) are
compressed using the same page size, as specified in the
CREATE TABLE or
ALTER TABLE statement. Table
attributes such as ROW_FORMAT and
KEY_BLOCK_SIZE are not part of the
CREATE INDEX syntax, and are
ignored if they are specified (although you see them in the output
of the SHOW CREATE TABLE statement).
Restrictions on Compressed Tables
Because MySQL versions prior to 5.1 cannot process compressed
tables, using compression requires specifying the configuration
parameter
innodb_file_format=Barracuda, to
avoid accidentally introducing compatibility issues.
Table compression is also not available for the InnoDB
system tablespace.
The system tablespace (space 0, the ibdata*
files) can contain user data, but it also contains internal system
information, and therefore is never compressed. Thus, compression
applies only to tables (and indexes) stored in their own
tablespaces, that is, created with the
innodb_file_per_table option
enabled.
Compression applies to an entire table and all its associated
indexes, not to individual rows, despite the clause name
ROW_FORMAT.
Most often, the internal optimizations described in Section 5.4.6.5, “ InnoDB Data Storage and Compression ” ensure that the system runs well with compressed data. However, because the efficiency of compression depends on the nature of your data, you can make decisions that affect the performance of compressed tables:
Which tables to compress.
What compressed page size to use.
Whether to adjust the size of the buffer pool based on run-time performance characteristics, such as the amount of time the system spends compressing and uncompressing data. Whether the workload is more like a data warehouse (primarily queries) or an OLTP system (mix of queries and DML).
If the system performs DML operations on compressed tables, and the way the data is distributed leads to expensive compression failures at runtime, you might adjust additional advanced configuration options.
Use the guidelines in this section to help make those architectural and configuration choices. When you are ready to conduct long-term testing and put compressed tables into production, see Section 5.4.6.4, “Monitoring Compression at Runtime” for ways to verify the effectiveness of those choices under real-world conditions.
When to Use Compression
In general, compression works best on tables that include a reasonable number of character string columns and where the data is read far more often than it is written. Because there are no guaranteed ways to predict whether or not compression benefits a particular situation, always test with a specific workload and data set running on a representative configuration. Consider the following factors when deciding which tables to compress.
Data Characteristics and Compression
A key determinant of the efficiency of compression in reducing the
size of data files is the nature of the data itself. Recall that
compression works by identifying repeated strings of bytes in a
block of data. Completely randomized data is the worst case.
Typical data often has repeated values, and so compresses
effectively. Character strings often compress well, whether
defined in CHAR, VARCHAR,
TEXT or BLOB columns. On the
other hand, tables containing mostly binary data (integers or
floating point numbers) or data that is previously compressed (for
example JPEG or PNG images)
may not generally compress well, significantly or at all.
You choose whether to turn on compression for each InnoDB table. A
table and all of its indexes use the same (compressed)
page size. It might be that
the primary key
(clustered) index, which contains the data for all columns of a
table, compresses more effectively than the secondary indexes. For
those cases where there are long rows, the use of compression
might result in long column values being stored
“off-page”, as discussed in
Section 5.4.8.3, “DYNAMIC and COMPRESSED Row Formats”. Those overflow pages
may compress well. Given these considerations, for many
applications, some tables compress more effectively than others,
and you might find that your workload performs best only with a
subset of tables compressed.
To determine whether or not to compress a particular table,
conduct experiments. You can get a rough estimate of how
efficiently your data can be compressed by using a utility that
implements LZ77 compression (such as gzip or
WinZip) on a copy of the .ibd
file for an uncompressed table. You can expect less
compression from a MySQL compressed table than from file-based
compression tools, because MySQL compresses data in chunks based
on the page size, 16KB by
default. In addition to user data, the page format includes some
internal system data that is not compressed. File-based
compression utilities can examine much larger chunks of data, and
so might find more repeated strings in a huge file than MySQL can
find in an individual page.
Another way to test compression on a specific table is to copy
some data from your uncompressed table to a similar, compressed
table (having all the same indexes) and look at the size of the
resulting .ibd file. For example:
use test; set global innodb_file_per_table=1; set global innodb_file_format=Barracuda; set global autocommit=0; -- Create an uncompressed table with a million or two rows. create table big_table as select * from information_schema.columns; insert into big_table select * from big_table; insert into big_table select * from big_table; insert into big_table select * from big_table; insert into big_table select * from big_table; insert into big_table select * from big_table; insert into big_table select * from big_table; insert into big_table select * from big_table; insert into big_table select * from big_table; insert into big_table select * from big_table; insert into big_table select * from big_table; commit; alter table big_table add id int unsigned not null primary key auto_increment; show create table big_table\G select count(id) from big_table; -- Check how much space is needed for the uncompressed table. \! ls -l data/test/big_table.ibd create table key_block_size_4 like big_table; alter table key_block_size_4 key_block_size=4 row_format=compressed; insert into key_block_size_4 select * from big_table; commit; -- Check how much space is needed for a compressed table -- with particular compression settings. \! ls -l data/test/key_block_size_4.ibd
This experiment produced the following numbers, which of course could vary considerably depending on your table structure and data:
-rw-rw---- 1 cirrus staff 310378496 Jan 9 13:44 data/test/big_table.ibd -rw-rw---- 1 cirrus staff 83886080 Jan 9 15:10 data/test/key_block_size_4.ibd
To see whether compression is efficient for your particular workload:
For simple tests, use a MySQL instance with no other compressed tables and run queries against the
INFORMATION_SCHEMA.INNODB_CMPtable.For more elaborate tests involving workloads with multiple compressed tables, run queries against the
INFORMATION_SCHEMA.INNODB_CMP_PER_INDEXtable. Because the statistics in theINNODB_CMP_PER_INDEXtable are expensive to collect, you must enable the configuration optioninnodb_cmp_per_index_enabledbefore querying that table, and you might restrict such testing to a development server or a non-critical slave server.Run some typical SQL statements against the compressed table you are testing.
Examine the ratio of successful compression operations to overall compression operations by querying the
INFORMATION_SCHEMA.INNODB_CMPorINFORMATION_SCHEMA.INNODB_CMP_PER_INDEXtable, and comparingCOMPRESS_OPStoCOMPRESS_OPS_OK.If a high percentage of compression operations complete successfully, the table might be a good candidate for compression.
If you get a high proportion of compression failures, you can adjust
innodb_compression_level,innodb_compression_failure_threshold_pct, andinnodb_compression_pad_pct_maxoptions as described in Section 14.2.4.2.2, “Compression Enhancements for OLTP Workloads”, and try further tests.
Database Compression versus Application Compression
Decide whether to compress data in your application or in the table; do not use both types of compression for the same data. When you compress the data in the application and store the results in a compressed table, extra space savings are extremely unlikely, and the double compression just wastes CPU cycles.
Compressing in the Database
When enabled, MySQL table compression is automatic and applies to
all columns and index values. The columns can still be tested with
operators such as LIKE, and sort operations can
still use indexes even when the index values are compressed.
Because indexes are often a significant fraction of the total size
of a database, compression could result in significant savings in
storage, I/O or processor time. The compression and decompression
operations happen on the database server, which likely is a
powerful system that is sized to handle the expected load.
Compressing in the Application
If you compress data such as text in your application, before it is inserted into the database, You might save overhead for data that does not compress well by compressing some columns and not others. This approach uses CPU cycles for compression and uncompression on the client machine rather than the database server, which might be appropriate for a distributed application with many clients, or where the client machine has spare CPU cycles.
Hybrid Approach
Of course, it is possible to combine these approaches. For some applications, it may be appropriate to use some compressed tables and some uncompressed tables. It may be best to externally compress some data (and store it in uncompressed tables) and allow MySQL to compress (some of) the other tables in the application. As always, up-front design and real-life testing are valuable in reaching the right decision.
Workload Characteristics and Compression
In addition to choosing which tables to compress (and the page
size), the workload is another key determinant of performance. If
the application is dominated by reads, rather than updates, fewer
pages need to be reorganized and recompressed after the index page
runs out of room for the per-page “modification log”
that MySQL maintains for compressed data. If the updates
predominantly change non-indexed columns or those containing
BLOBs or large strings that happen to be stored
“off-page”, the overhead of compression may be
acceptable. If the only changes to a table are
INSERTs that use a monotonically increasing
primary key, and there are few secondary indexes, there is little
need to reorganize and recompress index pages. Since MySQL can
“delete-mark” and delete rows on compressed pages
“in place” by modifying uncompressed data,
DELETE operations on a table are relatively
efficient.
For some environments, the time it takes to load data can be as important as run-time retrieval. Especially in data warehouse environments, many tables may be read-only or read-mostly. In those cases, it might or might not be acceptable to pay the price of compression in terms of increased load time, unless the resulting savings in fewer disk reads or in storage cost is significant.
Fundamentally, compression works best when the CPU time is available for compressing and uncompressing data. Thus, if your workload is I/O bound, rather than CPU-bound, you might find that compression can improve overall performance. When you test your application performance with different compression configurations, test on a platform similar to the planned configuration of the production system.
Configuration Characteristics and Compression
Reading and writing database pages from and to disk is the slowest aspect of system performance. Compression attempts to reduce I/O by using CPU time to compress and uncompress data, and is most effective when I/O is a relatively scarce resource compared to processor cycles.
This is often especially the case when running in a multi-user environment with fast, multi-core CPUs. When a page of a compressed table is in memory, MySQL often uses additional memory, typically 16KB, in the buffer pool for an uncompressed copy of the page. The adaptive LRU algorithm attempts to balance the use of memory between compressed and uncompressed pages to take into account whether the workload is running in an I/O-bound or CPU-bound manner. Still, a configuration with more memory dedicated to the buffer pool tends to run better when using compressed tables than a configuration where memory is highly constrained.
Choosing the Compressed Page Size
The optimal setting of the compressed page size depends on the type and distribution of data that the table and its indexes contain. The compressed page size should always be bigger than the maximum record size, or operations may fail as noted in Section 5.4.6.5, “ Compression of B-Tree Pages ”.
Setting the compressed page size too large wastes some space, but the pages do not have to be compressed as often. If the compressed page size is set too small, inserts or updates may require time-consuming recompression, and the B-tree nodes may have to be split more frequently, leading to bigger data files and less efficient indexing.
Typically, you set the compressed page size to 8K or 4K bytes.
Given that the maximum row size for an InnoDB table is around 8K,
KEY_BLOCK_SIZE=8 is usually a safe choice.
Overall application performance, CPU and I/O utilization and the size of disk files are good indicators of how effective compression is for your application. This section builds on the performance tuning advice from Section 5.4.6.3, “Tuning Compression for InnoDB Tables”, and shows how to find problems that might not turn up during initial testing.
To dig deeper into performance considerations for compressed tables, you can monitor compression performance at runtime using the Information Schema tables described in Example 14.2, “Using the Compression Information Schema Tables”. These tables reflect the internal use of memory and the rates of compression used overall.
The INNODB_CMP table reports
information about compression activity for each compressed page
size (KEY_BLOCK_SIZE) in use. The information
in these tables is system-wide: it summarizes the compression
statistics across all compressed tables in your database. You can
use this data to help decide whether or not to compress a table by
examining these tables when no other compressed tables are being
accessed. It involves relatively low overhead on the server, so
you might query it periodically on a production server to check
the overall efficiency of the compression feature.
The INNODB_CMP_PER_INDEX table
reports information about compression activity for individual
tables and indexes. This information is more targeted and more
useful for evaluating compression efficiency and diagnosing
performance issues one table or index at a time. (Because that
each InnoDB table is represented as a clustered
index, MySQL does not make a big distinction between tables and
indexes in this context.) The
INNODB_CMP_PER_INDEX table does
involve substantial overhead, so it is more suitable for
development servers, where you can compare the effects of
different workloads, data,
and compression settings in isolation. To guard against imposing
this monitoring overhead by accident, you must enable the
innodb_cmp_per_index_enabled
configuration option before you can query the
INNODB_CMP_PER_INDEX table.
The key statistics to consider are the number of, and amount of
time spent performing, compression and uncompression operations.
Since MySQL splits B-tree nodes when they are too full to contain
the compressed data following a modification, compare the number
of “successful” compression operations with the
number of such operations overall. Based on the information in the
INNODB_CMP and
INNODB_CMP_PER_INDEX tables and
overall application performance and hardware resource utilization,
you might make changes in your hardware configuration, adjust the
size of the buffer pool, choose a different page size, or select a
different set of tables to compress.
If the amount of CPU time required for compressing and uncompressing is high, changing to faster or multi-core CPUs can help improve performance with the same data, application workload and set of compressed tables. Increasing the size of the buffer pool might also help performance, so that more uncompressed pages can stay in memory, reducing the need to uncompress pages that exist in memory only in compressed form.
A large number of compression operations overall (compared to the
number of INSERT, UPDATE and
DELETE operations in your application and the
size of the database) could indicate that some of your compressed
tables are being updated too heavily for effective compression. If
so, choose a larger page size, or be more selective about which
tables you compress.
If the number of “successful” compression operations
(COMPRESS_OPS_OK) is a high percentage of the
total number of compression operations
(COMPRESS_OPS), then the system is likely
performing well. If the ratio is low, then MySQL is reorganizing,
recompressing, and splitting B-tree nodes more often than is
desirable. In this case, avoid compressing some tables, or
increase KEY_BLOCK_SIZE for some of the
compressed tables. You might turn off compression for tables that
cause the number of “compression failures” in your
application to be more than 1% or 2% of the total. (Such a failure
ratio might be acceptable during a temporary operation such as a
data load).
This section describes some internal implementation details about MySQL compression for InnoDB tables. The information presented here may be helpful in tuning for performance, but is not necessary to know for basic use of compression.
Compression Algorithms
Some operating systems implement compression at the file system level. Files are typically divided into fixed-size blocks that are compressed into variable-size blocks, which easily leads into fragmentation. Every time something inside a block is modified, the whole block is recompressed before it is written to disk. These properties make this compression technique unsuitable for use in an update-intensive database system.
MySQL implements compression with the help of the well-known zlib library, which implements the LZ77 compression algorithm. This compression algorithm is mature, robust, and efficient in both CPU utilization and in reduction of data size. The algorithm is “lossless”, so that the original uncompressed data can always be reconstructed from the compressed form. LZ77 compression works by finding sequences of data that are repeated within the data to be compressed. The patterns of values in your data determine how well it compresses, but typical user data often compresses by 50% or more.
Unlike compression performed by an application, or compression
features of some other database management systems, InnoDB
compression applies both to user data and to indexes. In many
cases, indexes can constitute 40-50% or more of the total database
size, so this difference is significant. When compression is
working well for a data set, the size of the InnoDB data files
(the .idb files) is 25% to 50% of the
uncompressed size or possibly smaller. Depending on the
workload, this smaller
database can in turn lead to a reduction in I/O, and an increase
in throughput, at a modest cost in terms of increased CPU
utilization. You can adjust the balance between compression level
and CPU overhead by modifying the
innodb_compression_level
configuration option.
InnoDB Data Storage and Compression
All user data in InnoDB tables is stored in pages comprising a B-tree index (the clustered index). In some other database systems, this type of index is called an “index-organized table”. Each row in the index node contains the values of the (user-specified or system-generated) primary key and all the other columns of the table.
Secondary indexes in InnoDB tables are also B-trees, containing pairs of values: the index key and a pointer to a row in the clustered index. The pointer is in fact the value of the primary key of the table, which is used to access the clustered index if columns other than the index key and primary key are required. Secondary index records must always fit on a single B-tree page.
The compression of B-tree nodes (of both clustered and secondary
indexes) is handled differently from compression of
overflow pages used to
store long VARCHAR, BLOB, or
TEXT columns, as explained in the following
sections.
Compression of B-Tree Pages
Because they are frequently updated, B-tree pages require special treatment. It is important to minimize the number of times B-tree nodes are split, as well as to minimize the need to uncompress and recompress their content.
One technique MySQL uses is to maintain some system information in the B-tree node in uncompressed form, thus facilitating certain in-place updates. For example, this allows rows to be delete-marked and deleted without any compression operation.
In addition, MySQL attempts to avoid unnecessary uncompression and recompression of index pages when they are changed. Within each B-tree page, the system keeps an uncompressed “modification log” to record changes made to the page. Updates and inserts of small records may be written to this modification log without requiring the entire page to be completely reconstructed.
When the space for the modification log runs out, InnoDB uncompresses the page, applies the changes and recompresses the page. If recompression fails (a situation known as a compression failure), the B-tree nodes are split and the process is repeated until the update or insert succeeds.
To avoid frequent compression failures in write-intensive
workloads, such as for OLTP
applications, MySQL sometimes reserves some empty space (padding)
in the page, so that the modification log fills up sooner and the
page is recompressed while there is still enough room to avoid
splitting it. The amount of padding space left in each page varies
as the system keeps track of the frequency of page splits. On a
busy server doing frequent writes to compressed tables, you can
adjust the
innodb_compression_failure_threshold_pct,
and
innodb_compression_pad_pct_max
configuration options to fine-tune this mechanism.
Generally, MySQL requires that each B-tree page in an InnoDB table
can accommodate at least two records. For compressed tables, this
requirement has been relaxed. Leaf pages of B-tree nodes (whether
of the primary key or secondary indexes) only need to accommodate
one record, but that record must fit, in uncompressed form, in the
per-page modification log. If
innodb_strict_mode is
ON, MySQL checks the maximum row size during
CREATE TABLE or
CREATE INDEX. If the row does not
fit, the following error message is issued: ERROR HY000:
Too big row.
If you create a table when
innodb_strict_mode is OFF, and a
subsequent INSERT or UPDATE
statement attempts to create an index entry that does not fit in
the size of the compressed page, the operation fails with
ERROR 42000: Row size too large. (This error
message does not name the index for which the record is too large,
or mention the length of the index record or the maximum record
size on that particular index page.) To solve this problem,
rebuild the table with ALTER TABLE
and select a larger compressed page size
(KEY_BLOCK_SIZE), shorten any column prefix
indexes, or disable compression entirely with
ROW_FORMAT=DYNAMIC or
ROW_FORMAT=COMPACT.
Compressing BLOB, VARCHAR, and TEXT Columns
In an InnoDB table, BLOB,
VARCHAR, and
TEXT columns that are not part of
the primary key may be stored on separately allocated
overflow pages. We refer
to these columns as off-page
columns. Their values are stored on singly-linked lists of
overflow pages.
For tables created in ROW_FORMAT=DYNAMIC or
ROW_FORMAT=COMPRESSED, the values of
BLOB,
TEXT, or
VARCHAR columns may be stored fully
off-page, depending on their length and the length of the entire
row. For columns that are stored off-page, the clustered index
record only contains 20-byte pointers to the overflow pages, one
per column. Whether any columns are stored off-page depends on the
page size and the total size of the row. When the row is too long
to fit entirely within the page of the clustered index, MySQL
chooses the longest columns for off-page storage until the row
fits on the clustered index page. As noted above, if a row does
not fit by itself on a compressed page, an error occurs.
Tables created in older versions of MySQL use the
Antelope file format, which
supports only ROW_FORMAT=REDUNDANT and
ROW_FORMAT=COMPACT. In these formats, MySQL
stores the first 768 bytes of BLOB,
VARCHAR, and
TEXT columns in the clustered index
record along with the primary key. The 768-byte prefix is followed
by a 20-byte pointer to the overflow pages that contain the rest
of the column value.
When a table is in COMPRESSED format, all data
written to overflow pages is compressed “as is”; that
is, MySQL applies the zlib compression algorithm to the entire
data item. Other than the data, compressed overflow pages contain
an uncompressed header and trailer comprising a page checksum and
a link to the next overflow page, among other things. Therefore,
very significant storage savings can be obtained for longer
BLOB, TEXT, or
VARCHAR columns if the data is highly
compressible, as is often the case with text data. Image data,
such as JPEG, is typically already compressed
and so does not benefit much from being stored in a compressed
table; the double compression can waste CPU cycles for little or
no space savings.
The overflow pages are of the same size as other pages. A row containing ten columns stored off-page occupies ten overflow pages, even if the total length of the columns is only 8K bytes. In an uncompressed table, ten uncompressed overflow pages occupy 160K bytes. In a compressed table with an 8K page size, they occupy only 80K bytes. Thus, it is often more efficient to use compressed table format for tables with long column values.
Using a 16K compressed page size can reduce storage and I/O costs
for BLOB,
VARCHAR, or
TEXT columns, because such data
often compress well, and might therefore require fewer overflow
pages, even though the B-tree nodes themselves take as many pages
as in the uncompressed form.
Compression and the InnoDB Buffer Pool
In a compressed InnoDB table, every compressed page (whether 1K,
2K, 4K or 8K) corresponds to an uncompressed page of 16K bytes (or
a smaller size if
innodb_page_size is set). To
access the data in a page, MySQL reads the compressed page from
disk if it is not already in the
buffer pool, then
uncompresses the page to its original form. This section describes
how InnoDB manages the buffer pool with respect to pages of
compressed tables.
To minimize I/O and to reduce the need to uncompress a page, at times the buffer pool contains both the compressed and uncompressed form of a database page. To make room for other required database pages, MySQL can evict from the buffer pool an uncompressed page, while leaving the compressed page in memory. Or, if a page has not been accessed in a while, the compressed form of the page might be written to disk, to free space for other data. Thus, at any given time, the buffer pool might contain both the compressed and uncompressed forms of the page, or only the compressed form of the page, or neither.
MySQL keeps track of which pages to keep in memory and which to evict using a least-recently-used (LRU) list, so that hot (frequently accessed) data tends to stay in memory. When compressed tables are accessed, MySQL uses an adaptive LRU algorithm to achieve an appropriate balance of compressed and uncompressed pages in memory. This adaptive algorithm is sensitive to whether the system is running in an I/O-bound or CPU-bound manner. The goal is to avoid spending too much processing time uncompressing pages when the CPU is busy, and to avoid doing excess I/O when the CPU has spare cycles that can be used for uncompressing compressed pages (that may already be in memory). When the system is I/O-bound, the algorithm prefers to evict the uncompressed copy of a page rather than both copies, to make more room for other disk pages to become memory resident. When the system is CPU-bound, MySQL prefers to evict both the compressed and uncompressed page, so that more memory can be used for “hot” pages and reducing the need to uncompress data in memory only in compressed form.
Compression and the InnoDB Redo Log Files
Before a compressed page is written to a
data file, MySQL writes a
copy of the page to the redo log (if it has been recompressed
since the last time it was written to the database). This is done
to ensure that redo logs are usable for
crash recovery, even in
the unlikely case that the zlib library is
upgraded and that change introduces a compatibility problem with
the compressed data. Therefore, some increase in the size of
log files, or a need for more
frequent checkpoints, can
be expected when using compression. The amount of increase in the
log file size or checkpoint frequency depends on the number of
times compressed pages are modified in a way that requires
reorganization and recompression.
Note that compressed tables use a different file format for the redo log and the per-table tablespaces than in MySQL 5.1 and earlier. The MySQL Enterprise Backup product supports this latest Barracuda file format for compressed InnoDB tables. The older InnoDB Hot Backup product can only back up tables using the file format Antelope, and thus does not support compressed InnoDB tables.
Specifying ROW_FORMAT=COMPRESSED or
KEY_BLOCK_SIZE in CREATE
TABLE or ALTER TABLE
statements produces the following warnings if the Barracuda file
format is not enabled. You can view them with the SHOW
WARNINGS statement.
| Level | Code | Message |
|---|---|---|
| Warning | 1478 | InnoDB: KEY_BLOCK_SIZE requires
innodb_file_per_table. |
| Warning | 1478 | InnoDB: KEY_BLOCK_SIZE requires innodb_file_format=1 |
| Warning | 1478 | InnoDB: ignoring
KEY_BLOCK_SIZE= |
| Warning | 1478 | InnoDB: ROW_FORMAT=COMPRESSED requires
innodb_file_per_table. |
| Warning | 1478 | InnoDB: assuming ROW_FORMAT=COMPACT. |
Notes:
By default, these messages are only warnings, not errors, and the table is created without compression, as if the options were not specified.
When
innodb_strict_modeis enabled, MySQL generates an error, not a warning, for these cases. The table is not created if the current configuration does not permit using compressed tables.
The “non-strict” behavior lets you import a
mysqldump file into a database that does not
support compressed tables, even if the source database contained
compressed tables. In that case, MySQL creates the table in
ROW_FORMAT=COMPACT instead of preventing the
operation.
To import the dump file into a new database, and have the tables
re-created as they exist in the original database, ensure the
server has the proper settings for the configuration parameters
innodb_file_format and
innodb_file_per_table.
The attribute KEY_BLOCK_SIZE is permitted only
when ROW_FORMAT is specified as
COMPRESSED or is omitted. Specifying a
KEY_BLOCK_SIZE with any other
ROW_FORMAT generates a warning that you can
view with SHOW WARNINGS. However, the table is
non-compressed; the specified KEY_BLOCK_SIZE is
ignored).
| Level | Code | Message |
|---|---|---|
| Warning | 1478 | InnoDB: ignoring KEY_BLOCK_SIZE= |
If you are running with
innodb_strict_mode enabled, the
combination of a KEY_BLOCK_SIZE with any
ROW_FORMAT other than
COMPRESSED generates an error, not a warning,
and the table is not created.
Table 5.5, “Meaning of CREATE TABLE and
ALTER TABLE options”
summarizes how the various options on CREATE
TABLE and ALTER TABLE are
handled.
Table 5.5. Meaning of CREATE TABLE and
ALTER TABLE options
| Option | Usage | Description |
|---|---|---|
ROW_FORMAT=REDUNDANT | Storage format used prior to MySQL 5.0.3 | Less efficient than ROW_FORMAT=COMPACT; for backward
compatibility |
ROW_FORMAT=COMPACT | Default storage format since MySQL 5.0.3 | Stores a prefix of 768 bytes of long column values in the clustered index page, with the remaining bytes stored in an overflow page |
ROW_FORMAT=DYNAMIC | Available only with
innodb_file_format=Barracuda | Store values within the clustered index page if they fit; if not, stores only a 20-byte pointer to an overflow page (no prefix) |
ROW_FORMAT=COMPRESSED | Available only with
innodb_file_format=Barracuda | Compresses the table and indexes using zlib to default compressed page
size of 8K bytes; implies
ROW_FORMAT=DYNAMIC |
KEY_BLOCK_SIZE= | Available only with
innodb_file_format=Barracuda | Specifies compressed page size of 1, 2, 4, 8 or 16 kilobytes; implies
ROW_FORMAT=DYNAMIC and
ROW_FORMAT=COMPRESSED |
Table 5.6, “CREATE/ALTER TABLE Warnings and Errors when InnoDB
Strict Mode is OFF”
summarizes error conditions that occur with certain combinations
of configuration parameters and options on the
CREATE TABLE or
ALTER TABLE statements, and how the
options appear in the output of SHOW TABLE
STATUS.
When innodb_strict_mode is
OFF, MySQL creates or alters the table, but
ignores certain settings as shown below. You can see the warning
messages in the MySQL error log. When
innodb_strict_mode is
ON, these specified combinations of options
generate errors, and the table is not created or altered. To see
the full description of the error condition, issue the
SHOW ERRORS statement: example:
mysql>CREATE TABLE x (id INT PRIMARY KEY, c INT)->ENGINE=INNODB KEY_BLOCK_SIZE=33333;ERROR 1005 (HY000): Can't create table 'test.x' (errno: 1478) mysql>SHOW ERRORS;+-------+------+-------------------------------------------+ | Level | Code | Message | +-------+------+-------------------------------------------+ | Error | 1478 | InnoDB: invalid KEY_BLOCK_SIZE=33333. | | Error | 1005 | Can't create table 'test.x' (errno: 1478) | +-------+------+-------------------------------------------+ 2 rows in set (0.00 sec)
Table 5.6. CREATE/ALTER TABLE Warnings and Errors when InnoDB
Strict Mode is OFF
| Syntax | Warning or Error Condition | Resulting ROW_FORMAT, as shown in SHOW TABLE
STATUS |
|---|---|---|
ROW_FORMAT=REDUNDANT | None | REDUNDANT |
ROW_FORMAT=COMPACT | None | COMPACT |
ROW_FORMAT=COMPRESSED or
ROW_FORMAT=DYNAMIC or
KEY_BLOCK_SIZE is specified | Ignored unless both
innodb_file_format=Barracuda
and innodb_file_per_table
are enabled | COMPACT |
Invalid KEY_BLOCK_SIZE is specified (not 1, 2, 4, 8
or 16) | KEY_BLOCK_SIZE is ignored | the requested one, or COMPACT by default |
ROW_FORMAT=COMPRESSED and valid
KEY_BLOCK_SIZE are specified | None; KEY_BLOCK_SIZE specified is used, not the 8K
default | COMPRESSED |
KEY_BLOCK_SIZE is specified with
REDUNDANT, COMPACT
or DYNAMIC row format | KEY_BLOCK_SIZE is ignored | REDUNDANT, COMPACT or
DYNAMIC |
ROW_FORMAT is not one of
REDUNDANT, COMPACT,
DYNAMIC or
COMPRESSED | Ignored if recognized by the MySQL parser. Otherwise, an error is issued. | COMPACT or N/A |
When innodb_strict_mode is
ON, MySQL rejects invalid
ROW_FORMAT or KEY_BLOCK_SIZE
parameters. For compatibility with earlier versions of MySQL,
strict mode is not enabled by default; instead, MySQL issues
warnings (not errors) for ignored invalid parameters.
Note that it is not possible to see the chosen
KEY_BLOCK_SIZE using SHOW TABLE
STATUS. The statement SHOW CREATE
TABLE displays the KEY_BLOCK_SIZE
(even if it was ignored when creating the table). The real
compressed page size of the table cannot be displayed by MySQL.
As InnoDB evolves, new on-disk data structures are sometimes
required to support new features. Features such as compressed tables
(see Section 5.4.6, “Working with InnoDB Compressed Tables”), and long variable-length
columns stored off-page (see Section 5.4.8, “How InnoDB Stores Variable-Length Columns”)
require data file formats that are not compatible with prior
versions of InnoDB. These features both require use of the new
Barracuda file format.
All other new features are compatible with the original Antelope file format and do not require the Barracuda file format.
This section discusses enabling file formats for new InnoDB tables, verifying compatibility of different file formats between MySQL releases, identifying the file format in use, downgrading the file format, and file format names that may be used in the future.
Named File Formats. InnoDB 1.1 has the idea of a named file format and a configuration parameter to enable the use of features that require use of that format. The new file format is the Barracuda format, and the original InnoDB file format is called Antelope. Compressed tables and the new row format that stores long columns “off-page” require the use of the Barracuda file format or newer. Future versions of InnoDB may introduce a series of file formats, identified with the names of animals, in ascending alphabetic order.
The configuration parameter
innodb_file_format controls
whether such statements as CREATE
TABLE and ALTER TABLE can
be used to create tables that depend on support for the
Barracuda file format.
Although Oracle recommends using the Barracuda format for new tables where practical, in MySQL 5.5 the default file format is still Antelope, for maximum compatibility with replication configurations containing different MySQL releases.
The file format is a dynamic, global parameter that can be
specified in the MySQL option file (my.cnf or
my.ini) or changed with the SET
GLOBAL command.
InnoDB 1.1 incorporates several checks to guard against the possible crashes and data corruptions that might occur if you run an older release of the MySQL server on InnoDB data files using a newer file format. These checks take place when the server is started, and when you first access a table. This section describes these checks, how you can control them, and error and warning conditions that might arise.
Backward Compatibility
Considerations of backward compatibility only apply when using a recent version of InnoDB (the InnoDB Plugin, or MySQL 5.5 and higher with InnoDB 1.1) alongside an older one (MySQL 5.1 or earlier, with the built-in InnoDB rather than the InnoDB Plugin). To minimize the chance of compatibility issues, you can standardize on the InnoDB Plugin for all your MySQL 5.1 and earlier database servers.
In general, a newer version of InnoDB may create a table or index that cannot safely be read or written with a prior version of InnoDB without risk of crashes, hangs, wrong results or corruptions. InnoDB 1.1 includes a mechanism to guard against these conditions, and to help preserve compatibility among database files and versions of InnoDB. This mechanism lets you take advantage of some new features of an InnoDB release (such as performance improvements and bug fixes), and still preserve the option of using your database with a prior version of InnoDB, by preventing accidental use of new features that create downward-incompatible disk files.
If a version of InnoDB supports a particular file format (whether or not that format is the default), you can query and update any table that requires that format or an earlier format. Only the creation of new tables using new features is limited based on the particular file format enabled. Conversely, if a tablespace contains a table or index that uses a file format that is not supported by the currently running software, it cannot be accessed at all, even for read access.
The only way to “downgrade” an InnoDB tablespace to
an earlier file format is to copy the data to a new table, in a
tablespace that uses the earlier format. This can be done with the
ALTER TABLE statement, as described
in Section 5.4.7.4, “Downgrading the File Format”.
The easiest way to determine the file format of an existing InnoDB
tablespace is to examine the properties of the table it contains,
using the SHOW TABLE STATUS command or querying
the table INFORMATION_SCHEMA.TABLES. If the
Row_format of the table is reported as
'Compressed' or 'Dynamic',
the tablespace containing the table uses the Barracuda format.
Otherwise, it uses the prior InnoDB file format, Antelope.
Internal Details
Every InnoDB per-table tablespace (represented by a
*.ibd file) file is labeled with a file format
identifier. The system tablespace (represented by the
ibdata files) is tagged with the
“highest” file format in use in a group of InnoDB
database files, and this tag is checked when the files are opened.
Creating a compressed table, or a table with
ROW_FORMAT=DYNAMIC, updates the file header for
the corresponding .ibd file and the table type
in the InnoDB data dictionary with the identifier for the
Barracuda file format. From that point forward, the table cannot
be used with a version of InnoDB that does not support this new
file format. To protect against anomalous behavior, InnoDB version
5.0.21 and later performs a compatibility check when the table is
opened. (In many cases, the ALTER
TABLE statement recreates a table and thus changes its
properties. The special case of adding or dropping indexes without
rebuilding the table is described in
Fast Index Creation in the InnoDB Storage Engine.)
Definition of ib-file set
To avoid confusion, for the purposes of this discussion we define the term “ib-file set” to mean the set of operating system files that InnoDB manages as a unit. The ib-file set includes the following files:
The system tablespace (one or more
ibdatafiles) that contain internal system information (including internal catalogs and undo information) and may include user data and indexes.Zero or more single-table tablespaces (also called “file per table” files, named
*.ibdfiles).InnoDB log files; usually two,
ib_logfile0andib_logfile1. Used for crash recovery and in backups.
An “ib-file set” does not include the corresponding
.frm files that contain metadata about InnoDB
tables. The .frm files are created and managed
by MySQL, and can sometimes get out of sync with the internal
metadata in InnoDB.
Multiple tables, even from more than one database, can be stored in a single “ib-file set”. (In MySQL, a “database” is a logical collection of tables, what other systems refer to as a “schema” or “catalog”.)
To prevent possible crashes or data corruptions when InnoDB
opens an ib-file set, it checks that it can fully support the
file formats in use within the ib-file set. If the system is
restarted following a crash, or a “fast shutdown”
(i.e., innodb_fast_shutdown is
greater than zero), there may be on-disk data structures (such
as redo or undo entries, or doublewrite pages) that are in a
“too-new” format for the current software. During
the recovery process, serious damage can be done to your data
files if these data structures are accessed. The startup check
of the file format occurs before any recovery process begins,
thereby preventing consistency issues with the new tables or
startup problems for the MySQL server.
Beginning with version InnoDB 1.0.1, the system tablespace
records an identifier or tag for the “highest” file
format used by any table in any of the tablespaces that is part
of the ib-file set. Checks against this file format tag are
controlled by the configuration parameter
innodb_file_format_check, which
is ON by default.
If the file format tag in the system tablespace is newer or
higher than the highest version supported by the particular
currently executing software and if
innodb_file_format_check is
ON, the following error is issued when the
server is started:
InnoDB: Error: the system tablespace is in a file format that this version doesn't support
You can also set
innodb_file_format to a file
format name. Doing so prevents InnoDB from starting if the
current software does not support the file format specified. It
also sets the “high water mark” to the value you
specify. The ability to set
innodb_file_format_check will
be useful (with future releases of InnoDB) if you manually
“downgrade” all of the tables in an ib-file set (as
described in Downgrading the InnoDB Storage Engine). You can then
rely on the file format check at startup if you subsequently use
an older version of InnoDB to access the ib-file set.
In some limited circumstances, you might want to start the
server and use an ib-file set that is in a “too
new” format (one that is not supported by the software
you are using). If you set the configuration parameter
innodb_file_format_check to
OFF, InnoDB opens the database, but issues
this warning message in the error log:
InnoDB: Warning: the system tablespace is in a file format that this version doesn't support
This is a very dangerous setting, as it permits the recovery
process to run, possibly corrupting your database if the
previous shutdown was a crash or “fast shutdown”.
You should only set
innodb_file_format_check to
OFF if you are sure that the previous
shutdown was done with
innodb_fast_shutdown=0, so that essentially
no recovery process occurs. In a future release, this
parameter setting may be renamed from OFF
to UNSAFE. (However, until there are newer
releases of InnoDB that support additional file formats, even
disabling the startup checking is in fact
“safe”.)
The parameter
innodb_file_format_check
affects only what happens when a database is opened, not
subsequently. Conversely, the parameter
innodb_file_format (which
enables a specific format) only determines whether or not a new
table can be created in the enabled format and has no effect on
whether or not a database can be opened.
The file format tag is a “high water mark”, and as
such it is increased after the server is started, if a table in
a “higher” format is created or an existing table
is accessed for read or write (assuming its format is
supported). If you access an existing table in a format higher
than the format the running software supports, the system
tablespace tag is not updated, but table-level compatibility
checking applies (and an error is issued), as described in
Section 5.4.7.2.2, “Compatibility Check When a Table Is Opened”.
Any time the high water mark is updated, the value of
innodb_file_format_check is
updated as well, so the command SELECT
@@innodb_file_format_check; displays the name of the
newest file format known to be used by tables in the currently
open ib-file set and supported by the currently executing
software.
To best illustrate this behavior, consider the scenario described in Table 5.7, “InnoDB Data File Compatibility and Related InnoDB Parameters”. Imagine that some future version of InnoDB supports the Cheetah format and that an ib-file set has been used with that version.
Table 5.7. InnoDB Data File Compatibility and Related InnoDB Parameters
| innodb file format check | innodb file format | Highest file format used in ib-file set | Highest file format supported by InnoDB | Result |
|---|---|---|---|---|
OFF | Antelope or Barracuda | Barracuda | Barracuda | Database can be opened; tables can be created which require Antelope or Barracuda file format |
OFF | Antelope or Barracuda | Cheetah | Barracuda | Database can be opened with a warning, since the database contains files in a “too new” format; tables can be created in Antelope or Barracuda file format; tables in Cheetah format cannot be accessed |
OFF | Cheetah | Barracuda | Barracuda | Database cannot be opened;
innodb_file_format
cannot be set to Cheetah |
ON | Antelope or Barracuda | Barracuda | Barracuda | Database can be opened; tables can be created in Antelope or Barracuda file format |
ON | Antelope or Barracuda | Cheetah | Barracuda | Database cannot be opened, since the database contains files in a “too new” format (Cheetah) |
ON | Cheetah | Barracuda | Barracuda | Database cannot be opened;
innodb_file_format
cannot be set to Cheetah |
When a table is first accessed, InnoDB (including some releases prior to InnoDB 1.0) checks that the file format of the tablespace in which the table is stored is fully supported. This check prevents crashes or corruptions that would otherwise occur when tables using a “too new” data structure are encountered.
All tables using any file format supported by a release can be
read or written (assuming the user has sufficient privileges).
The setting of the system configuration parameter
innodb_file_format can prevent
creating a new table that uses specific file formats, even if
they are supported by a given release. Such a setting might be
used to preserve backward compatibility, but it does not prevent
accessing any table that uses any supported format.
As noted in Named File Formats, versions of MySQL older than 5.0.21 cannot reliably use database files created by newer versions if a new file format was used when a table was created. To prevent various error conditions or corruptions, InnoDB checks file format compatibility when it opens a file (for example, upon first access to a table). If the currently running version of InnoDB does not support the file format identified by the table type in the InnoDB data dictionary, MySQL reports the following error:
ERROR 1146 (42S02): Table 'test.t1' doesn't exist
InnoDB also writes a message to the error log:
InnoDB: tabletest/t1: unknown table type33
The table type should be equal to the tablespace flags, which contains the file format version as discussed in Section 5.4.7.3, “Identifying the File Format in Use”.
Versions of InnoDB prior to MySQL 4.1 did not include table format identifiers in the database files, and versions prior to MySQL 5.0.21 did not include a table format compatibility check. Therefore, there is no way to ensure proper operations if a table in a “too new” format is used with versions of InnoDB prior to 5.0.21.
The file format management capability in InnoDB 1.0 and higher (tablespace tagging and run-time checks) allows InnoDB to verify as soon as possible that the running version of software can properly process the tables existing in the database.
If you permit InnoDB to open a database containing files in a
format it does not support (by setting the parameter
innodb_file_format_check to
OFF), the table-level checking described in
this section still applies.
Users are strongly urged not to use database files that contain Barracuda file format tables with releases of InnoDB older than the MySQL 5.1 with the InnoDB Plugin. It is possible to “downgrade” such tables to the Antelope format with the procedure described in Section 5.4.7.4, “Downgrading the File Format”.
After you enable a given
innodb_file_format, this change
applies only to newly created tables rather than existing ones. If
you do create a new table, the tablespace containing the table is
tagged with the “earliest” or “simplest”
file format that is required for the table's features. For
example, if you enable file format Barracuda, and create a new
table that is not compressed and does not use
ROW_FORMAT=DYNAMIC, the new tablespace that
contains the table is tagged as using file format Antelope.
It is easy to identify the file format used by a given tablespace
or table. The table uses the Barracuda format if the
Row_format reported by SHOW CREATE
TABLE or INFORMATION_SCHEMA.TABLES is
one of 'Compressed' or
'Dynamic'. (The Row_format
is a separate column; ignore the contents of the
Create_options column, which may contain the
string ROW_FORMAT.) If the table in a
tablespace uses neither of those features, the file uses the
format supported by prior releases of InnoDB, now called file
format Antelope. Then, the Row_format is one of
'Redundant' or 'Compact'.
Internal Details
InnoDB has two different file formats (Antelope and Barracuda) and four different row formats (Redundant, Compact, Dynamic, and Compressed). The Antelope file format contains Redundant and Compact row formats. A tablespace that uses the Barracuda file format uses either the Dynamic or Compressed row format.
File and row format information is written in the tablespace flags
(a 32-bit number) in the *.ibd file in the 4
bytes starting at position 54 of the file, most significant byte
first (the first byte of the file is byte zero). On some systems,
you can display these bytes in hexadecimal with the command
od -t x1 -j 54 -N 4
. If all bytes
are zero, the tablespace uses the Antelope file format, which is
the format used by the standard InnoDB storage engine up to
version 5.1. The system tablespace will always have zero in the
tablespace flags.
tablename.ibd
The first 10 bits of the tablespace flags can be described this way:
Bit 0: Zero for Antelope, and bits 1 to 5 will also be zero. One for Barracuda, and bits 1 to 5 may be set.
Bits 1 to 4: A 4 bit number representing the compressed page size. 0 = not compressed, 1 = 1k, 2 = 2k, 3 = 4k, 4 = 8k.
Bit 5: Same value as Bit 0, zero for Antelope, one for Barracuda. If bits 0 and 5 are set and bits 1 to 4 are not, the row format is Dynamic.
Bits 6 to 9: A 4-bit number indicating the physical page size of the tablespace. 0=16k (original default), 3=4k, 4=8k, 5=16k. These are the only valid values for My SQL 5.6 and later.
Bit 10: Tablespace location. 0 = default, 1 = used
DATA DIRECTORYinCREATE TABLEto choose the tablespace location.
Tablespace flags are similar to table flags found in the InnoDB
dictionary table, “SYS_TABLES”.
They differ in the meaning of bit 0 and bits 6 to 10. Table
flags will set bit 0 to one if the row format of a particular
table is “Compact”. Tablespace flags cannot do that
since the system tablespace can contain both Redundant and
Compact row formats. So, for tablespace flags, bit 0 and bit 5
are always the same value.
Table flags can be viewed by issuing the command:
SELECT * FROM INFORMATION_SCHEMA.INNODB_SYS_TABLES;
The first 7 bits of the table flags can be described this way:
Bit 0: Zero for Redundant row format, and bits 1 to 5 will be zero. One for Compact row format, and bits 1 to 5 may be set.
Bits 1 to 4: A 4 bit number representing the compressed page size. 0 = not compressed, 1 = 1k, 2 = 2k, 3 = 4k, 4 = 8k.
Bit 5: Zero for Antelope file format, and one for Barracuda file format. If bit 5 is set and bits 1 to 4 are not, the row format is Dynamic. Also, if bit 5 is set, bit 0 must also be set.
Bit 6: Tablespace location. 0 = default, 1 =
DATA DIRECTORYwas used inCREATE TABLEto choose a tablespace location.
If bits 7 to 31 are not zero, the table is corrupt or the
SYS_TABLES record is corrupt, and the table
cannot be used.
Each InnoDB tablespace file (with a name matching
*.ibd) is tagged with the file format used to
create its table and indexes. The way to downgrade the tablespace
is to re-create the table and its indexes. The easiest way to
recreate a table and its indexes is to use the command:
ALTER TABLEtROW_FORMAT=COMPACT;
on each table that you want to downgrade. The
COMPACT row format uses the file format
Antelope. It was introduced in MySQL 5.0.3.
The file format used by the standard built-in InnoDB in MySQL 5.1 is the Antelope format. The file format introduced with InnoDB Plugin 1.0 is the Barracuda format. Although no new features have been announced that would require additional new file formats, the InnoDB file format mechanism allows for future enhancements.
For the sake of completeness, these are the file format names that might be used for future file formats: Antelope, Barracuda, Cheetah, Dragon, Elk, Fox, Gazelle, Hornet, Impala, Jaguar, Kangaroo, Leopard, Moose, Nautilus, Ocelot, Porpoise, Quail, Rabbit, Shark, Tiger, Urchin, Viper, Whale, Xenops, Yak and Zebra. These file formats correspond to the internal identifiers 0..25.
This section discusses how certain InnoDB features, such as table
compression and off-page
storage of long columns, are controlled by the
ROW_FORMAT clause of the
CREATE TABLE statement. It discusses
considerations for choosing the right row format and compatibility
of row formats between MySQL releases.
The storage for rows and associated columns affects performance for queries and DML operations. As more rows fit into a single disk page, queries and index lookups can work faster, less cache memory is required in the InnoDB buffer pool, and less I/O is required to write out updated values for the numeric and short string columns.
The data in each InnoDB table is divided into pages. The pages that make up each table are arranged in a tree data structure called a B-tree index. Table data and secondary indexes both use this type of structure. The B-tree index that represents an entire table is known as the clustered index, which is organized according to the primary key columns. The nodes of the index data structure contain the values of all the columns in that row (for the clustered index) or the index columns and the primary key columns (for secondary indexes).
Variable-length columns are an exception to this rule. Columns
such as BLOB and VARCHAR
that are too long to fit on a B-tree page are stored on separately
allocated disk pages called
overflow pages. We call
such columns off-page
columns. The values of these columns are stored in
singly-linked lists of overflow pages, and each such column has
its own list of one or more overflow pages. In some cases, all or
a prefix of the long column value is stored in the B-tree, to
avoid wasting storage and eliminating the need to read a separate
page.
This section describes the clauses you can use with the
CREATE TABLE and
ALTER TABLE statements to control
how these variable-length columns are represented:
ROW_FORMAT and
KEY_BLOCK_SIZE. To use these clauses, you might
also need to change the settings for the
innodb_file_per_table and
innodb_file_format configuration
options.
You specify the row format for a table with the
ROW_FORMAT clause of the
CREATE TABLE and
ALTER TABLE statements.
This section discusses the DYNAMIC and
COMPRESSED row formats for InnoDB tables. You
can only create these kinds of tables when the
innodb_file_format configuration
option is set to Barracuda. (The Barracuda file
format also allows the COMPACT and
REDUNDANT row formats.)
When a table is created with ROW_FORMAT=DYNAMIC
or ROW_FORMAT=COMPRESSED, long column values
are stored fully off-page, and the clustered index record contains
only a 20-byte pointer to the overflow page.
Whether any columns are stored off-page depends on the page size and the total size of the row. When the row is too long, InnoDB chooses the longest columns for off-page storage until the clustered index record fits on the B-tree page.
The DYNAMIC row format maintains the efficiency
of storing the entire row in the index node if it fits (as do the
COMPACT and REDUNDANT
formats), but this new format avoids the problem of filling B-tree
nodes with a large number of data bytes of long columns. The
DYNAMIC format is based on the idea that if a
portion of a long data value is stored off-page, it is usually
most efficient to store all of the value off-page. With
DYNAMIC format, shorter columns are likely to
remain in the B-tree node, minimizing the number of overflow pages
needed for any given row.
The COMPRESSED row format uses similar internal
details for off-page storage as the DYNAMIC row
format, with additional storage and performance considerations
from the table and index data being compressed and using smaller
page sizes. With the COMPRESSED row format, the
option KEY_BLOCK_SIZE controls how much column
data is stored in the clustered index, and how much is placed on
overflow pages. For full details about the
COMPRESSED row format, see
Section 5.4.6, “Working with InnoDB Compressed Tables”.
Early versions of InnoDB used an unnamed file format (now called
Antelope) for database files.
With that file format, tables are defined with
ROW_FORMAT=COMPACT or
ROW_FORMAT=REDUNDANT. InnoDB stores up to the
first 768 bytes of variable-length columns (such as
BLOB and VARCHAR) in the
index record within the B-tree node, with the remainder stored on
the overflow pages.
To preserve compatibility with those prior versions, tables
created with the newest InnoDB default to the
COMPACT row format. See
Section 5.4.8.3, “DYNAMIC and COMPRESSED Row Formats” for information about
the newer DYNAMIC and
COMPRESSED row formats.
With the Antelope file format, if the value of a column is 768
bytes or less, no overflow page is needed, and some savings in I/O
may result, since the value is in the B-tree node. This works well
for relatively short BLOBs, but may cause
B-tree nodes to fill with data rather than key values, reducing
their efficiency. Tables with many BLOB columns
could cause B-tree nodes to become too full of data, and contain
too few rows, making the entire index less efficient than if the
rows were shorter or if the column values were stored off-page.
- 5.5.1. Overview of Online DDL
- 5.5.2. Performance and Concurrency Considerations for Online DDL
- 5.5.3. SQL Syntax for Online DDL
- 5.5.4. Combining or Separating DDL Statements
- 5.5.5. Examples of Online DDL
- 5.5.6. Implementation Details of Online DDL
- 5.5.7. How Crash Recovery Works with Online DDL
- 5.5.8. Online DDL for Partitioned
InnoDBTables - 5.5.9. Limitations of Online DDL
You can perform several kinds of
online DDL operations on
InnoDB tables: that is, allowing
DML operations and
queries on the table while the
DDL is in progress, performing the
operation “in-place” without rebuilding the entire
table, or both. This enhancement has the following benefits:
It improves responsiveness and availability in busy production environments, where making a table unavailable for minutes or hours whenever you modify its indexes or column definitions is not practical.
It lets you adjust the balance between performance and concurrency during the DDL operation, by choosing whether to block access to the table entirely (
LOCK=EXCLUSIVEclause), allow queries but not DML (LOCK=SHAREDclause), or allow full query and DML access to the table (LOCK=NONEclause). When you omit theLOCKclause or specifyLOCK=DEFAULT, MySQL allows as much concurrency as possible depending on the type of operation.By doing the changes in-place where possible, rather than creating a new copy of the table, it avoids temporary increases in disk space usage and the I/O overhead of copying the table and reconstructing all the secondary indexes.
Historically, many DDL operations
on InnoDB tables were expensive. Many
ALTER TABLE operations worked by
creating a new, empty table defined with the requested table
options and indexes, then copying the existing rows to the new
table one-by-one, updating the indexes as the rows were inserted.
After all rows from the original table were copied, the old table
was dropped and the copy was renamed with the name of the original
table.
MySQL 5.5, and MySQL 5.1 with the InnoDB Plugin, optimized
CREATE INDEX and DROP INDEX
to avoid the table-copying behavior. That feature was known as
Fast Index Creation. MySQL 5.6 enhances many other types of
ALTER TABLE operations to avoid
copying the table. Another enhancement allows
SELECT queries and
INSERT,
UPDATE, and
DELETE
(DML) statements to proceed while
the table is being altered. In MySQL 5.7,
ALTER TABLE RENAME
INDEX was also enhanced to avoid table copying. This
combination of features is now known as
online DDL.
This new mechanism also means that you can generally speed the overall process of creating and loading a table and associated indexes by creating the table with without any secondary indexes, then adding the secondary indexes after the data is loaded.
Although no syntax changes are required in the
CREATE INDEX or
DROP INDEX commands, some factors
affect the performance, space usage, and semantics of this
operation (see Section 5.5.9, “Limitations of Online DDL”).
The online DDL enhancements in MySQL 5.6 improve many DDL
operations that formerly required a table copy, blocked DML
operations on the table, or both.
Table 5.8, “Summary of Online Status for DDL Operations” shows the
variations of the ALTER TABLE
statement and shows how the online DDL feature applies to each
one.
With the exception of ALTER TABLE
partitioning clauses, online DDL operations for partitioned
InnoDB tables follow the same rules that apply
to regular InnoDB tables. For more information,
see Section 5.5.8, “Online DDL for Partitioned InnoDB Tables”.
The “In-Place?” column shows which operations allow the
ALGORITHM=INPLACEclause; the preferred value is “Yes”.The “Copies Table?” column shows which operations are able to avoid the expensive table-copying operation; the preferred value is “No”. This column is mostly the reverse of the “In-Place?” column, except that a few operations allow
ALGORITHM=INPLACEbut still involve some amount of table copying.The “Allows Concurrent DML?” column shows which operations can be performed fully online; the preferred value is “Yes”. You can specify
LOCK=NONEto assert that full concurrency is allowed during the DDL, but MySQL automatically allows this level of concurrency when possible. When concurrent DML is allowed, concurrent queries are also always allowed.The “Allows Concurrent Queries?” column shows which DDL operations allow queries on the table while the operation is in progress; the preferred value is “Yes”. Concurrent query is allowed during all online DDL operations. It is shown with “Yes” listed for all cells, for reference purposes. You can specify
LOCK=SHAREDto assert that concurrent queries are allowed during the DDL, but MySQL automatically allows this level of concurrency when possible.The “Notes” column explains any exceptions to the “yes/no” values of the other columns, such as when the answer depends on the setting of a configuration option or some other clause in the DDL statement. The values “Yes*” and “No*” indicate that an answer depends on these additional notes.
Table 5.8. Summary of Online Status for DDL Operations
| Operation | In-Place? | Copies Table? | Allows Concurrent DML? | Allows Concurrent Query? | Notes |
|---|---|---|---|---|---|
CREATE INDEX, ADD INDEX | Yes* | No* | Yes | Yes | Some restrictions for FULLTEXT index; see next row.
Currently, the operation is not in-place (that is, it
copies the table) if the same index being created was also
dropped by an earlier clause in the same
ALTER TABLE statement. |
ADD FULLTEXT INDEX | Yes | No* | No | Yes | Creating the first FULLTEXT index for a table
involves a table copy, unless there is a user-supplied
FTS_DOC_ID column. Subsequent
FULLTEXT indexes on the same table can
be created in-place. |
RENAME INDEX | Yes | No | No | No | |
DROP INDEX | Yes | No | Yes | Yes | |
| Set default value for a column | Yes | No | Yes | Yes | Modifies .frm file only, not the data file. |
| Change auto-increment value for a column | Yes | No | Yes | Yes | Modifies a value stored in memory, not the data file. |
| Add a foreign key constraint | Yes* | No* | Yes | Yes | To avoid copying the table, disable
foreign_key_checks during
constraint creation. |
| Drop a foreign key constraint | Yes | No | Yes | Yes | The foreign_key_checks option can be
enabled or disabled. |
| Rename a column | Yes* | No* | Yes* | Yes | To allow concurrent DML, keep the same data type and only change the column name. |
| Add a column | Yes | Yes | Yes* | Yes | Concurrent DML is not allowed when adding an
auto-increment
column. Although ALGORITHM=INPLACE is
allowed, the data is reorganized substantially, so it is
still an expensive operation. |
| Drop a column | Yes | Yes | Yes | Yes | Although ALGORITHM=INPLACE is allowed, the data is
reorganized substantially, so it is still an expensive
operation. |
| Reorder columns | Yes | Yes | Yes | Yes | Although ALGORITHM=INPLACE is allowed, the data is
reorganized substantially, so it is still an expensive
operation. |
Change ROW_FORMAT property | Yes | Yes | Yes | Yes | Although ALGORITHM=INPLACE is allowed, the data is
reorganized substantially, so it is still an expensive
operation. |
Change KEY_BLOCK_SIZE property | Yes | Yes | Yes | Yes | Although ALGORITHM=INPLACE is allowed, the data is
reorganized substantially, so it is still an expensive
operation. |
Make column NULL | Yes | Yes | Yes | Yes | Although ALGORITHM=INPLACE is allowed, the data is
reorganized substantially, so it is still an expensive
operation. |
Make column NOT NULL | Yes* | Yes | Yes | Yes | When SQL_MODE includes
strict_all_tables or
strict_all_tables, the operation fails
if the column contains any nulls. Although
ALGORITHM=INPLACE is allowed, the data
is reorganized substantially, so it is still an expensive
operation. |
| Change data type of column | No* | Yes* | No | Yes | Exception: VARCHAR size may be increased
using online ALTER TABLE.
See
Section 14.2.5.12, “Increase VARCHAR Size Online”. |
| Add primary key | Yes* | Yes | Yes | Yes | Although ALGORITHM=INPLACE is allowed, the data is
reorganized substantially, so it is still an expensive
operation. ALGORITHM=INPLACE is not
allowed under certain conditions if columns have to be
converted to NOT NULL. See
Example 5.9, “Creating and Dropping the Primary Key”. |
| Drop primary key and add another | Yes | Yes | Yes | Yes | ALGORITHM=INPLACE is only allowed when you add a new
primary key in the same ALTER
TABLE; the data is reorganized substantially, so
it is still an expensive operation. |
| Drop primary key | No | Yes | No | Yes | Restrictions apply when you drop a primary key primary key without
adding a new one in the same ALTER
TABLE statement. |
| Convert character set | No | Yes | No | Yes | Rebuilds the table if the new character encoding is different. |
| Specify character set | No | Yes | No | Yes | Rebuilds the table if the new character encoding is different. |
Rebuild with FORCE option | No | Yes | No | Yes | Acts like the ALGORITHM=COPY clause or the setting
old_alter_table=1. |
The following sections shows the basic syntax, and usage notes related to online DDL, for each of the major operations that can be performed with concurrent DML, in-place, or both:
Secondary Indexes
Create secondary indexes:
CREATE INDEXornameONtable(col_list)ALTER TABLE. (Creating a atableADD INDEXname(col_list)FULLTEXTindex still requires locking the table.)Drop secondary indexes:
DROP INDEXornameONtable;ALTER TABLEtableDROP INDEXname
Creating and dropping secondary indexes on
InnoDB tables skips the table-copying behavior,
the same as in MySQL 5.5 and MySQL 5.1 with the
InnoDB Plugin.
In MySQL 5.6 and higher, the table remains available for read and
write operations while the index is being created or dropped. The
CREATE INDEX or
DROP INDEX statement only finishes
after all transactions that are accessing the table are completed,
so that the initial state of the index reflects the most recent
contents of the table. Previously, modifying the table while an
index was being created or dropped typically resulted in a
deadlock that cancelled the
INSERT,
UPDATE, or
DELETE statement on the table.
Column Properties
Set a default value for a column:
ALTER TABLEortblALTER COLUMNcolSET DEFAULTliteralALTER TABLEtblALTER COLUMNcolDROP DEFAULTThe default values for columns are stored in the .frm file for the table, not the
InnoDBdata dictionary.Changing the auto-increment value for a column:
ALTER TABLEtableAUTO_INCREMENT=next_value;Especially in a distributed system using replication or sharding, you sometimes reset the auto-increment counter for a table to a specific value. The next row inserted into the table uses the specified value for its auto-increment column. You might also use this technique in a data warehousing environment where you periodically empty all the tables and reload them, and you can restart the auto-increment sequence from 1.
Renaming a column:
ALTER TABLEtblCHANGEold_col_namenew_col_namedatatypeWhen you keep the same data type and
[NOT] NULLattribute, only changing the column name, this operation can always be performed online.As part of this enhancement, you can now rename a column that is part of a foreign key constraint, which was not allowed before. The foreign key definition is automatically updated to use the new column name. Renaming a column participating in a foreign key only works with the in-place mode of
ALTER TABLE. If you use theALGORITHM=COPYclause, or some other condition causes the command to useALGORITHM=COPYbehind the scenes, theALTER TABLEstatement will fail.Extending
VARCHARsize using an in-placeALTER TABLEstatement, as in this example:ALTER TABLE t1 ALGORITHM=INPLACE, CHANGE COLUMN c1 c1 VARCHAR(255);
The number of length bytes required by a
VARCHARcolumn must remain the same. ForVARCHARvalues of 0 to 255, one length byte is required to encode the value. ForVARCHARvalues of 256 bytes or more, two length bytes are required. As a result, in-placeALTER TABLEonly supports increasingVARCHARsize from 0 to 255 bytes or increasingVARCHARsize from a value equal to or greater than 256 bytes. In-placeALTER TABLEdoes not support increasingVARCHARsize from less than 256 bytes to a value equal to or greater than 256 bytes. In this case, the number of required length bytes would change from 1 to 2, which is only supported by a table copy (ALGORITHM=COPY).Decreasing
VARCHARsize using in-placeALTER TABLEis not supported. DecreasingVARCHARsize requires a table copy (ALGORITHM=COPY).
Foreign Keys
Adding or dropping a foreign key constraint:
ALTER TABLE
tbl1ADD CONSTRAINTfk_nameFOREIGN KEYindex(col1) REFERENCEStbl2(col2)referential_actions; ALTER TABLEtblDROP FOREIGN KEYfk_name;Dropping a foreign key can be performed online with the
foreign_key_checksoption enabled or disabled. Creating a foreign key online requiresforeign_key_checksto be disabled.If you do not know the names of the foreign key constraints on a particular table, issue the following statement and find the constraint name in the
CONSTRAINTclause for each foreign key:show create table
table\GOr, query the
information_schema.table_constraintstable and use theconstraint_nameandconstraint_typecolumns to identify the foreign key names.As a consequence of this enhancement, you can now also drop a foreign key and its associated index in a single statement, which previously required separate statements in a strict order:
ALTER TABLE
tableDROP FOREIGN KEYconstraint, DROP INDEXindex;
If foreign keys are
already present in the table being altered (that is, it is a
child table containing any
FOREIGN KEY ... REFERENCE clauses), additional
restrictions apply to online DDL operations, even those not
directly involving the foreign key columns:
Concurrent DML is disallowed during online DDL operations on such child tables. (This restriction is being evaluated as a bug and might be lifted.)
An
ALTER TABLEon the child table could also wait for another transaction to commit, if a change to the parent table caused associated changes in the child table through anON UPDATEorON DELETEclause using theCASCADEorSET NULLparameters.
In the same way, if a table is the
parent table in a foreign
key relationship, even though it does not contain any
FOREIGN KEY clauses, it could wait for the
ALTER TABLE to complete if an
INSERT,
UPDATE, or
DELETE statement caused an
ON UPDATE or ON DELETE
action in the child table.
Notes on ALGORITHM=COPY
Any ALTER TABLE operation run with
the ALGORITHM=COPY clause prevents concurrent
DML operations. Concurrent queries are still allowed. That is, a
table-copying operation always includes at least the concurrency
restrictions of LOCK=SHARED (allow queries but
not DML). You can further restrict concurrency for such operations
by specifying LOCK=EXCLUSIVE (prevent DML and
queries).
Concurrent DML but Table Copy Still Required
Some other ALTER TABLE operations
allow concurrent DML, and are faster than MySQL 5.5 and prior: the
table-copying operation is optimized, even though a table copy is
still required:
Adding, dropping, or reordering columns.
Adding or dropping a primary key.
Changing the
ROW_FORMATorKEY_BLOCK_SIZEproperties for a table.Changing the nullable status for a column.
As your database schema evolves with new columns, data types,
constraints, indexes, and so on, keep your
CREATE TABLE statements up to
date with the latest table definitions. Even with the
performance improvements of online DDL, it is more efficient to
create stable database structures at the beginning, rather than
creating part of the schema and then issuing
ALTER TABLE statements afterward.
The main exception to this guideline is for secondary indexes on tables with large numbers of rows. It is typically most efficient to create the table with all details specified except the secondary indexes, load the data, then create the secondary indexes. You can use the same technique with foreign keys (load the data first, then set up the foreign keys) if you know the initial data is clean and do not need consistency checks during the loading process.
Whatever sequence of CREATE
TABLE, CREATE INDEX,
ALTER TABLE, and similar
statements went into putting a table together, you can capture
the SQL needed to reconstruct the current form of the table by
issuing the statement SHOW CREATE TABLE
(uppercase
table\G\G required for tidy formatting). This output
shows clauses such as numeric precision, NOT
NULL, and CHARACTER SET that are
sometimes added behind the scenes, and you might otherwise leave
out when cloning the table on a new system or setting up foreign
key columns with identical type.
Online DDL improves several aspects of MySQL operation, such as performance, concurrency, availability, and scalability:
Because queries and DML operations on the table can proceed while the DDL is in progress, applications that access the table are more responsive. Reduced locking and waiting for other resources all throughout the MySQL server leads to greater scalability, even for operations not involving the table being altered.
For in-place operations, by avoiding the disk I/O and CPU cycles to rebuild the table, you minimize the overall load on the database and maintain good performance and high throughput during the DDL operation.
For in-place operations, because less data is read into the buffer pool than if all the data was copied, you avoid purging frequently accessed data from memory, which formerly could cause a temporary performance dip after a DDL operation.
If an online operation requires temporary files,
InnoDB creates them in the temporary file
directory, not the directory containing the original table. If
this directory is not large enough to hold such files, you may
need to set the tmpdir system
variable to a different directory. (See
Section C.5.4.4, “Where MySQL Stores Temporary Files”.)
Locking Options for Online DDL
While an InnoDB table is being changed by a DDL operation, the
table may or may not be
locked, depending on the
internal workings of that operation and the
LOCK clause of the ALTER
TABLE statement. By default, MySQL uses as little
locking as possible during a DDL operation; you specify the clause
either to make the locking more restrictive than it normally would
be (thus limiting concurrent DML, or DML and queries), or to
ensure that some expected degree of locking is allowed for an
operation. If the LOCK clause specifies a level
of locking that is not available for that specific kind of DDL
operation, such as LOCK=SHARED or
LOCK=NONE while creating or dropping a primary
key, the clause works like an assertion, causing the statement to
fail with an error. The following list shows the different
possibilities for the LOCK clause, from the
most permissive to the most restrictive:
For DDL operations with
LOCK=NONE, both queries and concurrent DML are allowed. This clause makes theALTER TABLEfail if the kind of DDL operation cannot be performed with the requested type of locking, so specifyLOCK=NONEif keeping the table fully available is vital and it is OK to cancel the DDL if that is not possible. For example, you might use this clause in DDLs for tables involving customer signups or purchases, to avoid making those tables unavailable by mistakenly issuing an expensiveALTER TABLEstatement.For DDL operations with
LOCK=SHARED, any writes to the table (that is, DML operations) are blocked, but the data in the table can be read. This clause makes theALTER TABLEfail if the kind of DDL operation cannot be performed with the requested type of locking, so specifyLOCK=SHAREDif keeping the table available for queries is vital and it is OK to cancel the DDL if that is not possible. For example, you might use this clause in DDLs for tables in a data warehouse, where it is OK to delay data load operations until the DDL is finished, but queries cannot be delayed for long periods.For DDL operations with
LOCK=DEFAULT, or with theLOCKclause omitted, MySQL uses the lowest level of locking that is available for that kind of operation, allowing concurrent queries, DML, or both wherever possible. This is the setting to use when making pre-planned, pre-tested changes that you know will not cause any availability problems based on the workload for that table.For DDL operations with
LOCK=EXCLUSIVE, both queries and DML operations are blocked. This clause makes theALTER TABLEfail if the kind of DDL operation cannot be performed with the requested type of locking, so specifyLOCK=EXCLUSIVEif the primary concern is finishing the DDL in the shortest time possible, and it is OK to make applications wait when they try to access the table. You might also useLOCK=EXCLUSIVEif the server is supposed to be idle, to avoid unexpected accesses to the table.
An online DDL statement for an InnoDB table always waits for currently executing transactions that are accessing the table to commit or roll back, because it requires exclusive access to the table for a brief period while the DDL statement is being prepared. Likewise, it requires exclusive access to the table for a brief time before finishing. Thus, an online DDL statement waits for any transactions that are started while the DDL is in progress, and query or modify the table, to commit or roll back before the DDL completes.
Because there is some processing work involved with recording the changes made by concurrent DML operations, then applying those changes at the end, an online DDL operation could take longer overall than the old-style mechanism that blocks table access from other sessions. The reduction in raw performance is balanced against better responsiveness for applications that use the table. When evaluating the ideal techniques for changing table structure, consider end-user perception of performance, based on factors such as load times for web pages.
A newly created InnoDB secondary index contains only the committed
data in the table at the time the CREATE
INDEX or ALTER TABLE
statement finishes executing. It does not contain any uncommitted
values, old versions of values, or values marked for deletion but
not yet removed from the old index.
Performance of In-Place versus Table-Copying DDL Operations
The raw performance of an online DDL operation is largely determined by whether the operation is performed in-place, or requires copying and rebuilding the entire table. See Table 5.8, “Summary of Online Status for DDL Operations” to see what kinds of operations can be performed in-place, and any requirements for avoiding table-copy operations.
The performance speedup from in-place DDL applies to operations on secondary indexes, not to the primary key index. The rows of an InnoDB table are stored in a clustered index organized based on the primary key, forming what some database systems call an “index-organized table”. Because the table structure is so closely tied to the primary key, redefining the primary key still requires copying the data.
When an operation on the primary key uses
ALGORITHM=INPLACE, even though the data is
still copied, it is more efficient than using
ALGORITHM=COPY because:
No undo logging or associated redo logging is required for
ALGORITHM=INPLACE. These operations add overhead to DDL statements that useALGORITHM=COPY.The secondary index entries are pre-sorted, and so can be loaded in order.
The change buffer is not used, because there are no random-access inserts into the secondary indexes.
To judge the relative performance of online DDL operations, you
can run such operations on a big InnoDB table
using current and earlier versions of MySQL. You can also run all
the performance tests under the latest MySQL version, simulating
the previous DDL behavior for the “before” results,
by setting the old_alter_table
system variable. Issue the statement set
old_alter_table=1 in the session, and measure DDL
performance to record the “before” figures. Then
set old_alter_table=0 to re-enable the newer,
faster behavior, and run the DDL operations again to record the
“after” figures.
For a basic idea of whether a DDL operation does its changes in-place or performs a table copy, look at the “rows affected” value displayed after the command finishes. For example, here are lines you might see after doing different types of DDL operations:
Changing the default value of a column (super-fast, does not affect the table data at all):
Query OK, 0 rows affected (0.07 sec)
Adding an index (takes time, but
0 rows affectedshows that the table is not copied):Query OK, 0 rows affected (21.42 sec)
Changing the data type of a column (takes substantial time and does require rebuilding all the rows of the table):
Query OK, 1671168 rows affected (1 min 35.54 sec)
NoteChanging the data type of a column requires rebuilding all the rows of the table with the exception of changing
VARCHARsize, which may be performed using onlineALTER TABLE. See Section 14.2.5.12, “Increase VARCHAR Size Online”.
For example, before running a DDL operation on a big table, you might check whether the operation will be fast or slow as follows:
Clone the table structure.
Populate the cloned table with a tiny amount of data.
Run the DDL operation on the cloned table.
Check whether the “rows affected” value is zero or not. A non-zero value means the operation will require rebuilding the entire table, which might require special planning. For example, you might do the DDL operation during a period of scheduled downtime, or on each replication slave server one at a time.
For a deeper understanding of the reduction in MySQL processing,
examine the performance_schema and
INFORMATION_SCHEMA tables related to
InnoDB before and after DDL operations, to see
the number of physical reads, writes, memory allocations, and so
on.
Typically, you do not need to do anything special to enable
online DDL when using the
ALTER TABLE statement for
InnoDB tables. See
Table 5.8, “Summary of Online Status for DDL Operations” for the kinds of
DDL operations that can be performed in-place, allowing concurrent
DML, or both. Some variations require particular combinations of
configuration settings or ALTER
TABLE clauses.
You can control the various aspects of a particular online DDL
operation by using the LOCK and
ALGORITHM clauses of the
ALTER TABLE statement. These
clauses come at the end of the statement, separated from the table
and column specifications by commas. The LOCK
clause is useful for fine-tuning the degree of concurrent access
to the table. The ALGORITHM clause is primarily
intended for performance comparisons and as a fallback to the
older table-copying behavior in case you encounter any issues with
existing DDL code. For example:
To avoid accidentally making the table unavailable for reads, writes, or both, you could specify a clause on the
ALTER TABLEstatement such asLOCK=NONE(allow both reads and writes) orLOCK=SHARED(allow reads). The operation halts immediately if the requested level of concurrency is not available.To compare performance, you could run one statement with
ALGORITHM=INPLACEand another withALGORITHM=COPY, as an alternative to setting theold_alter_tableconfiguration option.To avoid the chance of tying up the server by running an
ALTER TABLEthat copied the table, you could includeALGORITHM=INPLACEso the statement halts immediately if it cannot use the in-place mechanism. See Table 5.8, “Summary of Online Status for DDL Operations” for a list of the DDL operations that can or cannot be performed in-place.
See Section 5.5.2, “Performance and Concurrency Considerations for Online DDL” for more
details about the LOCK clause. For full
examples of using online DDL, see
Section 5.5.5, “Examples of Online DDL”.
Before the introduction of online
DDL, it was common practice to combine many DDL operations
into a single ALTER TABLE
statement. Because each ALTER TABLE
statement involved copying and rebuilding the table, it was more
efficient to make several changes to the same table at once, since
those changes could all be done with a single rebuild operation
for the table. The downside was that SQL code involving DDL
operations was harder to maintain and to reuse in different
scripts. If the specific changes were different each time, you
might have to construct a new complex ALTER
TABLE for each slightly different scenario.
For DDL operations that can be done in-place, as shown in
Table 5.8, “Summary of Online Status for DDL Operations”, now you can
separate them into individual ALTER
TABLE statements for easier scripting and maintenance,
without sacrificing efficiency. For example, you might take a
complicated statement such as:
alter table t1 add index i1(c1), add unique index i2(c2), change c4_old_name c4_new_name integer unsigned;
and break it down into simpler parts that can be tested and performed independently, such as:
alter table t1 add index i1(c1); alter table t1 add unique index i2(c2); alter table t1 change c4_old_name c4_new_name integer unsigned not null;
You might still use multi-part ALTER
TABLE statements for:
Operations that must be performed in a specific sequence, such as creating an index followed by a foreign key constraint that uses that index.
Operations all using the same specific
LOCKclause, that you want to either succeed or fail as a group.Operations that cannot be performed in-place, that is, that still copy and rebuild the table.
Operations for which you specify
ALGORITHM=COPYorold_alter_table=1, to force the table-copying behavior if needed for precise backward-compatibility in specialized scenarios.
Here are code examples showing some operations whose performance, concurrency, and scalability are improved by the latest online DDL enhancements.
Example 5.1, “Schema Setup Code for Online DDL Experiments” sets up tables named
BIG_TABLEandSMALL_TABLEused in the subsequent examples.Example 5.2, “Speed and Efficiency of CREATE INDEX and DROP INDEX” illustrates the performance aspects of creating and dropping indexes.
Example 5.3, “Concurrent DML During CREATE INDEX and DROP INDEX” shows queries and DML statements running during a
DROP INDEXoperation.Example 5.4, “Renaming a Column” demonstrates the speed improvement for renaming a column, and shows the care needed to keep the data type precisely the same when doing the rename operation.
Example 5.5, “Dropping Foreign Keys” demonstrates how foreign keys work with online DDL. Because two tables are involved in foreign key operations, there are extra locking considerations. Thus, tables with foreign keys sometimes have restrictions for online DDL operations.
Example 5.6, “Changing Auto-Increment Value” demonstrates how auto-increment columns work with online DDL. Tables with auto-increment columns sometimes have restrictions for online DDL operations.
Example 5.7, “Controlling Concurrency with the
LOCKClause” demonstrates the options to permit or restrict concurrent queries and DML operations while an online DDL operation is in progress. It shows the situations when the DDL statement might wait, or the concurrent transaction might wait, or the concurrent transaction might cancel a DML statement due to a deadlock error.Example 5.8, “Schema Setup Code for Online DDL Experiments” demonstrates creating and dropping multiple indexes in a single statement, which can be more efficient than using a separate statement for each index operation.
Example 5.9, “Creating and Dropping the Primary Key” demonstrates how it is more efficient to define a primary key when creating the table, and relatively expensive to add one later.
Example 5.1. Schema Setup Code for Online DDL Experiments
Here is the code that sets up the initial tables used in these demonstrations:
/* Setup code for the online DDL demonstration: - Set up some config variables. - Create 2 tables that are clones of one of the INFORMATION_SCHEMA tables that always has some data. The "small" table has a couple of thousand rows. For the "big" table, keep doubling the data until it reaches over a million rows. - Set up a primary key for the sample tables, since we are demonstrating InnoDB aspects. */ set autocommit = 0; set foreign_key_checks = 1; set global innodb_file_per_table = 1; set old_alter_table=0; prompt mysql: use test; \! echo "Setting up 'small' table:" drop table if exists small_table; create table small_table as select * from information_schema.columns; alter table small_table add id int unsigned not null primary key auto_increment; select count(id) from small_table; \! echo "Setting up 'big' table:" drop table if exists big_table; create table big_table as select * from information_schema.columns; show create table big_table\G insert into big_table select * from big_table; insert into big_table select * from big_table; insert into big_table select * from big_table; insert into big_table select * from big_table; insert into big_table select * from big_table; insert into big_table select * from big_table; insert into big_table select * from big_table; insert into big_table select * from big_table; insert into big_table select * from big_table; insert into big_table select * from big_table; commit; alter table big_table add id int unsigned not null primary key auto_increment; select count(id) from big_table;
Running this code gives this output, condensed for brevity and with the most important points bolded:
Setting up 'small' table: Query OK, 0 rows affected (0.01 sec) Query OK, 1678 rows affected (0.13 sec) Records: 1678 Duplicates: 0 Warnings: 0 Query OK, 1678 rows affected (0.07 sec) Records: 1678 Duplicates: 0 Warnings: 0 +-----------+ | count(id) | +-----------+ | 1678 | +-----------+ 1 row in set (0.00 sec) Setting up 'big' table: Query OK, 0 rows affected (0.16 sec) Query OK, 1678 rows affected (0.17 sec) Records: 1678 Duplicates: 0 Warnings: 0 *************************** 1. row *************************** Table: big_table Create Table: CREATE TABLE `big_table` ( `TABLE_CATALOG` varchar(512) CHARACTER SET utf8 NOT NULL DEFAULT '', `TABLE_SCHEMA` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '', `TABLE_NAME` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '', `COLUMN_NAME` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '', `ORDINAL_POSITION` bigint(21) unsigned NOT NULL DEFAULT '0', `COLUMN_DEFAULT` longtext CHARACTER SET utf8, `IS_NULLABLE` varchar(3) CHARACTER SET utf8 NOT NULL DEFAULT '', `DATA_TYPE` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '', `CHARACTER_MAXIMUM_LENGTH` bigint(21) unsigned DEFAULT NULL, `CHARACTER_OCTET_LENGTH` bigint(21) unsigned DEFAULT NULL, `NUMERIC_PRECISION` bigint(21) unsigned DEFAULT NULL, `NUMERIC_SCALE` bigint(21) unsigned DEFAULT NULL, `DATETIME_PRECISION` bigint(21) unsigned DEFAULT NULL, `CHARACTER_SET_NAME` varchar(32) CHARACTER SET utf8 DEFAULT NULL, `COLLATION_NAME` varchar(32) CHARACTER SET utf8 DEFAULT NULL, `COLUMN_TYPE` longtext CHARACTER SET utf8 NOT NULL, `COLUMN_KEY` varchar(3) CHARACTER SET utf8 NOT NULL DEFAULT '', `EXTRA` varchar(30) CHARACTER SET utf8 NOT NULL DEFAULT '', `PRIVILEGES` varchar(80) CHARACTER SET utf8 NOT NULL DEFAULT '', `COLUMN_COMMENT` varchar(1024) CHARACTER SET utf8 NOT NULL DEFAULT '' ) ENGINE=InnoDB DEFAULT CHARSET=latin1 1 row in set (0.00 sec) Query OK, 1678 rows affected (0.09 sec) Records: 1678 Duplicates: 0 Warnings: 0 Query OK, 3356 rows affected (0.07 sec) Records: 3356 Duplicates: 0 Warnings: 0 Query OK, 6712 rows affected (0.17 sec) Records: 6712 Duplicates: 0 Warnings: 0 Query OK, 13424 rows affected (0.44 sec) Records: 13424 Duplicates: 0 Warnings: 0 Query OK, 26848 rows affected (0.63 sec) Records: 26848 Duplicates: 0 Warnings: 0 Query OK, 53696 rows affected (1.72 sec) Records: 53696 Duplicates: 0 Warnings: 0 Query OK, 107392 rows affected (3.02 sec) Records: 107392 Duplicates: 0 Warnings: 0 Query OK, 214784 rows affected (6.28 sec) Records: 214784 Duplicates: 0 Warnings: 0 Query OK, 429568 rows affected (13.25 sec) Records: 429568 Duplicates: 0 Warnings: 0 Query OK, 859136 rows affected (28.16 sec) Records: 859136 Duplicates: 0 Warnings: 0 Query OK, 0 rows affected (0.03 sec) Query OK, 1718272 rows affected (1 min 9.22 sec) Records: 1718272 Duplicates: 0 Warnings: 0 +-----------+ | count(id) | +-----------+ | 1718272 | +-----------+ 1 row in set (1.75 sec)
Example 5.2. Speed and Efficiency of CREATE INDEX and DROP INDEX
Here is a sequence of statements demonstrating the relative
speed of CREATE INDEX and
DROP INDEX statements. For a
small table, the elapsed time is less than a second whether we
use the fast or slow technique, so we look at the “rows
affected” output to verify which operations can avoid the
table rebuild. For a large table, the difference in efficiency
is obvious because skipping the table rebuild saves substantial
time.
\! clear \! echo "=== Create and drop index (small table, new/fast technique) ===" \! echo \! echo "Data size (kilobytes) before index created: " \! du -k data/test/small_table.ibd create index i_dtyp_small on small_table (data_type), algorithm=inplace; \! echo "Data size after index created: " \! du -k data/test/small_table.ibd drop index i_dtyp_small on small_table, algorithm=inplace; -- Compare against the older slower DDL. \! echo "=== Create and drop index (small table, old/slow technique) ===" \! echo \! echo "Data size (kilobytes) before index created: " \! du -k data/test/small_table.ibd create index i_dtyp_small on small_table (data_type), algorithm=copy; \! echo "Data size after index created: " \! du -k data/test/small_table.ibd drop index i_dtyp_small on small_table, algorithm=copy; -- In the above example, we examined the "rows affected" number, -- ideally looking for a zero figure. Let's try again with a larger -- sample size, where we'll see that the actual time taken can -- vary significantly. \! echo "=== Create and drop index (big table, new/fast technique) ===" \! echo \! echo "Data size (kilobytes) before index created: " \! du -k data/test/big_table.ibd create index i_dtyp_big on big_table (data_type), algorithm=inplace; \! echo "Data size after index created: " \! du -k data/test/big_table.ibd drop index i_dtyp_big on big_table, algorithm=inplace; \! echo "=== Create and drop index (big table, old/slow technique) ===" \! echo \! echo "Data size (kilobytes) before index created: " \! du -k data/test/big_table.ibd create index i_dtyp_big on big_table (data_type), algorithm=copy; \! echo "Data size after index created: " \! du -k data/test/big_table.ibd drop index i_dtyp_big on big_table, algorithm=copy;
Running this code gives this output, condensed for brevity and with the most important points bolded:
Query OK, 0 rows affected (0.00 sec) === Create and drop index (small table, new/fast technique) === Data size (kilobytes) before index created: 384 data/test/small_table.ibd Query OK, 0 rows affected (0.04 sec) Records: 0 Duplicates: 0 Warnings: 0 Data size after index created: 432 data/test/small_table.ibd Query OK, 0 rows affected (0.02 sec) Records: 0 Duplicates: 0 Warnings: 0 Query OK, 0 rows affected (0.00 sec) === Create and drop index (small table, old/slow technique) === Data size (kilobytes) before index created: 432 data/test/small_table.ibd Query OK, 1678 rows affected (0.12 sec) Records: 1678 Duplicates: 0 Warnings: 0 Data size after index created: 448 data/test/small_table.ibd Query OK, 1678 rows affected (0.10 sec) Records: 1678 Duplicates: 0 Warnings: 0 Query OK, 0 rows affected (0.00 sec) === Create and drop index (big table, new/fast technique) === Data size (kilobytes) before index created: 315392 data/test/big_table.ibd Query OK, 0 rows affected (33.32 sec) Records: 0 Duplicates: 0 Warnings: 0 Data size after index created: 335872 data/test/big_table.ibd Query OK, 0 rows affected (0.02 sec) Records: 0 Duplicates: 0 Warnings: 0 Query OK, 0 rows affected (0.00 sec) === Create and drop index (big table, old/slow technique) === Data size (kilobytes) before index created: 335872 data/test/big_table.ibd Query OK, 1718272 rows affected (1 min 5.01 sec) Records: 1718272 Duplicates: 0 Warnings: 0 Data size after index created: 348160 data/test/big_table.ibd Query OK, 1718272 rows affected (46.59 sec) Records: 1718272 Duplicates: 0 Warnings: 0
Example 5.3. Concurrent DML During CREATE INDEX and DROP INDEX
Here are some snippets of code that I ran in separate
mysql sessions connected to the same
database, to illustrate DML statements (insert, update, or
delete) running at the same time as CREATE
INDEX and DROP INDEX.
/* CREATE INDEX statement to run against a table while insert/update/delete statements are modifying the column being indexed. */ -- We'll run this script in one session, while simultaneously creating and dropping -- an index on test/big_table.table_name in another session. use test; create index i_concurrent on big_table(table_name);
/* DROP INDEX statement to run against a table while insert/update/delete statements are modifying the column being indexed. */ -- We'll run this script in one session, while simultaneously creating and dropping -- an index on test/big_table.table_name in another session. use test; drop index i_concurrent on big_table;
/* Some queries and insert/update/delete statements to run against a table while an index is being created or dropped. Previously, these operations would have stalled during the index create/drop period and possibly timed out or deadlocked. */ -- We'll run this script in one session, while simultaneously creating and dropping -- an index on test/big_table.table_name in another session. -- In our test instance, that column has about 1.7M rows, with 136 different values. -- Sample values: COLUMNS (20480), ENGINES (6144), EVENTS (24576), FILES (38912), TABLES (21504), VIEWS (10240). set autocommit = 0; use test; select distinct character_set_name from big_table where table_name = 'FILES'; delete from big_table where table_name = 'FILES'; select distinct character_set_name from big_table where table_name = 'FILES'; -- I'll issue the final rollback interactively, not via script, -- the better to control the timing. -- rollback;
Running this code gives this output, condensed for brevity and with the most important points bolded:
mysql: source concurrent_ddl_create.sql Database changed Query OK, 0 rows affected (1 min 25.15 sec) Records: 0 Duplicates: 0 Warnings: 0 mysql: source concurrent_ddl_drop.sql Database changed Query OK, 0 rows affected (24.98 sec) Records: 0 Duplicates: 0 Warnings: 0 mysql: source concurrent_dml.sql Query OK, 0 rows affected (0.00 sec) Database changed +--------------------+ | character_set_name | +--------------------+ | NULL | | utf8 | +--------------------+ 2 rows in set (0.32 sec) Query OK, 38912 rows affected (1.84 sec) Empty set (0.01 sec) mysql: rollback; Query OK, 0 rows affected (1.05 sec)
Example 5.4. Renaming a Column
Here is a demonstration of using ALTER
TABLE to rename a column. We use the new, fast DDL
mechanism to change the name, then the old, slow DDL mechanism
(with old_alter_table=1) to restore the
original column name.
Notes:
Because the syntax for renaming a column also involves re-specifying the data type, be very careful to specify exactly the same data type to avoid a costly table rebuild. In this case, we checked the output of
show create tableand copied any clauses such astable\GCHARACTER SETandNOT NULLfrom the original column definition.Again, renaming a column for a small table is fast enough that we need to examine the “rows affected” number to verify that the new DDL mechanism is more efficient than the old one. With a big table, the difference in elapsed time makes the improvement obvious.
/* Run through a sequence of 'rename column' statements. Because this operation involves only metadata, not table data, it is fast for big and small tables, with new or old DDL mechanisms. */ \! clear \! echo "Rename column (fast technique, small table):" alter table small_table change `IS_NULLABLE` `NULLABLE` varchar(3) character set utf8 not null, algorithm=inplace; \! echo "Rename back to original name (slow technique):" alter table small_table change `NULLABLE` `IS_NULLABLE` varchar(3) character set utf8 not null, algorithm=copy; \! echo "Rename column (fast technique, big table):" alter table big_table change `IS_NULLABLE` `NULLABLE` varchar(3) character set utf8 not null, algorithm=inplace; \! echo "Rename back to original name (slow technique):" alter table big_table change `NULLABLE` `IS_NULLABLE` varchar(3) character set utf8 not null, algorithm=copy;
Running this code gives this output, condensed for brevity and with the most important points bolded:
Rename column (fast technique, small table): Query OK, 0 rows affected (0.05 sec) Query OK, 0 rows affected (0.13 sec) Records: 0 Duplicates: 0 Warnings: 0 Rename back to original name (slow technique): Query OK, 0 rows affected (0.00 sec) Query OK, 1678 rows affected (0.35 sec) Records: 1678 Duplicates: 0 Warnings: 0 Rename column (fast technique, big table): Query OK, 0 rows affected (0.00 sec) Query OK, 0 rows affected (0.11 sec) Records: 0 Duplicates: 0 Warnings: 0 Rename back to original name (slow technique): Query OK, 0 rows affected (0.00 sec) Query OK, 1718272 rows affected (1 min 0.00 sec) Records: 1718272 Duplicates: 0 Warnings: 0 Query OK, 0 rows affected (0.00 sec)
Example 5.5. Dropping Foreign Keys
Here is a demonstration of foreign keys, including improvement to the speed of dropping a foreign key constraint.
/* Demonstrate aspects of foreign keys that are or aren't affected by the DDL improvements. - Create a new table with only a few values to serve as the parent table. - Set up the 'small' and 'big' tables as child tables using a foreign key. - Verify that the ON DELETE CASCADE clause makes changes ripple from parent to child tables. - Drop the foreign key constraints, and optionally associated indexes. (This is the operation that is sped up.) */ \! clear -- Make sure foreign keys are being enforced, and allow -- rollback after doing some DELETEs that affect both -- parent and child tables. set foreign_key_checks = 1; set autocommit = 0; -- Create a parent table, containing values that we know are already present -- in the child tables. drop table if exists schema_names; create table schema_names (id int unsigned not null primary key auto_increment, schema_name varchar(64) character set utf8 not null, index i_schema (schema_name)) as select distinct table_schema schema_name from small_table; show create table schema_names\G show create table small_table\G show create table big_table\G -- Creating the foreign key constraint still involves a table rebuild when foreign_key_checks=1, -- as illustrated by the "rows affected" figure. alter table small_table add constraint small_fk foreign key i_table_schema (table_schema) references schema_names(schema_name) on delete cascade; alter table big_table add constraint big_fk foreign key i_table_schema (table_schema) references schema_names(schema_name) on delete cascade; show create table small_table\G show create table big_table\G select schema_name from schema_names order by schema_name; select count(table_schema) howmany, table_schema from small_table group by table_schema; select count(table_schema) howmany, table_schema from big_table group by table_schema; -- big_table is the parent table. -- schema_names is the parent table. -- big_table is the child table. -- (One row in the parent table can have many "children" in the child table.) -- Changes to the parent table can ripple through to the child table. -- For example, removing the value 'test' from schema_names.schema_name will -- result in the removal of 20K or so rows from big_table. delete from schema_names where schema_name = 'test'; select schema_name from schema_names order by schema_name; select count(table_schema) howmany, table_schema from small_table group by table_schema; select count(table_schema) howmany, table_schema from big_table group by table_schema; -- Because we've turned off autocommit, we can still get back those deleted rows -- if the DELETE was issued by mistake. rollback; select schema_name from schema_names order by schema_name; select count(table_schema) howmany, table_schema from small_table group by table_schema; select count(table_schema) howmany, table_schema from big_table group by table_schema; -- All of the cross-checking between parent and child tables would be -- deadly slow if there wasn't the requirement for the corresponding -- columns to be indexed! -- But we can get rid of the foreign key using a fast operation -- that doesn't rebuild the table. -- If we didn't specify a constraint name when setting up the foreign key, we would -- have to find the auto-generated name such as 'big_table_ibfk_1' in the -- output from 'show create table'. -- For the small table, we'll drop the foreign key and the associated index. -- Having an index on a small table is less critical. \! echo "DROP FOREIGN KEY and INDEX from small_table:" alter table small_table drop foreign key small_fk, drop index small_fk; -- For the big table, we'll drop the foreign key and leave the associated index. -- If we are still doing queries that reference the indexed column, the index is -- very important to avoid a full table scan of the big table. \! echo "DROP FOREIGN KEY from big_table:" alter table big_table drop foreign key big_fk; show create table small_table\G show create table big_table\G
Running this code gives this output, condensed for brevity and with the most important points bolded:
Query OK, 0 rows affected (0.00 sec)
Query OK, 0 rows affected (0.00 sec)
Query OK, 0 rows affected (0.01 sec)
Query OK, 4 rows affected (0.03 sec)
Records: 4 Duplicates: 0 Warnings: 0
*************************** 1. row ***************************
Table: schema_names
Create Table: CREATE TABLE `schema_names` (
`id` int(10) unsigned NOT NULL AUTO_INCREMENT,
`schema_name` varchar(64) CHARACTER SET utf8 NOT NULL,
PRIMARY KEY (`id`),
KEY `i_schema` (`schema_name`)
) ENGINE=InnoDB AUTO_INCREMENT=8 DEFAULT CHARSET=latin1
1 row in set (0.00 sec)
*************************** 1. row ***************************
Table: small_table
Create Table: CREATE TABLE `small_table` (
`TABLE_CATALOG` varchar(512) CHARACTER SET utf8 NOT NULL DEFAULT '',
`TABLE_SCHEMA` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`TABLE_NAME` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`COLUMN_NAME` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`ORDINAL_POSITION` bigint(21) unsigned NOT NULL DEFAULT '0',
`COLUMN_DEFAULT` longtext CHARACTER SET utf8,
`IS_NULLABLE` varchar(3) CHARACTER SET utf8 NOT NULL,
`DATA_TYPE` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`CHARACTER_MAXIMUM_LENGTH` bigint(21) unsigned DEFAULT NULL,
`CHARACTER_OCTET_LENGTH` bigint(21) unsigned DEFAULT NULL,
`NUMERIC_PRECISION` bigint(21) unsigned DEFAULT NULL,
`NUMERIC_SCALE` bigint(21) unsigned DEFAULT NULL,
`DATETIME_PRECISION` bigint(21) unsigned DEFAULT NULL,
`CHARACTER_SET_NAME` varchar(32) CHARACTER SET utf8 DEFAULT NULL,
`COLLATION_NAME` varchar(32) CHARACTER SET utf8 DEFAULT NULL,
`COLUMN_TYPE` longtext CHARACTER SET utf8 NOT NULL,
`COLUMN_KEY` varchar(3) CHARACTER SET utf8 NOT NULL DEFAULT '',
`EXTRA` varchar(30) CHARACTER SET utf8 NOT NULL DEFAULT '',
`PRIVILEGES` varchar(80) CHARACTER SET utf8 NOT NULL DEFAULT '',
`COLUMN_COMMENT` varchar(1024) CHARACTER SET utf8 NOT NULL DEFAULT '',
`id` int(10) unsigned NOT NULL AUTO_INCREMENT,
PRIMARY KEY (`id`)
) ENGINE=InnoDB AUTO_INCREMENT=1679 DEFAULT CHARSET=latin1
1 row in set (0.00 sec)
*************************** 1. row ***************************
Table: big_table
Create Table: CREATE TABLE `big_table` (
`TABLE_CATALOG` varchar(512) CHARACTER SET utf8 NOT NULL DEFAULT '',
`TABLE_SCHEMA` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`TABLE_NAME` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`COLUMN_NAME` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`ORDINAL_POSITION` bigint(21) unsigned NOT NULL DEFAULT '0',
`COLUMN_DEFAULT` longtext CHARACTER SET utf8,
`IS_NULLABLE` varchar(3) CHARACTER SET utf8 NOT NULL,
`DATA_TYPE` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`CHARACTER_MAXIMUM_LENGTH` bigint(21) unsigned DEFAULT NULL,
`CHARACTER_OCTET_LENGTH` bigint(21) unsigned DEFAULT NULL,
`NUMERIC_PRECISION` bigint(21) unsigned DEFAULT NULL,
`NUMERIC_SCALE` bigint(21) unsigned DEFAULT NULL,
`DATETIME_PRECISION` bigint(21) unsigned DEFAULT NULL,
`CHARACTER_SET_NAME` varchar(32) CHARACTER SET utf8 DEFAULT NULL,
`COLLATION_NAME` varchar(32) CHARACTER SET utf8 DEFAULT NULL,
`COLUMN_TYPE` longtext CHARACTER SET utf8 NOT NULL,
`COLUMN_KEY` varchar(3) CHARACTER SET utf8 NOT NULL DEFAULT '',
`EXTRA` varchar(30) CHARACTER SET utf8 NOT NULL DEFAULT '',
`PRIVILEGES` varchar(80) CHARACTER SET utf8 NOT NULL DEFAULT '',
`COLUMN_COMMENT` varchar(1024) CHARACTER SET utf8 NOT NULL DEFAULT '',
`id` int(10) unsigned NOT NULL AUTO_INCREMENT,
PRIMARY KEY (`id`),
KEY `big_fk` (`TABLE_SCHEMA`)
) ENGINE=InnoDB AUTO_INCREMENT=1718273 DEFAULT CHARSET=latin1
1 row in set (0.00 sec)
Query OK, 1678 rows affected (0.10 sec)
Records: 1678 Duplicates: 0 Warnings: 0
Query OK, 1718272 rows affected (1 min 14.54 sec)
Records: 1718272 Duplicates: 0 Warnings: 0
*************************** 1. row ***************************
Table: small_table
Create Table: CREATE TABLE `small_table` (
`TABLE_CATALOG` varchar(512) CHARACTER SET utf8 NOT NULL DEFAULT '',
`TABLE_SCHEMA` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`TABLE_NAME` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`COLUMN_NAME` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`ORDINAL_POSITION` bigint(21) unsigned NOT NULL DEFAULT '0',
`COLUMN_DEFAULT` longtext CHARACTER SET utf8,
`IS_NULLABLE` varchar(3) CHARACTER SET utf8 NOT NULL,
`DATA_TYPE` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`CHARACTER_MAXIMUM_LENGTH` bigint(21) unsigned DEFAULT NULL,
`CHARACTER_OCTET_LENGTH` bigint(21) unsigned DEFAULT NULL,
`NUMERIC_PRECISION` bigint(21) unsigned DEFAULT NULL,
`NUMERIC_SCALE` bigint(21) unsigned DEFAULT NULL,
`DATETIME_PRECISION` bigint(21) unsigned DEFAULT NULL,
`CHARACTER_SET_NAME` varchar(32) CHARACTER SET utf8 DEFAULT NULL,
`COLLATION_NAME` varchar(32) CHARACTER SET utf8 DEFAULT NULL,
`COLUMN_TYPE` longtext CHARACTER SET utf8 NOT NULL,
`COLUMN_KEY` varchar(3) CHARACTER SET utf8 NOT NULL DEFAULT '',
`EXTRA` varchar(30) CHARACTER SET utf8 NOT NULL DEFAULT '',
`PRIVILEGES` varchar(80) CHARACTER SET utf8 NOT NULL DEFAULT '',
`COLUMN_COMMENT` varchar(1024) CHARACTER SET utf8 NOT NULL DEFAULT '',
`id` int(10) unsigned NOT NULL AUTO_INCREMENT,
PRIMARY KEY (`id`),
KEY `small_fk` (`TABLE_SCHEMA`),
CONSTRAINT `small_fk` FOREIGN KEY (`TABLE_SCHEMA`) REFERENCES `schema_names` (`schema_name`) ON DELETE CASCADE
) ENGINE=InnoDB AUTO_INCREMENT=1679 DEFAULT CHARSET=latin1
1 row in set (0.12 sec)
*************************** 1. row ***************************
Table: big_table
Create Table: CREATE TABLE `big_table` (
`TABLE_CATALOG` varchar(512) CHARACTER SET utf8 NOT NULL DEFAULT '',
`TABLE_SCHEMA` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`TABLE_NAME` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`COLUMN_NAME` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`ORDINAL_POSITION` bigint(21) unsigned NOT NULL DEFAULT '0',
`COLUMN_DEFAULT` longtext CHARACTER SET utf8,
`IS_NULLABLE` varchar(3) CHARACTER SET utf8 NOT NULL,
`DATA_TYPE` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`CHARACTER_MAXIMUM_LENGTH` bigint(21) unsigned DEFAULT NULL,
`CHARACTER_OCTET_LENGTH` bigint(21) unsigned DEFAULT NULL,
`NUMERIC_PRECISION` bigint(21) unsigned DEFAULT NULL,
`NUMERIC_SCALE` bigint(21) unsigned DEFAULT NULL,
`DATETIME_PRECISION` bigint(21) unsigned DEFAULT NULL,
`CHARACTER_SET_NAME` varchar(32) CHARACTER SET utf8 DEFAULT NULL,
`COLLATION_NAME` varchar(32) CHARACTER SET utf8 DEFAULT NULL,
`COLUMN_TYPE` longtext CHARACTER SET utf8 NOT NULL,
`COLUMN_KEY` varchar(3) CHARACTER SET utf8 NOT NULL DEFAULT '',
`EXTRA` varchar(30) CHARACTER SET utf8 NOT NULL DEFAULT '',
`PRIVILEGES` varchar(80) CHARACTER SET utf8 NOT NULL DEFAULT '',
`COLUMN_COMMENT` varchar(1024) CHARACTER SET utf8 NOT NULL DEFAULT '',
`id` int(10) unsigned NOT NULL AUTO_INCREMENT,
PRIMARY KEY (`id`),
KEY `big_fk` (`TABLE_SCHEMA`),
CONSTRAINT `big_fk` FOREIGN KEY (`TABLE_SCHEMA`) REFERENCES `schema_names` (`schema_name`) ON DELETE CASCADE
) ENGINE=InnoDB AUTO_INCREMENT=1718273 DEFAULT CHARSET=latin1
1 row in set (0.01 sec)
+--------------------+
| schema_name |
+--------------------+
| information_schema |
| mysql |
| performance_schema |
| test |
+--------------------+
4 rows in set (0.00 sec)
+---------+--------------------+
| howmany | table_schema |
+---------+--------------------+
| 563 | information_schema |
| 286 | mysql |
| 786 | performance_schema |
| 43 | test |
+---------+--------------------+
4 rows in set (0.01 sec)
+---------+--------------------+
| howmany | table_schema |
+---------+--------------------+
| 576512 | information_schema |
| 292864 | mysql |
| 804864 | performance_schema |
| 44032 | test |
+---------+--------------------+
4 rows in set (2.10 sec)
Query OK, 1 row affected (1.52 sec)
+--------------------+
| schema_name |
+--------------------+
| information_schema |
| mysql |
| performance_schema |
+--------------------+
3 rows in set (0.00 sec)
+---------+--------------------+
| howmany | table_schema |
+---------+--------------------+
| 563 | information_schema |
| 286 | mysql |
| 786 | performance_schema |
+---------+--------------------+
3 rows in set (0.00 sec)
+---------+--------------------+
| howmany | table_schema |
+---------+--------------------+
| 576512 | information_schema |
| 292864 | mysql |
| 804864 | performance_schema |
+---------+--------------------+
3 rows in set (1.74 sec)
Query OK, 0 rows affected (0.60 sec)
+--------------------+
| schema_name |
+--------------------+
| information_schema |
| mysql |
| performance_schema |
| test |
+--------------------+
4 rows in set (0.00 sec)
+---------+--------------------+
| howmany | table_schema |
+---------+--------------------+
| 563 | information_schema |
| 286 | mysql |
| 786 | performance_schema |
| 43 | test |
+---------+--------------------+
4 rows in set (0.01 sec)
+---------+--------------------+
| howmany | table_schema |
+---------+--------------------+
| 576512 | information_schema |
| 292864 | mysql |
| 804864 | performance_schema |
| 44032 | test |
+---------+--------------------+
4 rows in set (1.59 sec)
DROP FOREIGN KEY and INDEX from small_table:
Query OK, 0 rows affected (0.02 sec)
Records: 0 Duplicates: 0 Warnings: 0
DROP FOREIGN KEY from big_table:
Query OK, 0 rows affected (0.02 sec)
Records: 0 Duplicates: 0 Warnings: 0
*************************** 1. row ***************************
Table: small_table
Create Table: CREATE TABLE `small_table` (
`TABLE_CATALOG` varchar(512) CHARACTER SET utf8 NOT NULL DEFAULT '',
`TABLE_SCHEMA` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`TABLE_NAME` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`COLUMN_NAME` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`ORDINAL_POSITION` bigint(21) unsigned NOT NULL DEFAULT '0',
`COLUMN_DEFAULT` longtext CHARACTER SET utf8,
`IS_NULLABLE` varchar(3) CHARACTER SET utf8 NOT NULL,
`DATA_TYPE` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`CHARACTER_MAXIMUM_LENGTH` bigint(21) unsigned DEFAULT NULL,
`CHARACTER_OCTET_LENGTH` bigint(21) unsigned DEFAULT NULL,
`NUMERIC_PRECISION` bigint(21) unsigned DEFAULT NULL,
`NUMERIC_SCALE` bigint(21) unsigned DEFAULT NULL,
`DATETIME_PRECISION` bigint(21) unsigned DEFAULT NULL,
`CHARACTER_SET_NAME` varchar(32) CHARACTER SET utf8 DEFAULT NULL,
`COLLATION_NAME` varchar(32) CHARACTER SET utf8 DEFAULT NULL,
`COLUMN_TYPE` longtext CHARACTER SET utf8 NOT NULL,
`COLUMN_KEY` varchar(3) CHARACTER SET utf8 NOT NULL DEFAULT '',
`EXTRA` varchar(30) CHARACTER SET utf8 NOT NULL DEFAULT '',
`PRIVILEGES` varchar(80) CHARACTER SET utf8 NOT NULL DEFAULT '',
`COLUMN_COMMENT` varchar(1024) CHARACTER SET utf8 NOT NULL DEFAULT '',
`id` int(10) unsigned NOT NULL AUTO_INCREMENT,
PRIMARY KEY (`id`)
) ENGINE=InnoDB AUTO_INCREMENT=1679 DEFAULT CHARSET=latin1
1 row in set (0.00 sec)
*************************** 1. row ***************************
Table: big_table
Create Table: CREATE TABLE `big_table` (
`TABLE_CATALOG` varchar(512) CHARACTER SET utf8 NOT NULL DEFAULT '',
`TABLE_SCHEMA` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`TABLE_NAME` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`COLUMN_NAME` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`ORDINAL_POSITION` bigint(21) unsigned NOT NULL DEFAULT '0',
`COLUMN_DEFAULT` longtext CHARACTER SET utf8,
`IS_NULLABLE` varchar(3) CHARACTER SET utf8 NOT NULL,
`DATA_TYPE` varchar(64) CHARACTER SET utf8 NOT NULL DEFAULT '',
`CHARACTER_MAXIMUM_LENGTH` bigint(21) unsigned DEFAULT NULL,
`CHARACTER_OCTET_LENGTH` bigint(21) unsigned DEFAULT NULL,
`NUMERIC_PRECISION` bigint(21) unsigned DEFAULT NULL,
`NUMERIC_SCALE` bigint(21) unsigned DEFAULT NULL,
`DATETIME_PRECISION` bigint(21) unsigned DEFAULT NULL,
`CHARACTER_SET_NAME` varchar(32) CHARACTER SET utf8 DEFAULT NULL,
`COLLATION_NAME` varchar(32) CHARACTER SET utf8 DEFAULT NULL,
`COLUMN_TYPE` longtext CHARACTER SET utf8 NOT NULL,
`COLUMN_KEY` varchar(3) CHARACTER SET utf8 NOT NULL DEFAULT '',
`EXTRA` varchar(30) CHARACTER SET utf8 NOT NULL DEFAULT '',
`PRIVILEGES` varchar(80) CHARACTER SET utf8 NOT NULL DEFAULT '',
`COLUMN_COMMENT` varchar(1024) CHARACTER SET utf8 NOT NULL DEFAULT '',
`id` int(10) unsigned NOT NULL AUTO_INCREMENT,
PRIMARY KEY (`id`),
KEY `big_fk` (`TABLE_SCHEMA`)
) ENGINE=InnoDB AUTO_INCREMENT=1718273 DEFAULT CHARSET=latin1
1 row in set (0.00 sec)
Example 5.6. Changing Auto-Increment Value
Here is an illustration of increasing the
auto-increment lower
limit for a table column, demonstrating how this operation now
avoids a table rebuild, plus some other fun facts about
InnoDB auto-increment columns.
/*
If this script is run after foreign_key.sql, the schema_names table is
already set up. But to allow this script to run multiple times without
running into duplicate ID errors, we set up the schema_names table
all over again.
*/
\! clear
\! echo "=== Adjusting the Auto-Increment Limit for a Table ==="
\! echo
drop table if exists schema_names;
create table schema_names (id int unsigned not null primary key auto_increment,
schema_name varchar(64) character set utf8 not null, index i_schema (schema_name))
as select distinct table_schema schema_name from small_table;
\! echo "Initial state of schema_names table. AUTO_INCREMENT is included in SHOW CREATE TABLE output."
\! echo "Note how MySQL reserved a block of IDs, but only needed 4 of them in this transaction, so the next inserted values would get IDs 8 and 9."
show create table schema_names\G
select * from schema_names order by id;
\! echo "Inserting even a tiny amount of data can produce gaps in the ID sequence."
insert into schema_names (schema_name) values ('eight'), ('nine');
\! echo "Bumping auto-increment lower limit to 20 (fast mechanism):"
alter table schema_names auto_increment=20, algorithm=inplace;
\! echo "Inserting 2 rows that should get IDs 20 and 21:"
insert into schema_names (schema_name) values ('foo'), ('bar');
commit;
\! echo "Bumping auto-increment lower limit to 30 (slow mechanism):"
alter table schema_names auto_increment=30, algorithm=copy;
\! echo "Inserting 2 rows that should get IDs 30 and 31:"
insert into schema_names (schema_name) values ('bletch'),('baz');
commit;
select * from schema_names order by id;
\! echo "Final state of schema_names table. AUTO_INCREMENT value shows the next inserted row would get ID=32."
show create table schema_names\G
Running this code gives this output, condensed for brevity and with the most important points bolded:
=== Adjusting the Auto-Increment Limit for a Table ===
Query OK, 0 rows affected (0.01 sec)
Query OK, 4 rows affected (0.02 sec)
Records: 4 Duplicates: 0 Warnings: 0
Initial state of schema_names table. AUTO_INCREMENT is included in SHOW CREATE TABLE output.
Note how MySQL reserved a block of IDs, but only needed 4 of them in this transaction, so the next inserted values would get IDs 8 and 9.
*************************** 1. row ***************************
Table: schema_names
Create Table: CREATE TABLE `schema_names` (
`id` int(10) unsigned NOT NULL AUTO_INCREMENT,
`schema_name` varchar(64) CHARACTER SET utf8 NOT NULL,
PRIMARY KEY (`id`),
KEY `i_schema` (`schema_name`)
) ENGINE=InnoDB AUTO_INCREMENT=8 DEFAULT CHARSET=latin1
1 row in set (0.00 sec)
+----+--------------------+
| id | schema_name |
+----+--------------------+
| 1 | information_schema |
| 2 | mysql |
| 3 | performance_schema |
| 4 | test |
+----+--------------------+
4 rows in set (0.00 sec)
Inserting even a tiny amount of data can produce gaps in the ID sequence.
Query OK, 2 rows affected (0.00 sec)
Records: 2 Duplicates: 0 Warnings: 0
Query OK, 0 rows affected (0.00 sec)
Bumping auto-increment lower limit to 20 (fast mechanism):
Query OK, 0 rows affected (0.01 sec)
Records: 0 Duplicates: 0 Warnings: 0
Inserting 2 rows that should get IDs 20 and 21:
Query OK, 2 rows affected (0.00 sec)
Records: 2 Duplicates: 0 Warnings: 0
Query OK, 0 rows affected (0.00 sec)
Query OK, 0 rows affected (0.00 sec)
Bumping auto-increment lower limit to 30 (slow mechanism):
Query OK, 8 rows affected (0.02 sec)
Records: 8 Duplicates: 0 Warnings: 0
Inserting 2 rows that should get IDs 30 and 31:
Query OK, 2 rows affected (0.00 sec)
Records: 2 Duplicates: 0 Warnings: 0
Query OK, 0 rows affected (0.01 sec)
+----+--------------------+
| id | schema_name |
+----+--------------------+
| 1 | information_schema |
| 2 | mysql |
| 3 | performance_schema |
| 4 | test |
| 8 | eight |
| 9 | nine |
| 20 | foo |
| 21 | bar |
| 30 | bletch |
| 31 | baz |
+----+--------------------+
10 rows in set (0.00 sec)
Query OK, 0 rows affected (0.00 sec)
Final state of schema_names table. AUTO_INCREMENT value shows the next inserted row would get ID=32.
*************************** 1. row ***************************
Table: schema_names
Create Table: CREATE TABLE `schema_names` (
`id` int(10) unsigned NOT NULL AUTO_INCREMENT,
`schema_name` varchar(64) CHARACTER SET utf8 NOT NULL,
PRIMARY KEY (`id`),
KEY `i_schema` (`schema_name`)
) ENGINE=InnoDB AUTO_INCREMENT=32 DEFAULT CHARSET=latin1
1 row in set (0.00 sec)
Example 5.7. Controlling Concurrency with the LOCK Clause
This example shows how to use the LOCK clause
of the ALTER TABLE statement to
allow or deny concurrent access to the table while an online DDL
operation is in progress. The clause has settings that allow
queries and DML statements
(LOCK=NONE), just
queries
(LOCK=SHARED), or no concurrent access at all
(LOCK=EXCLUSIVE).
In one session, we run a succession of
ALTER TABLE statements to create
and drop an index, using different values for the
LOCK clause to see what happens with waiting
or deadlocking in either session. We are using the same
BIG_TABLE table as in previous examples,
starting with approximately 1.7 million rows. For illustration
purposes, we will index and query the
IS_NULLABLE column. (Although in real life it
would be silly to make an index for a tiny column with only 2
distinct values.)
mysql: desc big_table; +--------------------------+---------------------+------+-----+---------+----------------+ | Field | Type | Null | Key | Default | Extra | +--------------------------+---------------------+------+-----+---------+----------------+ | TABLE_CATALOG | varchar(512) | NO | | | | | TABLE_SCHEMA | varchar(64) | NO | | | | | TABLE_NAME | varchar(64) | NO | | | | | COLUMN_NAME | varchar(64) | NO | | | | | ORDINAL_POSITION | bigint(21) unsigned | NO | | 0 | | | COLUMN_DEFAULT | longtext | YES | | NULL | | | IS_NULLABLE | varchar(3) | NO | | | | ... +--------------------------+---------------------+------+-----+---------+----------------+ 21 rows in set (0.14 sec) mysql: alter table big_table add index i1(is_nullable); Query OK, 0 rows affected (20.71 sec) mysql: alter table big_table drop index i1; Query OK, 0 rows affected (0.02 sec) mysql: alter table big_table add index i1(is_nullable), lock=exclusive; Query OK, 0 rows affected (19.44 sec) mysql: alter table big_table drop index i1; Query OK, 0 rows affected (0.03 sec) mysql: alter table big_table add index i1(is_nullable), lock=shared; Query OK, 0 rows affected (16.71 sec) mysql: alter table big_table drop index i1; Query OK, 0 rows affected (0.05 sec) mysql: alter table big_table add index i1(is_nullable), lock=none; Query OK, 0 rows affected (12.26 sec) mysql: alter table big_table drop index i1; Query OK, 0 rows affected (0.01 sec) ... repeat statements like the above while running queries ... ... and DML statements at the same time in another session ...
Nothing dramatic happens in the session running the DDL
statements. Sometimes, an ALTER
TABLE takes unusually long because it is waiting for
another transaction to finish, when that transaction modified
the table during the DDL or queried the table before the DDL:
mysql: alter table big_table add index i1(is_nullable), lock=none; Query OK, 0 rows affected (59.27 sec) mysql: -- The previous ALTER took so long because it was waiting for all the concurrent mysql: -- transactions to commit or roll back. mysql: alter table big_table drop index i1; Query OK, 0 rows affected (41.05 sec) mysql: -- Even doing a SELECT on the table in the other session first causes mysql: -- the ALTER TABLE above to stall until the transaction mysql: -- surrounding the SELECT is committed or rolled back.
Here is the log from another session running concurrently, where
we issue queries and DML statements against the table before,
during, and after the DDL operations shown in the previous
listings. This first listing shows queries only. We expect the
queries to be allowed during DDL operations using
LOCK=NONE or LOCK=SHARED,
and for the query to wait until the DDL is finished if the
ALTER TABLE statement includes
LOCK=EXCLUSIVE.
mysql: show variables like 'autocommit'; +---------------+-------+ | Variable_name | Value | +---------------+-------+ | autocommit | ON | +---------------+-------+ 1 row in set (0.01 sec) mysql: -- A trial query before any ADD INDEX in the other session: mysql: -- Note: because autocommit is enabled, each mysql: -- transaction finishes immediately after the query. mysql: select distinct is_nullable from big_table; +-------------+ | is_nullable | +-------------+ | NO | | YES | +-------------+ 2 rows in set (4.49 sec) mysql: -- Index is being created with LOCK=EXCLUSIVE on the ALTER statement. mysql: -- The query waits until the DDL is finished before proceeding. mysql: select distinct is_nullable from big_table; +-------------+ | is_nullable | +-------------+ | NO | | YES | +-------------+ 2 rows in set (17.26 sec) mysql: -- Index is being created with LOCK=SHARED on the ALTER statement. mysql: -- The query returns its results while the DDL is in progress. mysql: -- The same thing happens with LOCK=NONE on the ALTER statement. mysql: select distinct is_nullable from big_table; +-------------+ | is_nullable | +-------------+ | NO | | YES | +-------------+ 2 rows in set (3.11 sec) mysql: -- Once the index is created, and with no DDL in progress, mysql: -- queries referencing the indexed column are very fast: mysql: select count(*) from big_table where is_nullable = 'YES'; +----------+ | count(*) | +----------+ | 411648 | +----------+ 1 row in set (0.20 sec) mysql: select distinct is_nullable from big_table; +-------------+ | is_nullable | +-------------+ | NO | | YES | +-------------+ 2 rows in set (0.00 sec)
Now in this concurrent session, we run some transactions
including DML statements, or a combination of DML statements and
queries. We use DELETE statements
to illustrate predictable, verifiable changes to the table.
Because the transactions in this part can span multiple
statements, we run these tests with
autocommit turned off.
mysql: set global autocommit = off; Query OK, 0 rows affected (0.00 sec) mysql: -- Count the rows that will be involved in our DELETE statements: mysql: select count(*) from big_table where is_nullable = 'YES'; +----------+ | count(*) | +----------+ | 411648 | +----------+ 1 row in set (0.95 sec) mysql: -- After this point, any DDL statements back in the other session mysql: -- stall until we commit or roll back. mysql: delete from big_table where is_nullable = 'YES' limit 11648; Query OK, 11648 rows affected (0.14 sec) mysql: select count(*) from big_table where is_nullable = 'YES'; +----------+ | count(*) | +----------+ | 400000 | +----------+ 1 row in set (1.04 sec) mysql: rollback; Query OK, 0 rows affected (0.09 sec) mysql: select count(*) from big_table where is_nullable = 'YES'; +----------+ | count(*) | +----------+ | 411648 | +----------+ 1 row in set (0.93 sec) mysql: -- OK, now we're going to try that during index creation with LOCK=NONE. mysql: delete from big_table where is_nullable = 'YES' limit 11648; Query OK, 11648 rows affected (0.21 sec) mysql: -- We expect that now there will be 400000 'YES' rows left: mysql: select count(*) from big_table where is_nullable = 'YES'; +----------+ | count(*) | +----------+ | 400000 | +----------+ 1 row in set (1.25 sec) mysql: -- In the other session, the ALTER TABLE is waiting before finishing, mysql: -- because _this_ transaction hasn't committed or rolled back yet. mysql: rollback; Query OK, 0 rows affected (0.11 sec) mysql: select count(*) from big_table where is_nullable = 'YES'; +----------+ | count(*) | +----------+ | 411648 | +----------+ 1 row in set (0.19 sec) mysql: -- The ROLLBACK left the table in the same state we originally found it. mysql: -- Now let's make a permanent change while the index is being created, mysql: -- again with ALTER TABLE ... , LOCK=NONE. mysql: -- First, commit so the DROP INDEX in the other shell can finish; mysql: -- the previous SELECT started a transaction that accessed the table. mysql: commit; Query OK, 0 rows affected (0.00 sec) mysql: -- Now we add the index back in the other shell, then issue DML in this one mysql: -- while the DDL is running. mysql: delete from big_table where is_nullable = 'YES' limit 11648; Query OK, 11648 rows affected (0.23 sec) mysql: commit; Query OK, 0 rows affected (0.01 sec) mysql: -- In the other shell, the ADD INDEX has finished. mysql: select count(*) from big_table where is_nullable = 'YES'; +----------+ | count(*) | +----------+ | 400000 | +----------+ 1 row in set (0.19 sec) mysql: -- At the point the new index is finished being created, it contains entries mysql: -- only for the 400000 'YES' rows left when all concurrent transactions are finished. mysql: mysql: -- Now we will run a similar test, while ALTER TABLE ... , LOCK=SHARED is running. mysql: -- We expect a query to complete during the ALTER TABLE, but for the DELETE mysql: -- to run into some kind of issue. mysql: commit; Query OK, 0 rows affected (0.00 sec) mysql: -- As expected, the query returns results while the LOCK=SHARED DDL is running: mysql: select count(*) from big_table where is_nullable = 'YES'; +----------+ | count(*) | +----------+ | 400000 | +----------+ 1 row in set (2.07 sec) mysql: -- The DDL in the other session is not going to finish until this transaction mysql: -- is committed or rolled back. If we tried a DELETE now and it waited because mysql: -- of LOCK=SHARED on the DDL, both transactions would wait forever (deadlock). mysql: -- MySQL detects this condition and cancels the attempted DML statement. mysql: delete from big_table where is_nullable = 'YES' limit 100000; ERROR 1213 (40001): Deadlock found when trying to get lock; try restarting transaction mysql: -- The transaction here is still going, so in the other shell, the ADD INDEX operation mysql: -- is waiting for this transaction to commit or roll back. mysql: rollback; Query OK, 0 rows affected (0.00 sec) mysql: -- Now let's try issuing a query and some DML, on one line, while running mysql: -- ALTER TABLE ... , LOCK=EXCLUSIVE in the other shell. mysql: -- Notice how even the query is held up until the DDL is finished. mysql: -- By the time the DELETE is issued, there is no conflicting access mysql: -- to the table and we avoid the deadlock error. mysql: select count(*) from big_table where is_nullable = 'YES'; delete from big_table where is_nullable = 'YES' limit 100000; +----------+ | count(*) | +----------+ | 400000 | +----------+ 1 row in set (15.98 sec) Query OK, 100000 rows affected (2.81 sec) mysql: select count(*) from big_table where is_nullable = 'YES'; +----------+ | count(*) | +----------+ | 300000 | +----------+ 1 row in set (0.17 sec) mysql: rollback; Query OK, 0 rows affected (1.36 sec) mysql: select count(*) from big_table where is_nullable = 'YES'; +----------+ | count(*) | +----------+ | 400000 | +----------+ 1 row in set (0.19 sec) mysql: commit; Query OK, 0 rows affected (0.00 sec) mysql: -- Next, we try ALTER TABLE ... , LOCK=EXCLUSIVE in the other session mysql: -- and only issue DML, not any query, in the concurrent transaction here. mysql: delete from big_table where is_nullable = 'YES' limit 100000; Query OK, 100000 rows affected (16.37 sec) mysql: -- That was OK because the ALTER TABLE did not have to wait for the transaction mysql: -- here to complete. The DELETE in this session waited until the index was ready. mysql: select count(*) from big_table where is_nullable = 'YES'; +----------+ | count(*) | +----------+ | 300000 | +----------+ 1 row in set (0.16 sec) mysql: commit; Query OK, 0 rows affected (0.00 sec)
In the preceding example listings, we learned that:
The
LOCKclause forALTER TABLEis set off from the rest of the statement by a comma.Online DDL operations might wait before starting, until any prior transactions that access the table are committed or rolled back.
Online DDL operations might wait before completing, until any concurrent transactions that access the table are committed or rolled back.
While an online DDL operation is running, concurrent queries are relatively straightforward, as long as the
ALTER TABLEstatement usesLOCK=NONEorLOCK=SHARED.Pay attention to whether
autocommitis turned on or off. If it is turned off, be careful to end transactions in other sessions (even just queries) before performing DDL operations on the table.With
LOCK=SHARED, concurrent transactions that mix queries and DML could encounter deadlock errors and have to be restarted after the DDL is finished.With
LOCK=NONE, concurrent transactions can freely mix queries and DML. The DDL operation waits until the concurrent transactions are committed or rolled back.With
LOCK=NONE, concurrent transactions can freely mix queries and DML, but those transactions wait until the DDL operation is finished before they can access the table.
Example 5.8. Schema Setup Code for Online DDL Experiments
You can create multiple indexes on a table with one
ALTER TABLE statement. This is
relatively efficient, because the clustered index of the table
needs to be scanned only once (although the data is sorted
separately for each new index). For example:
CREATE TABLE T1(A INT PRIMARY KEY, B INT, C CHAR(1)) ENGINE=InnoDB; INSERT INTO T1 VALUES (1,2,'a'), (2,3,'b'), (3,2,'c'), (4,3,'d'), (5,2,'e'); COMMIT; ALTER TABLE T1 ADD INDEX (B), ADD UNIQUE INDEX (C);
The above statements create table T1 with the
primary key on column A, insert several rows,
then build two new indexes on columns B and
C. If there were many rows inserted into
T1 before the ALTER
TABLE statement, this approach is much more efficient
than creating all the secondary indexes before loading the data.
Because dropping InnoDB secondary indexes also does not require
any copying of table data, it is equally efficient to drop
multiple indexes with a single ALTER
TABLE statement or multiple DROP
INDEX statements:
ALTER TABLE T1 DROP INDEX B, DROP INDEX C;
or:
DROP INDEX B ON T1; DROP INDEX C ON T1;
Example 5.9. Creating and Dropping the Primary Key
Restructuring the clustered
index for an InnoDB table always
requires copying the table data. Thus, it is best to define the
primary key when you
create a table, rather than issuing ALTER TABLE ... ADD
PRIMARY KEY later, to avoid rebuilding the table.
Defining a PRIMARY KEY later causes the data
to be copied, as in the following example:
CREATE TABLE T2 (A INT, B INT); INSERT INTO T2 VALUES (NULL, 1); ALTER TABLE T2 ADD PRIMARY KEY (B);
When you create a UNIQUE or PRIMARY
KEY index, MySQL must do some extra work. For
UNIQUE indexes, MySQL checks that the table
contains no duplicate values for the key. For a PRIMARY
KEY index, MySQL also checks that none of the
PRIMARY KEY columns contains a
NULL.
When you add a primary key using the
ALGORITHM=COPY clause, MySQL actually
converts NULL values in the associated
columns to default values: 0 for numbers, the empty string for
character-based columns and BLOBs, and January 1, 1975 for
dates. This is a non-standard behavior that Oracle recommends
you not rely on. Adding a primary key using
ALGORITHM=INPLACE is only allowed when the
SQL_MODE setting includes the
strict_trans_tables or
strict_all_tables flags; when the
SQL_MODEsetting is strict, ADD
PRIMARY KEY ... , ALGORITHM=INPLACE is allowed, but
the statement can still fail if the requested primary key
columns contain any NULL values. The
ALGORITHM=INPLACE behavior is more
standard-compliant.
The following example shows the different possibilities for the
ADD PRIMARY KEY clause. With the
ALGORITHM=COPY clause, the operation succeeds
despite the presence of NULL values in the
primary key columns; the data is silently changed, which could
cause problems. With the ALGORITHM=INPLACE
clause, the operation could fail for different reasons, because
this setting considers data integrity a high priority: the
statement gives an error if the
SQL_MODE setting is not
“strict” enough, or if the primary key columns
contain any NULL values. Once we address both
of those requirements, the ALTER
TABLE operation succeeds.
CREATE TABLE add_pk_via_copy (c1 INT, c2 VARCHAR(10), c3 DATETIME); INSERT INTO add_pk_via_copy VALUES (1,'a','...'),(NULL,NULL,NULL); ALTER TABLE add_pk_via_copy ADD PRIMARY KEY (c1,c2,c3), ALGORITHM=COPY; SELECT * FROM add_pk_via_copy; CREATE TABLE add_pk_via_inplace (c1 INT, c2 VARCHAR(10), c3 DATETIME); INSERT INTO add_pk_via_inplace VALUES (1,'a','...'),(NULL,NULL,NULL); SET sql_mode = 'strict_trans_tables'; ALTER TABLE add_pk_via_inplace ADD PRIMARY KEY (c1,c2,c3), ALGORITHM=COPY; SET sql_mode = ''; ALTER TABLE add_pk_via_inplace ADD PRIMARY KEY (c1,c2,c3), ALGORITHM=COPY; DELETE FROM add_pk_via_inplace WHERE c1 IS NULL OR c2 IS NULL OR c3 IS NULL; ALTER TABLE add_pk_via_inplace ADD PRIMARY KEY (c1,c2,c3), ALGORITHM=COPY; SELECT * FROM add_pk_via_inplace;
If you create a table without a primary key, InnoDB chooses one
for you, which can be the first UNIQUE key
defined on NOT NULL columns, or a
system-generated key. To avoid any uncertainty and the potential
space requirement for an extra hidden column, specify the
PRIMARY KEY clause as part of the
CREATE TABLE statement.
Each ALTER TABLE operation for an
InnoDB table is governed by several aspects:
Whether there is any change to the physical representation of the table, or whether it purely a change to metadata that can be done without touching the table itself.
Whether the volume of data in the table stays the same, increases, or decreases.
Whether a change in table data involves the clustered index, secondary indexes, or both.
Whether there are any foreign key relationships between the table being altered and some other table. The mechanics differ depending on whether the
foreign_key_checksconfiguration option is enabled or disabled.Whether the table is partitioned. Partitioning clauses of
ALTER TABLEare turned into low-level operations involving one or more tables, and those operations follow the regular rules for online DDL.Whether the table data must be copied, whether the table can be reorganized “in-place”, or a combination of both.
Whether the table contains any auto-increment columns.
What degree of locking is required, either by the nature of the underlying database operations, or a
LOCKclause that you specify in theALTER TABLEstatement.
This section explains how these factors affect the different kinds
of ALTER TABLE operations on
InnoDB tables.
Error Conditions for Online DDL
Here are the primary reasons why an online DDL operation could fail:
If a
LOCKclause specifies a low degree of locking (SHAREDorNONE) that is not compatible with the particular type of DDL operation.If a timeout occurs while waiting to get an exclusive lock on the table, which is needed briefly during the initial and final phases of the DDL operation.
If the
tmpdirfile system runs out of disk space, while MySQL writes temporary sort files on disk during index creation.If the
ALTER TABLEtakes so long, and concurrent DML modifies the table so much, that the size of the temporary online long exceeds the value of theinnodb_online_alter_log_max_sizeconfiguration option. This condition causes aDB_ONLINE_LOG_TOO_BIGerror.If concurrent DML makes changes to the table that are allowed with the original table definition, but not with the new one. The operation only fails at the very end, when MySQL tries to apply all the changes from concurrent DML statements. For example, you might insert duplicate values into a column while a unique index is being created, or you might insert
NULLvalues into a column while creating a primary key index on that column. The changes made by the concurrent DML take precedence, and theALTER TABLEoperation is effectively rolled back.
Although the configuration option
innodb_file_per_table has a
dramatic effect on the representation for an
InnoDB table, all online DDL operations work
equally well whether that option is enabled or disabled, and
whether the table is physically located in its own
.ibd file or inside the
system tablespace.
InnoDB has two types of indexes: the clustered index representing all the data in the table, and optional secondary indexes to speed up queries. Since the clustered index contains the data values in its B-tree nodes, adding or dropping a clustered index does involve copying the data, and creating a new copy of the table. A secondary index, however, contains only the index key and the value of the primary key. This type of index can be created or dropped without copying the data in the clustered index. Because each secondary index contains copies of the primary key values (used to access the clustered index when needed), when you change the definition of the primary key, all secondary indexes are recreated as well.
Dropping a secondary index is simple. Only the internal InnoDB system tables and the MySQL data dictionary tables are updated to reflect the fact that the index no longer exists. InnoDB returns the storage used for the index to the tablespace that contained it, so that new indexes or additional table rows can use the space.
To add a secondary index to an existing table, InnoDB scans the table, and sorts the rows using memory buffers and temporary files in order by the values of the secondary index key columns. The B-tree is then built in key-value order, which is more efficient than inserting rows into an index in random order. Because the B-tree nodes are split when they fill, building the index in this way results in a higher fill-factor for the index, making it more efficient for subsequent access.
Primary Key and Secondary Key Indexes
Historically, the MySQL server and InnoDB have
each kept their own metadata about table and index structures. The
MySQL server stores this information in
.frm files that are not
protected by a transactional mechanism, while
InnoDB has its own
data dictionary as
part of the system
tablespace. If a DDL operation was interrupted by a crash
or other unexpected event partway through, the metadata could be
left inconsistent between these two locations, causing problems
such as startup errors or inability to access the table that was
being altered. Now that InnoDB is the default
storage engine, addressing such issues is a high priority. These
enhancements to DDL operations reduce the window of opportunity
for such issues to occur.
Although no data is lost if the server crashes while an
ALTER TABLE statement is executing,
the crash recovery
process is different for
clustered indexes and
secondary indexes.
If the server crashes while creating an InnoDB secondary index,
upon recovery, MySQL drops any partially created indexes. You must
re-run the ALTER TABLE or
CREATE INDEX statement.
When a crash occurs during the creation of an InnoDB clustered index, recovery is more complicated, because the data in the table must be copied to an entirely new clustered index. Remember that all InnoDB tables are stored as clustered indexes. In the following discussion, we use the word table and clustered index interchangeably.
MySQL creates the new clustered index by copying the existing data from the original InnoDB table to a temporary table that has the desired index structure. Once the data is completely copied to this temporary table, the original table is renamed with a different temporary table name. The temporary table comprising the new clustered index is renamed with the name of the original table, and the original table is dropped from the database.
If a system crash occurs while creating a new clustered index, no data is lost, but you must complete the recovery process using the temporary tables that exist during the process. Since it is rare to re-create a clustered index or re-define primary keys on large tables, or to encounter a system crash during this operation, this manual does not provide information on recovering from this scenario.
With the exception of ALTER TABLE
partitioning clauses, online DDL operations for partitioned
InnoDB tables follow the same rules that apply
to regular InnoDB tables. Online DDL rules are
outlined in Table 5.8, “Summary of Online Status for DDL Operations”.
ALTER TABLE partitioning clauses do
not go through the same internal online DDL API as regular
non-partitioned InnoDB tables, and are only
allowed in conjunction with ALGORITHM=DEFAULT
and LOCK=DEFAULT.
If you use an ALTER TABLE partitioning clause in an
ALTER TABLE statement, the
partitioned table will be “re-partitioned” using the
ALTER TABLE COPY
algorithm. In other words, a new partitioned table is created with
the new partitioning scheme. The newly created table will include
any changes applied by the ALTER
TABLE statement and the table data will be copied into
the new table structure.
If you do not change the table's partitioning using
ALTER TABLE partitioning clauses or
perform any other partition management in your
ALTER TABLE statement,
ALTER TABLE will use the
INPLACE algorithm on each table partition. Be
aware, however, that when INPLACE
ALTER TABLE operations are
performed on each partition, there will be increased demand on
system resources due to operations being performed on multiple
partitions.
Even though partitioning clauses of the ALTER
TABLE statement do not go through the same internal
online DDL API as regular non-partitioned
InnoDB tables, MySQL still attempts to minimize
data copying and locking where possible:
ADD PARTITIONandDROP PARTITIONfor tables partitioned byRANGEorLISTdo not copy any existing data.TRUNCATE PARTITIONdoes not copy any existing data, for all types of partitioned tables.Concurrent queries are allowed during
ADD PARTITIONandCOALESCE PARTITIONfor tables partitioned byHASHorLIST. MySQL copies the data while holding a shared lock.For
REORGANIZE PARTITION,REBUILD PARTITION, orADD PARTITIONorCOALESCE PARTITIONfor a table partitioned byLINEAR HASHorLIST, concurrent queries are allowed. Data from the affected partitions is copied while holding a shared lock.
Full-text search (FTS) and foreign keys are not supported by
InnoDB partitioned tables. For more
information, see Section 12.9.5, “Full-Text Restrictions” and
Section 17.6.2, “Partitioning Limitations Relating to Storage Engines”.
Take the following limitations into account when running online DDL operations:
During an online DDL operation that copies the table, files are written to the temporary directory (
$TMPDIRon Unix,%TEMP%on Windows, or the directory specified by the--tmpdirconfiguration variable). Each temporary file is large enough to hold one column in the new table or index, and each one is removed as soon as it is merged into the final table or index.An
ALTER TABLEstatement that containsDROP INDEXandADD INDEXclauses that both name the same index uses a table copy, not Fast Index Creation.The table is copied, rather than using Fast Index Creation when you create an index on a
TEMPORARY TABLE. This has been reported as MySQL Bug #39833.InnoDB handles error cases when users attempt to drop indexes needed for foreign keys. See section Section 14.2.5.9, “Better Error Handling when Dropping Indexes” for details.
The
ALTER TABLEclauseLOCK=NONEis not allowed if there areON...CASCADEorON...SET NULLconstraints on the table.During each online DDL
ALTER TABLEstatement, regardless of theLOCKclause, there are brief periods at the beginning and end requiring an exclusive lock on the table (the same kind of lock specified by theLOCK=EXCLUSIVEclause). Thus, an online DDL operation might wait before starting if there is a long-running transaction performing inserts, updates, deletes, orSELECT ... FOR UPDATEon that table; and an online DDL operation might wait before finishing if a similar long-running transaction was started while theALTER TABLEwas in progress.When running an online
ALTER TABLEoperation, the thread that runs theALTER TABLEoperation will apply an “online log” of DML operations that were run concurrently on the same table from other connection threads. When the DML operations are applied, it is possible to encounter a duplicate key entry error (ERROR 1062 (23000): Duplicate entry), even if the duplicate entry is only temporary and would be reverted by a later entry in the “online log”. This is similar to the idea of a foreign key constraint check inInnoDBin which constraints must hold during a transaction.OPTIMIZE TABLEfor anInnoDBtable is mapped to anALTER TABLEoperation to rebuild the table and update index statistics and free unused space in the clustered index. This operation does not use fast index creation. Secondary indexes are not created as efficiently because keys are inserted in the order they appeared in the primary key.
In some cases, you might want to run multiple instances of MySQL on a single machine. You might want to test a new MySQL release while leaving an existing production setup undisturbed. Or you might want to give different users access to different mysqld servers that they manage themselves. (For example, you might be an Internet Service Provider that wants to provide independent MySQL installations for different customers.)
It is possible to use a different MySQL server binary per instance, or use the same binary for multiple instances, or any combination of the two approaches. For example, you might run a server from MySQL 5.6 and one from MySQL 5.7, to see how different versions handle a given workload. Or you might run multiple instances of the current production version, each managing a different set of databases.
Whether or not you use distinct server binaries, each instance that
you run must be configured with unique values for several operating
parameters. This eliminates the potential for conflict between
instances. Parameters can be set on the command line, in option
files, or by setting environment variables. See
Section 4.2.3, “Specifying Program Options”. To see the values used by a given
instance, connect to it and execute a SHOW
VARIABLES statement.
The primary resource managed by a MySQL instance is the data
directory. Each instance should use a different data directory, the
location of which is specified using the
--datadir=
option. For methods of configuring each instance with its own data
directory, and warnings about the dangers of failing to do so, see
Section 5.6.1, “Setting Up Multiple Data Directories”.
path
In addition to using different data directories, several other options must have different values for each server instance:
--portcontrols the port number for TCP/IP connections. Alternatively, if the host has multiple network addresses, you can use--bind-addressto cause each server to listen to a different address.--socketcontrols the Unix socket file path on Unix or the named pipe name on Windows. On Windows, it is necessary to specify distinct pipe names only for those servers configured to permit named-pipe connections.--shared-memory-base-name=nameThis option is used only on Windows. It designates the shared-memory name used by a Windows server to permit clients to connect using shared memory. It is necessary to specify distinct shared-memory names only for those servers configured to permit shared-memory connections.
This option indicates the path name of the file in which the server writes its process ID.
If you use the following log file options, their values must differ for each server:
For further discussion of log file options, see Section 5.2, “MySQL Server Logs”.
To achieve better performance, you can specify the following option differently for each server, to spread the load between several physical disks:
Having different temporary directories also makes it easier to determine which MySQL server created any given temporary file.
If you have multiple MySQL installations in different locations, you
can specify the base directory for each installation with the
--basedir=
option. This causes each instance to automatically use a different
data directory, log files, and PID file because the default for each
of those parameters is relative to the base directory. In that case,
the only other options you need to specify are the
path--socket and
--port options. Suppose that you
install different versions of MySQL using tar
file binary distributions. These install in different locations, so
you can start the server for each installation using the command
bin/mysqld_safe under its corresponding base
directory. mysqld_safe determines the proper
--basedir option to pass to
mysqld, and you need specify only the
--socket and
--port options to
mysqld_safe.
As discussed in the following sections, it is possible to start
additional servers by specifying appropriate command options or by
setting environment variables. However, if you need to run multiple
servers on a more permanent basis, it is more convenient to use
option files to specify for each server those option values that
must be unique to it. The
--defaults-file option is useful for
this purpose.
Each MySQL Instance on a machine should have its own data
directory. The location is specified using the
--datadir=
option.
path
There are different methods of setting up a data directory for a new instance:
Create a new data directory.
Copy an existing data directory.
The following discussion provides more detail about each method.
Normally, you should never have two servers that update data in the same databases. This may lead to unpleasant surprises if your operating system does not support fault-free system locking. If (despite this warning) you run multiple servers using the same data directory and they have logging enabled, you must use the appropriate options to specify log file names that are unique to each server. Otherwise, the servers try to log to the same files.
Even when the preceding precautions are observed, this kind of
setup works only with MyISAM and
MERGE tables, and not with any of the other
storage engines. Also, this warning against sharing a data
directory among servers always applies in an NFS environment.
Permitting multiple MySQL servers to access a common data
directory over NFS is a very bad idea. The
primary problem is that NFS is the speed bottleneck. It is not
meant for such use. Another risk with NFS is that you must
devise a way to ensure that two or more servers do not interfere
with each other. Usually NFS file locking is handled by the
lockd daemon, but at the moment there is no
platform that performs locking 100% reliably in every situation.
Create a New Data Directory
With this method, the data directory will be in the same state as when you first install MySQL. It will have the default set of MySQL accounts and no user data.
On Unix, initialize the data directory by running mysql_install_db. See Section 2.10.1, “Unix Postinstallation Procedures”.
On Windows, the data directory is included in the MySQL distribution:
MySQL Zip archive distributions for Windows contain an unmodified data directory. You can unpack such a distribution into a temporary location, then copy it
datadirectory to where you are setting up the new instance.Windows MSI package installers create and set up the data directory that the installed server will use, but also create a pristine “template” data directory named
dataunder the installation directory. After an installation has been performed using an MSI package, the template data directory can be copied to set up additional MySQL instances.
Copy an Existing Data Directory
With this method, any MySQL accounts or user data present in the data directory are carried over to the new data directory.
Stop the existing MySQL instance using the data directory. This must be a clean shutdown so that the instance flushes any pending changes to disk.
Copy the data directory to the location where the new data directory should be.
Copy the
my.cnformy.inioption file used by the existing instance. This serves as a basis for the new instance.Modify the new option file so that any pathnames referring to the original data directory refer to the new data directory. Also, modify any other options that must be unique per instance, such as the TCP/IP port number and the log files. For a list of parameters that must be unique per instance, see Section 5.6, “Running Multiple MySQL Instances on One Machine”.
Start the new instance, telling it to use the new option file.
You can run multiple servers on Windows by starting them manually from the command line, each with appropriate operating parameters, or by installing several servers as Windows services and running them that way. General instructions for running MySQL from the command line or as a service are given in Section 2.3, “Installing MySQL on Microsoft Windows”. The following sections describe how to start each server with different values for those options that must be unique per server, such as the data directory. These options are listed in Section 5.6, “Running Multiple MySQL Instances on One Machine”.
The procedure for starting a single MySQL server manually from
the command line is described in
Section 2.3.5.5, “Starting MySQL from the Windows Command Line”. To start multiple
servers this way, you can specify the appropriate options on the
command line or in an option file. It is more convenient to
place the options in an option file, but it is necessary to make
sure that each server gets its own set of options. To do this,
create an option file for each server and tell the server the
file name with a --defaults-file
option when you run it.
Suppose that you want to run mysqld on port
3307 with a data directory of C:\mydata1,
and mysqld-debug on port 3308 with a data
directory of C:\mydata2. Use this
procedure:
Make sure that each data directory exists, including its own copy of the
mysqldatabase that contains the grant tables.Create two option files. For example, create one file named
C:\my-opts1.cnfthat looks like this:[mysqld] datadir = C:/mydata1 port = 3307
Create a second file named
C:\my-opts2.cnfthat looks like this:[mysqld] datadir = C:/mydata2 port = 3308
Use the
--defaults-fileoption to start each server with its own option file:C:\>
C:\mysql\bin\mysqld --defaults-file=C:\my-opts1.cnfC:\>C:\mysql\bin\mysqld-debug --defaults-file=C:\my-opts2.cnfEach server starts in the foreground (no new prompt appears until the server exits later), so you will need to issue those two commands in separate console windows.
To shut down the servers, connect to each using the appropriate port number:
C:\>C:\mysql\bin\mysqladmin --port=3307 shutdownC:\>C:\mysql\bin\mysqladmin --port=3308 shutdown
Servers configured as just described permit clients to connect
over TCP/IP. If your version of Windows supports named pipes and
you also want to permit named-pipe connections, use the
mysqld or mysqld-debug
server and specify options that enable the named pipe and
specify its name. Each server that supports named-pipe
connections must use a unique pipe name. For example, the
C:\my-opts1.cnf file might be written like
this:
[mysqld] datadir = C:/mydata1 port = 3307 enable-named-pipe socket = mypipe1
Modify C:\my-opts2.cnf similarly for use by
the second server. Then start the servers as described
previously.
A similar procedure applies for servers that you want to permit
shared-memory connections. Enable such connections with the
--shared-memory option and
specify a unique shared-memory name for each server with the
--shared-memory-base-name option.
On Windows, a MySQL server can run as a Windows service. The procedures for installing, controlling, and removing a single MySQL service are described in Section 2.3.5.7, “Starting MySQL as a Windows Service”.
To set up multiple MySQL services, you must make sure that each instance uses a different service name in addition to the other parameters that must be unique per instance.
For the following instructions, suppose that you want to run the
mysqld server from two different versions of
MySQL that are installed at C:\mysql-5.5.9
and C:\mysql-5.7.3,
respectively. (This might be the case if you are running 5.5.9
as your production server, but also want to conduct tests using
5.7.3.)
To install MySQL as a Windows service, use the
--install or --install-manual
option. For information about these options, see
Section 2.3.5.7, “Starting MySQL as a Windows Service”.
Based on the preceding information, you have several ways to set up multiple services. The following instructions describe some examples. Before trying any of them, shut down and remove any existing MySQL services.
Approach 1: Specify the options for all services in one of the standard option files. To do this, use a different service name for each server. Suppose that you want to run the 5.5.9 mysqld using the service name of
mysqld1and the 5.7.3 mysqld using the service namemysqld2. In this case, you can use the[mysqld1]group for 5.5.9 and the[mysqld2]group for 5.7.3. For example, you can set upC:\my.cnflike this:# options for mysqld1 service [mysqld1] basedir = C:/mysql-5.5.9 port = 3307 enable-named-pipe socket = mypipe1 # options for mysqld2 service [mysqld2] basedir = C:/mysql-5.7.3 port = 3308 enable-named-pipe socket = mypipe2
Install the services as follows, using the full server path names to ensure that Windows registers the correct executable program for each service:
C:\>
C:\mysql-5.5.9\bin\mysqld --install mysqld1C:\>C:\mysql-5.7.3\bin\mysqld --install mysqld2To start the services, use the services manager, or use NET START with the appropriate service names:
C:\>
NET START mysqld1C:\>NET START mysqld2To stop the services, use the services manager, or use NET STOP with the appropriate service names:
C:\>
NET STOP mysqld1C:\>NET STOP mysqld2Approach 2: Specify options for each server in separate files and use
--defaults-filewhen you install the services to tell each server what file to use. In this case, each file should list options using a[mysqld]group.With this approach, to specify options for the 5.5.9 mysqld, create a file
C:\my-opts1.cnfthat looks like this:[mysqld] basedir = C:/mysql-5.5.9 port = 3307 enable-named-pipe socket = mypipe1
For the 5.7.3 mysqld, create a file
C:\my-opts2.cnfthat looks like this:[mysqld] basedir = C:/mysql-5.7.3 port = 3308 enable-named-pipe socket = mypipe2
Install the services as follows (enter each command on a single line):
C:\>
C:\mysql-5.5.9\bin\mysqld --install mysqld1--defaults-file=C:\my-opts1.cnfC:\>C:\mysql-5.7.3\bin\mysqld --install mysqld2--defaults-file=C:\my-opts2.cnfWhen you install a MySQL server as a service and use a
--defaults-fileoption, the service name must precede the option.After installing the services, start and stop them the same way as in the preceding example.
To remove multiple services, use mysqld
--remove for each one, specifying a service name
following the --remove option. If
the service name is the default (MySQL), you
can omit it.
One way is to run multiple MySQL instances on Unix is to compile different servers with different default TCP/IP ports and Unix socket files so that each one listens on different network interfaces. Compiling in different base directories for each installation also results automatically in a separate, compiled-in data directory, log file, and PID file location for each server.
Assume that an existing 5.6 server is configured for
the default TCP/IP port number (3306) and Unix socket file
(/tmp/mysql.sock). To configure a new
5.7.3 server to have different operating parameters,
use a CMake command something like this:
shell>cmake . -DMYSQL_TCP_PORT=port_number\-DMYSQL_UNIX_ADDR=file_name\-DCMAKE_INSTALL_PREFIX=/usr/local/mysql-5.7.3
Here, port_number and
file_name must be different from the
default TCP/IP port number and Unix socket file path name, and the
CMAKE_INSTALL_PREFIX value should
specify an installation directory different from the one under
which the existing MySQL installation is located.
If you have a MySQL server listening on a given port number, you can use the following command to find out what operating parameters it is using for several important configurable variables, including the base directory and Unix socket file name:
shell> mysqladmin --host=host_name --port=port_number variables
With the information displayed by that command, you can tell what option values not to use when configuring an additional server.
If you specify localhost as the host name,
mysqladmin defaults to using a Unix socket file
connection rather than TCP/IP. To explicitly specify the
connection protocol, use the
--protocol={TCP|SOCKET|PIPE|MEMORY}
option.
You need not compile a new MySQL server just to start with a different Unix socket file and TCP/IP port number. It is also possible to use the same server binary and start each invocation of it with different parameter values at runtime. One way to do so is by using command-line options:
shell> mysqld_safe --socket=file_name --port=port_number
To start a second server, provide different
--socket and
--port option values, and pass a
--datadir=
option to mysqld_safe so that the server uses a
different data directory.
path
Alternatively, put the options for each server in a different
option file, then start each server using a
--defaults-file option that
specifies the path to the appropriate option file. For example, if
the option files for two server instances are named
/usr/local/mysql/my.cnf and
/usr/local/mysql/my.cnf2, start the servers
like this: command:
shell>mysqld_safe --defaults-file=/usr/local/mysql/my.cnfshell>mysqld_safe --defaults-file=/usr/local/mysql/my.cnf2
Another way to achieve a similar effect is to use environment variables to set the Unix socket file name and TCP/IP port number:
shell>MYSQL_UNIX_PORT=/tmp/mysqld-new.sockshell>MYSQL_TCP_PORT=3307shell>export MYSQL_UNIX_PORT MYSQL_TCP_PORTshell>mysql_install_db --user=mysqlshell>mysqld_safe --datadir=/path/to/datadir &
This is a quick way of starting a second server to use for testing. The nice thing about this method is that the environment variable settings apply to any client programs that you invoke from the same shell. Thus, connections for those clients are automatically directed to the second server.
Section 2.12, “Environment Variables”, includes a list of other environment variables you can use to affect MySQL programs.
On Unix, the mysqld_multi script provides another way to start multiple servers. See Section 4.3.4, “mysqld_multi — Manage Multiple MySQL Servers”.
To connect with a client program to a MySQL server that is listening to different network interfaces from those compiled into your client, you can use one of the following methods:
Start the client with
--host=host_name--port=to connect using TCP/IP to a remote server, withport_number--host=127.0.0.1--port=to connect using TCP/IP to a local server, or withport_number--host=localhost--socket=to connect to a local server using a Unix socket file or a Windows named pipe.file_nameStart the client with
--protocol=TCPto connect using TCP/IP,--protocol=SOCKETto connect using a Unix socket file,--protocol=PIPEto connect using a named pipe, or--protocol=MEMORYto connect using shared memory. For TCP/IP connections, you may also need to specify--hostand--portoptions. For the other types of connections, you may need to specify a--socketoption to specify a Unix socket file or Windows named-pipe name, or a--shared-memory-base-nameoption to specify the shared-memory name. Shared-memory connections are supported only on Windows.On Unix, set the
MYSQL_UNIX_PORTandMYSQL_TCP_PORTenvironment variables to point to the Unix socket file and TCP/IP port number before you start your clients. If you normally use a specific socket file or port number, you can place commands to set these environment variables in your.loginfile so that they apply each time you log in. See Section 2.12, “Environment Variables”.Specify the default Unix socket file and TCP/IP port number in the
[client]group of an option file. For example, you can useC:\my.cnfon Windows, or the.my.cnffile in your home directory on Unix. See Section 4.2.3.3, “Using Option Files”.In a C program, you can specify the socket file or port number arguments in the
mysql_real_connect()call. You can also have the program read option files by callingmysql_options(). See Section 21.8.7, “C API Function Descriptions”.If you are using the Perl
DBD::mysqlmodule, you can read options from MySQL option files. For example:$dsn = "DBI:mysql:test;mysql_read_default_group=client;" . "mysql_read_default_file=/usr/local/mysql/data/my.cnf"; $dbh = DBI->connect($dsn, $user, $password);See Section 21.10, “MySQL Perl API”.
Other programming interfaces may provide similar capabilities for reading option files.
The DTrace probes in the MySQL server are designed to provide
information about the execution of queries within MySQL and the
different areas of the system being utilized during that process.
The organization and triggering of the probes means that the
execution of an entire query can be monitored with one level of
probes (query-start and
query-done) but by monitoring other probes you
can get successively more detailed information about the execution
of the query in terms of the locks used, sort methods and even
row-by-row and storage-engine level execution information.
The DTrace probes are organized so that you can follow the entire query process, from the point of connection from a client, through the query execution, row-level operations, and back out again. You can think of the probes as being fired within a specific sequence during a typical client connect/execute/disconnect sequence, as shown in the following figure.
Global information is provided in the arguments to the DTrace probes
at various levels. Global information, that is, the connection ID
and user/host and where relevant the query string, is provided at
key levels (connection-start,
command-start, query-start,
and query-exec-start). As you go deeper into the
probes, it is assumed either you are only interested in the
individual executions (row-level probes provide information on the
database and table name only), or that you will combine the
row-level probes with the notional parent probes to provide the
information about a specific query. Examples of this will be given
as the format and arguments of each probe are provided.
For more information on DTrace and writing DTrace scripts, read the DTrace User Guide.
MySQL 5.7 includes support for DTrace probes on Solaris
10 Update 5 (Solaris 5/08) on SPARC, x86 and x86_64 platforms.
Probes are also supported on Mac OS X 10.4 and higher. Enabling the
probes should be automatic on these platforms. To explicitly enable
or disable the probes during building, use the
-DENABLE_DTRACE=1 or
-DENABLE_DTRACE=0 option to
CMake.
- 5.7.1.1. Connection Probes
- 5.7.1.2. Command Probes
- 5.7.1.3. Query Probes
- 5.7.1.4. Query Parsing Probes
- 5.7.1.5. Query Cache Probes
- 5.7.1.6. Query Execution Probes
- 5.7.1.7. Row-Level Probes
- 5.7.1.8. Read Row Probes
- 5.7.1.9. Index Probes
- 5.7.1.10. Lock Probes
- 5.7.1.11. Filesort Probes
- 5.7.1.12. Statement Probes
- 5.7.1.13. Network Probes
- 5.7.1.14. Keycache Probes
MySQL supports the following static probes, organized into groups of functionality.
Table 5.9. MySQL DTrace Probes
| Group | Probes |
|---|---|
| Connection | connection-start, connection-done |
| Command | command-start, command-done |
| Query | query-start, query-done |
| Query Parsing | query-parse-start,
query-parse-done |
| Query Cache | query-cache-hit, query-cache-miss |
| Query Execution | query-exec-start, query-exec-done |
| Row Level | insert-row-start, insert-row-done |
update-row-start, update-row-done | |
delete-row-start, delete-row-done | |
| Row Reads | read-row-start, read-row-done |
| Index Reads | index-read-row-start,
index-read-row-done |
| Lock | handler-rdlock-start,
handler-rdlock-done |
handler-wrlock-start,
handler-wrlock-done | |
handler-unlock-start,
handler-unlock-done | |
| Filesort | filesort-start, filesort-done |
| Statement | select-start, select-done |
insert-start, insert-done | |
insert-select-start,
insert-select-done | |
update-start, update-done | |
multi-update-start,
multi-update-done | |
delete-start, delete-done | |
multi-delete-start,
multi-delete-done | |
| Network | net-read-start, net-read-done,
net-write-start,
net-write-done |
| Keycache | keycache-read-start,
keycache-read-block,
keycache-read-done,
keycache-read-hit,
keycache-read-miss,
keycache-write-start,
keycache-write-block,
keycache-write-done |
When extracting the argument data from the probes, each argument
is available as
arg, starting
with Narg0. To identify each argument within
the definitions they are provided with a descriptive name, but
you must access the information using the corresponding
arg parameter.
N
The connection-start and
connection-done probes enclose a connection
from a client, regardless of whether the connection is through a
socket or network connection.
connection-start(connectionid, user, host) connection-done(status, connectionid)
connection-start: Triggered after a connection and successful login/authentication have been completed by a client. The arguments contain the connection information:connectionid: Anunsigned longcontaining the connection ID. This is the same as the process ID shown as theIdvalue in the output fromSHOW PROCESSLIST.user: The username used when authenticating. The value will be blank for the anonymous user.host: The host of the client connection. For a connection made using UNIX sockets, the value will be blank.
connection-done: Triggered just as the connection to the client has been closed. The arguments are:status: The status of the connection when it was closed. A logout operation will have a value of 0; any other termination of the connection has a nonzero value.connectionid: The connection ID of the connection that was closed.
The following D script will quantify and summarize the average duration of individual connections, and provide a count, dumping the information every 60 seconds:
#!/usr/sbin/dtrace -s
mysql*:::connection-start
{
self->start = timestamp;
}
mysql*:::connection-done
/self->start/
{
@ = quantize(((timestamp - self->start)/1000000));
self->start = 0;
}
tick-60s
{
printa(@);
}When executed on a server with a large number of clients you might see output similar to this:
1 57413 :tick-60s
value ------------- Distribution ------------- count
-1 | 0
0 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ 30011
1 | 59
2 | 5
4 | 20
8 | 29
16 | 18
32 | 27
64 | 30
128 | 11
256 | 10
512 | 1
1024 | 6
2048 | 8
4096 | 9
8192 | 8
16384 | 2
32768 | 1
65536 | 1
131072 | 0
262144 | 1
524288 | 0
The command probes are executed before and after a client
command is executed, including any SQL statement that might be
executed during that period. Commands include operations such as
the initialization of the DB, use of the
COM_CHANGE_USER operation (supported by the
MySQL protocol), and manipulation of prepared statements. Many
of these commands are used only by the MySQL client API from
various connectors such as PHP and Java.
command-start(connectionid, command, user, host) command-done(status)
command-start: Triggered when a command is submitted to the server.connectionid: The connection ID of the client executing the command.command: An integer representing the command that was executed. Possible values are shown in the following table.Value Name Description 00 COM_SLEEP Internal thread state 01 COM_QUIT Close connection 02 COM_INIT_DB Select database ( USE ...)03 COM_QUERY Execute a query 04 COM_FIELD_LIST Get a list of fields 05 COM_CREATE_DB Create a database (deprecated) 06 COM_DROP_DB Drop a database (deprecated) 07 COM_REFRESH Refresh connection 08 COM_SHUTDOWN Shutdown server 09 COM_STATISTICS Get statistics 10 COM_PROCESS_INFO Get processes ( SHOW PROCESSLIST)11 COM_CONNECT Initialize connection 12 COM_PROCESS_KILL Kill process 13 COM_DEBUG Get debug information 14 COM_PING Ping 15 COM_TIME Internal thread state 16 COM_DELAYED_INSERT Internal thread state 17 COM_CHANGE_USER Change user 18 COM_BINLOG_DUMP Used by a replication slave or mysqlbinlog to initiate a binary log read 19 COM_TABLE_DUMP Used by a replication slave to get the master table information 20 COM_CONNECT_OUT Used by a replication slave to log a connection to the server 21 COM_REGISTER_SLAVE Used by a replication slave during registration 22 COM_STMT_PREPARE Prepare a statement 23 COM_STMT_EXECUTE Execute a statement 24 COM_STMT_SEND_LONG_DATA Used by a client when requesting extended data 25 COM_STMT_CLOSE Close a prepared statement 26 COM_STMT_RESET Reset a prepared statement 27 COM_SET_OPTION Set a server option 28 COM_STMT_FETCH Fetch a prepared statement user: The user executing the command.host: The client host.
command-done: Triggered when the command execution completes. Thestatusargument contains 0 if the command executed successfully, or 1 if the statement was terminated before normal completion.
The command-start and
command-done probes are best used when
combined with the statement probes to get an idea of overall
execution time.
The query-start and
query-done probes are triggered when a
specific query is received by the server and when the query has
been completed and the information has been successfully sent to
the client.
query-start(query, connectionid, database, user, host) query-done(status)
query-start: Triggered after the query string has been received from the client. The arguments are:query: The full text of the submitted query.connectionid: The connection ID of the client that submitted the query. The connection ID equals the connection ID returned when the client first connects and theIdvalue in the output fromSHOW PROCESSLIST.database: The database name on which the query is being executed.user: The username used to connect to the server.host: The hostname of the client.
query-done: Triggered once the query has been executed and the information has been returned to the client. The probe includes a single argument,status, which returns 0 when the query is successfully executed and 1 if there was an error.
You can get a simple report of the execution time for each query using the following D script:
#!/usr/sbin/dtrace -s
#pragma D option quiet
dtrace:::BEGIN
{
printf("%-20s %-20s %-40s %-9s\n", "Who", "Database", "Query", "Time(ms)");
}
mysql*:::query-start
{
self->query = copyinstr(arg0);
self->connid = arg1;
self->db = copyinstr(arg2);
self->who = strjoin(copyinstr(arg3),strjoin("@",copyinstr(arg4)));
self->querystart = timestamp;
}
mysql*:::query-done
{
printf("%-20s %-20s %-40s %-9d\n",self->who,self->db,self->query,
(timestamp - self->querystart) / 1000000);
}When executing the above script you should get a basic idea of the execution time of your queries:
shell> ./query.d Who Database Query Time(ms) root@localhost test select * from t1 order by i limit 10 0 root@localhost test set global query_cache_size=0 0 root@localhost test select * from t1 order by i limit 10 776 root@localhost test select * from t1 order by i limit 10 773 root@localhost test select * from t1 order by i desc limit 10 795
The query parsing probes are triggered before the original SQL statement is parsed and when the parsing of the statement and determination of the execution model required to process the statement has been completed:
query-parse-start(query) query-parse-done(status)
query-parse-start: Triggered just before the statement is parsed by the MySQL query parser. The single argument,query, is a string containing the full text of the original query.query-parse-done: Triggered when the parsing of the original statement has been completed. Thestatusis an integer describing the status of the operation. A0indicates that the query was successfully parsed. A1indicates that the parsing of the query failed.
For example, you could monitor the execution time for parsing a given query using the following D script:
#!/usr/sbin/dtrace -s
#pragma D option quiet
mysql*:::query-parse-start
{
self->parsestart = timestamp;
self->parsequery = copyinstr(arg0);
}
mysql*:::query-parse-done
/arg0 == 0/
{
printf("Parsing %s: %d microseconds\n", self->parsequery,((timestamp - self->parsestart)/1000));
}
mysql*:::query-parse-done
/arg0 != 0/
{
printf("Error parsing %s: %d microseconds\n", self->parsequery,((timestamp - self->parsestart)/1000));
}
In the above script a predicate is used on
query-parse-done so that different output is
generated based on the status value of the probe.
When running the script and monitoring the execution:
shell> ./query-parsing.d Error parsing select from t1 join (t2) on (t1.i = t2.i) order by t1.s,t1.i limit 10: 36 ms Parsing select * from t1 join (t2) on (t1.i = t2.i) order by t1.s,t1.i limit 10: 176 ms
The query cache probes are fired when executing any query. The
query-cache-hit query is triggered when a
query exists in the query cache and can be used to return the
query cache information. The arguments contain the original
query text and the number of rows returned from the query cache
for the query. If the query is not within the query cache, or
the query cache is not enabled, then the
query-cache-miss probe is triggered instead.
query-cache-hit(query, rows) query-cache-miss(query)
query-cache-hit: Triggered when the query has been found within the query cache. The first argument,query, contains the original text of the query. The second argument,rows, is an integer containing the number of rows in the cached query.query-cache-miss: Triggered when the query is not found within the query cache. The first argument,query, contains the original text of the query.
The query cache probes are best combined with a probe on the main query so that you can determine the differences in times between using or not using the query cache for specified queries. For example, in the following D script, the query and query cache information are combined into the information output during monitoring:
#!/usr/sbin/dtrace -s
#pragma D option quiet
dtrace:::BEGIN
{
printf("%-20s %-20s %-40s %2s %-9s\n", "Who", "Database", "Query", "QC", "Time(ms)");
}
mysql*:::query-start
{
self->query = copyinstr(arg0);
self->connid = arg1;
self->db = copyinstr(arg2);
self->who = strjoin(copyinstr(arg3),strjoin("@",copyinstr(arg4)));
self->querystart = timestamp;
self->qc = 0;
}
mysql*:::query-cache-hit
{
self->qc = 1;
}
mysql*:::query-cache-miss
{
self->qc = 0;
}
mysql*:::query-done
{
printf("%-20s %-20s %-40s %-2s %-9d\n",self->who,self->db,self->query,(self->qc ? "Y" : "N"),
(timestamp - self->querystart) / 1000000);
}When executing the script you can see the effects of the query cache. Initially the query cache is disabled. If you set the query cache size and then execute the query multiple times you should see that the query cache is being used to return the query data:
shell> ./query-cache.d root@localhost test select * from t1 order by i limit 10 N 1072 root@localhost set global query_cache_size=262144 N 0 root@localhost test select * from t1 order by i limit 10 N 781 root@localhost test select * from t1 order by i limit 10 Y 0
The query execution probe is triggered when the actual execution of the query starts, after the parsing and checking the query cache but before any privilege checks or optimization. By comparing the difference between the start and done probes you can monitor the time actually spent servicing the query (instead of just handling the parsing and other elements of the query).
query-exec-start(query, connectionid, database, user, host, exec_type) query-exec-done(status)
The information provided in the arguments for
query-start and
query-exec-start are almost identical and
designed so that you can choose to monitor either the entire
query process (using query-start) or only
the execution (using query-exec-start)
while exposing the core information about the user, client,
and query being executed.
query-exec-start: Triggered when the execution of a individual query is started. The arguments are:query: The full text of the submitted query.connectionid: The connection ID of the client that submitted the query. The connection ID equals the connection ID returned when the client first connects and theIdvalue in the output fromSHOW PROCESSLIST.database: The database name on which the query is being executed.user: The username used to connect to the server.host: The hostname of the client.exec_type: The type of execution. Execution types are determined based on the contents of the query and where it was submitted. The values for each type are shown in the following table.Value Description 0 Executed query from sql_parse, top-level query. 1 Executed prepared statement 2 Executed cursor statement 3 Executed query in stored procedure
query-exec-done: Triggered when the execution of the query has completed. The probe includes a single argument,status, which returns 0 when the query is successfully executed and 1 if there was an error.
The *row-{start,done} probes are triggered
each time a row operation is pushed down to a storage engine.
For example, if you execute an
INSERT statement with 100 rows of
data, then the insert-row-start and
insert-row-done probes will be triggered 100
times each, for each row insert.
insert-row-start(database, table) insert-row-done(status) update-row-start(database, table) update-row-done(status) delete-row-start(database, table) delete-row-done(status)
insert-row-start: Triggered before a row is inserted into a table.insert-row-done: Triggered after a row is inserted into a table.update-row-start: Triggered before a row is updated in a table.update-row-done: Triggered before a row is updated in a table.delete-row-start: Triggered before a row is deleted from a table.delete-row-done: Triggered before a row is deleted from a table.
The arguments supported by the probes are consistent for the
corresponding start and
done probes in each case:
database: The database name.table: The table name.status: The status; 0 for success or 1 for failure.
Because the row-level probes are triggered for each individual
row access, these probes can be triggered many thousands of
times each second, which may have a detrimental effect on both
the monitoring script and MySQL. The DTrace environment should
limit the triggering on these probes to prevent the performance
being adversely affected. Either use the probes sparingly, or
use counter or aggregation functions to report on these probes
and then provide a summary when the script terminates or as part
of a query-done or
query-exec-done probes.
The following example script summarizes the duration of each row operation within a larger query:
#!/usr/sbin/dtrace -s
#pragma D option quiet
dtrace:::BEGIN
{
printf("%-2s %-10s %-10s %9s %9s %-s \n",
"St", "Who", "DB", "ConnID", "Dur ms", "Query");
}
mysql*:::query-start
{
self->query = copyinstr(arg0);
self->who = strjoin(copyinstr(arg3),strjoin("@",copyinstr(arg4)));
self->db = copyinstr(arg2);
self->connid = arg1;
self->querystart = timestamp;
self->rowdur = 0;
}
mysql*:::query-done
{
this->elapsed = (timestamp - self->querystart) /1000000;
printf("%2d %-10s %-10s %9d %9d %s\n",
arg0, self->who, self->db,
self->connid, this->elapsed, self->query);
}
mysql*:::query-done
/ self->rowdur /
{
printf("%34s %9d %s\n", "", (self->rowdur/1000000), "-> Row ops");
}
mysql*:::insert-row-start
{
self->rowstart = timestamp;
}
mysql*:::delete-row-start
{
self->rowstart = timestamp;
}
mysql*:::update-row-start
{
self->rowstart = timestamp;
}
mysql*:::insert-row-done
{
self->rowdur += (timestamp-self->rowstart);
}
mysql*:::delete-row-done
{
self->rowdur += (timestamp-self->rowstart);
}
mysql*:::update-row-done
{
self->rowdur += (timestamp-self->rowstart);
}Running the above script with a query that inserts data into a table, you can monitor the exact time spent performing the raw row insertion:
St Who DB ConnID Dur ms Query 0 @localhost test 13 20767 insert into t1(select * from t2) 4827 -> Row ops
The read row probes are triggered at a storage engine level each
time a row read operation occurs. These probes are specified
within each storage engine (as opposed to the
*row-start probes which are in the storage
engine interface). These probes can therefore be used to monitor
individual storage engine row-level operations and performance.
Because these probes are triggered around the storage engine row
read interface, they may be hit a significant number of times
during a basic query.
read-row-start(database, table, scan_flag) read-row-done(status)
read-row-start: Triggered when a row is read by the storage engine from the specifieddatabaseandtable. Thescan_flagis set to 1 (true) when the read is part of a table scan (that is, a sequential read), or 0 (false) when the read is of a specific record.read-row-done: Triggered when a row read operation within a storage engine completes. Thestatusreturns 0 on success, or a positive value on failure.
The index probes are triggered each time a a row is read using one of the indexes for the specified table. The probe is triggered within the corresponding storage engine for the table.
index-read-row-start(database, table) index-read-row-done(status)
index-read-row-start: Triggered when a row is read by the storage engine from the specifieddatabaseandtable.index-read-row-done: Triggered when an indexed row read operation within a storage engine completes. Thestatusreturns 0 on success, or a positive value on failure.
The lock probes are called whenever an external lock is requested by MySQL for a table using the corresponding lock mechanism on the table as defined by the table's engine type. There are three different types of lock, the read lock, write lock, and unlock operations. Using the probes you can determine the duration of the external locking routine (that is, the time taken by the storage engine to implement the lock, including any time waiting for another lock to become free) and the total duration of the lock/unlock process.
handler-rdlock-start(database, table) handler-rdlock-done(status) handler-wrlock-start(database, table) handler-wrlock-done(status) handler-unlock-start(database, table) handler-unlock-done(status)
handler-rdlock-start: Triggered when a read lock is requested on the specifieddatabaseandtable.handler-wrlock-start: Triggered when a write lock is requested on the specifieddatabaseandtable.handler-unlock-start: Triggered when an unlock request is made on the specifieddatabaseandtable.handler-rdlock-done: Triggered when a read lock request completes. Thestatusis 0 if the lock operation succeeded, or>0on failure.handler-wrlock-done: Triggered when a write lock request completes. Thestatusis 0 if the lock operation succeeded, or>0on failure.handler-unlock-done: Triggered when an unlock request completes. Thestatusis 0 if the unlock operation succeeded, or>0on failure.
You can use arrays to monitor the locking and unlocking of individual tables and then calculate the duration of the entire table lock using the following script:
#!/usr/sbin/dtrace -s
#pragma D option quiet
mysql*:::handler-rdlock-start
{
self->rdlockstart = timestamp;
this->lockref = strjoin(copyinstr(arg0),strjoin("@",copyinstr(arg1)));
self->lockmap[this->lockref] = self->rdlockstart;
printf("Start: Lock->Read %s.%s\n",copyinstr(arg0),copyinstr(arg1));
}
mysql*:::handler-wrlock-start
{
self->wrlockstart = timestamp;
this->lockref = strjoin(copyinstr(arg0),strjoin("@",copyinstr(arg1)));
self->lockmap[this->lockref] = self->rdlockstart;
printf("Start: Lock->Write %s.%s\n",copyinstr(arg0),copyinstr(arg1));
}
mysql*:::handler-unlock-start
{
self->unlockstart = timestamp;
this->lockref = strjoin(copyinstr(arg0),strjoin("@",copyinstr(arg1)));
printf("Start: Lock->Unlock %s.%s (%d ms lock duration)\n",
copyinstr(arg0),copyinstr(arg1),
(timestamp - self->lockmap[this->lockref])/1000000);
}
mysql*:::handler-rdlock-done
{
printf("End: Lock->Read %d ms\n",
(timestamp - self->rdlockstart)/1000000);
}
mysql*:::handler-wrlock-done
{
printf("End: Lock->Write %d ms\n",
(timestamp - self->wrlockstart)/1000000);
}
mysql*:::handler-unlock-done
{
printf("End: Lock->Unlock %d ms\n",
(timestamp - self->unlockstart)/1000000);
}When executed, you should get information both about the duration of the locking process itself, and of the locks on a specific table:
Start: Lock->Read test.t2 End: Lock->Read 0 ms Start: Lock->Unlock test.t2 (25743 ms lock duration) End: Lock->Unlock 0 ms Start: Lock->Read test.t2 End: Lock->Read 0 ms Start: Lock->Unlock test.t2 (1 ms lock duration) End: Lock->Unlock 0 ms Start: Lock->Read test.t2 End: Lock->Read 0 ms Start: Lock->Unlock test.t2 (1 ms lock duration) End: Lock->Unlock 0 ms Start: Lock->Read test.t2 End: Lock->Read 0 ms
The filesort probes are triggered whenever a filesort operation
is applied to a table. For more information on filesort and the
conditions under which it occurs, see
Section 8.2.1.15, “ORDER BY Optimization”.
filesort-start(database, table) filesort-done(status, rows)
filesort-start: Triggered when the filesort operation starts on a table. The two arguments to the probe,databaseandtable, will identify the table being sorted.filesort-done: Triggered when the filesort operation completes. Two arguments are supplied, thestatus(0 for success, 1 for failure), and the number of rows sorted during the filesort process.
An example of this is in the following script, which tracks the duration of the filesort process in addition to the duration of the main query:
#!/usr/sbin/dtrace -s
#pragma D option quiet
dtrace:::BEGIN
{
printf("%-2s %-10s %-10s %9s %18s %-s \n",
"St", "Who", "DB", "ConnID", "Dur microsec", "Query");
}
mysql*:::query-start
{
self->query = copyinstr(arg0);
self->who = strjoin(copyinstr(arg3),strjoin("@",copyinstr(arg4)));
self->db = copyinstr(arg2);
self->connid = arg1;
self->querystart = timestamp;
self->filesort = 0;
self->fsdb = "";
self->fstable = "";
}
mysql*:::filesort-start
{
self->filesort = timestamp;
self->fsdb = copyinstr(arg0);
self->fstable = copyinstr(arg1);
}
mysql*:::filesort-done
{
this->elapsed = (timestamp - self->filesort) /1000;
printf("%2d %-10s %-10s %9d %18d Filesort on %s\n",
arg0, self->who, self->fsdb,
self->connid, this->elapsed, self->fstable);
}
mysql*:::query-done
{
this->elapsed = (timestamp - self->querystart) /1000;
printf("%2d %-10s %-10s %9d %18d %s\n",
arg0, self->who, self->db,
self->connid, this->elapsed, self->query);
}
Executing a query on a large table with an ORDER
BY clause that triggers a filesort, and then creating
an index on the table and then repeating the same query, you can
see the difference in execution speed:
St Who DB ConnID Dur microsec Query 0 @localhost test 14 11335469 Filesort on t1 0 @localhost test 14 11335787 select * from t1 order by i limit 100 0 @localhost test 14 466734378 create index t1a on t1 (i) 0 @localhost test 14 26472 select * from t1 order by i limit 100
The individual statement probes are provided to give specific
information about different statement types. For the start
probes the string of the query is provided as a the only
argument. Depending on the statement type, the information
provided by the corresponding done probe will differ. For all
done probes the status of the operation (0
for success, >0 for failure) is provided.
For SELECT,
INSERT,
INSERT ... (SELECT FROM
...), DELETE, and
DELETE FROM
t1,t2 operations the number of rows affected is
returned.
For UPDATE and
UPDATE t1,t2
... statements the number of rows matched and the
number of rows actually changed is provided. This is because the
number of rows actually matched by the corresponding
WHERE clause, and the number of rows changed
can differ. MySQL does not update the value of a row if the
value already matches the new setting.
select-start(query) select-done(status,rows) insert-start(query) insert-done(status,rows) insert-select-start(query) insert-select-done(status,rows) update-start(query) update-done(status,rowsmatched,rowschanged) multi-update-start(query) multi-update-done(status,rowsmatched,rowschanged) delete-start(query) delete-done(status,rows) multi-delete-start(query) multi-delete-done(status,rows)
select-start: Triggered before aSELECTstatement.select-done: Triggered at the end of aSELECTstatement.insert-start: Triggered before aINSERTstatement.insert-done: Triggered at the end of anINSERTstatement.insert-select-start: Triggered before anINSERT ... SELECTstatement.insert-select-done: Triggered at the end of anINSERT ... SELECTstatement.update-start: Triggered before anUPDATEstatement.update-done: Triggered at the end of anUPDATEstatement.multi-update-start: Triggered before anUPDATEstatement involving multiple tables.multi-update-done: Triggered at the end of anUPDATEstatement involving multiple tables.delete-start: Triggered before aDELETEstatement.delete-done: Triggered at the end of aDELETEstatement.multi-delete-start: Triggered before aDELETEstatement involving multiple tables.multi-delete-done: Triggered at the end of aDELETEstatement involving multiple tables.
The arguments for the statement probes are:
query: The query string.status: The status of the query.0for success, and>0for failure.rows: The number of rows affected by the statement. This returns the number rows found forSELECT, the number of rows deleted forDELETE, and the number of rows successfully inserted forINSERT.rowsmatched: The number of rows matched by theWHEREclause of anUPDATEoperation.rowschanged: The number of rows actually changed during anUPDATEoperation.
You use these probes to monitor the execution of these statement types without having to monitor the user or client executing the statements. A simple example of this is to track the execution times:
#!/usr/sbin/dtrace -s
#pragma D option quiet
dtrace:::BEGIN
{
printf("%-60s %-8s %-8s %-8s\n", "Query", "RowsU", "RowsM", "Dur (ms)");
}
mysql*:::update-start, mysql*:::insert-start,
mysql*:::delete-start, mysql*:::multi-delete-start,
mysql*:::multi-delete-done, mysql*:::select-start,
mysql*:::insert-select-start, mysql*:::multi-update-start
{
self->query = copyinstr(arg0);
self->querystart = timestamp;
}
mysql*:::insert-done, mysql*:::select-done,
mysql*:::delete-done, mysql*:::multi-delete-done, mysql*:::insert-select-done
/ self->querystart /
{
this->elapsed = ((timestamp - self->querystart)/1000000);
printf("%-60s %-8d %-8d %d\n",
self->query,
0,
arg1,
this->elapsed);
self->querystart = 0;
}
mysql*:::update-done, mysql*:::multi-update-done
/ self->querystart /
{
this->elapsed = ((timestamp - self->querystart)/1000000);
printf("%-60s %-8d %-8d %d\n",
self->query,
arg1,
arg2,
this->elapsed);
self->querystart = 0;
}When executed you can see the basic execution times and rows matches:
Query RowsU RowsM Dur (ms) select * from t2 0 275 0 insert into t2 (select * from t2) 0 275 9 update t2 set i=5 where i > 75 110 110 8 update t2 set i=5 where i < 25 254 134 12 delete from t2 where i < 5 0 0 0
Another alternative is to use the aggregation functions in DTrace to aggregate the execution time of individual statements together:
#!/usr/sbin/dtrace -s
#pragma D option quiet
mysql*:::update-start, mysql*:::insert-start,
mysql*:::delete-start, mysql*:::multi-delete-start,
mysql*:::multi-delete-done, mysql*:::select-start,
mysql*:::insert-select-start, mysql*:::multi-update-start
{
self->querystart = timestamp;
}
mysql*:::select-done
{
@statements["select"] = sum(((timestamp - self->querystart)/1000000));
}
mysql*:::insert-done, mysql*:::insert-select-done
{
@statements["insert"] = sum(((timestamp - self->querystart)/1000000));
}
mysql*:::update-done, mysql*:::multi-update-done
{
@statements["update"] = sum(((timestamp - self->querystart)/1000000));
}
mysql*:::delete-done, mysql*:::multi-delete-done
{
@statements["delete"] = sum(((timestamp - self->querystart)/1000000));
}
tick-30s
{
printa(@statements);
}The script just shown aggregates the times spent doing each operation, which could be used to help benchmark a standard suite of tests.
delete 0 update 0 insert 23 select 2484 delete 0 update 0 insert 39 select 10744 delete 0 update 26 insert 56 select 10944 delete 0 update 26 insert 2287 select 15985
The network probes monitor the transfer of information from the MySQL server and clients of all types over the network. The probes are defined as follows:
net-read-start() net-read-done(status, bytes) net-write-start(bytes) net-write-done(status)
net-read-start: Triggered when a network read operation is started.net-read-done: Triggered when the network read operation completes. Thestatusis anintegerrepresenting the return status for the operation,0for success and1for failure. Thebytesargument is an integer specifying the number of bytes read during the process.net-start-bytes: Triggered when data is written to a network socket. The single argument,bytes, specifies the number of bytes written to the network socket.net-write-done: Triggered when the network write operation has completed. The single argument,status, is an integer representing the return status for the operation,0for success and1for failure.
You can use the network probes to monitor the time spent reading
from and writing to network clients during execution. The
following D script provides an example of this. Both the
cumulative time for the read or write is calculated, and the
number of bytes. Note that the dynamic variable size has been
increased (using the dynvarsize option) to
cope with the rapid firing of the individual probes for the
network reads/writes.
#!/usr/sbin/dtrace -s
#pragma D option quiet
#pragma D option dynvarsize=4m
dtrace:::BEGIN
{
printf("%-2s %-30s %-10s %9s %18s %-s \n",
"St", "Who", "DB", "ConnID", "Dur microsec", "Query");
}
mysql*:::query-start
{
self->query = copyinstr(arg0);
self->who = strjoin(copyinstr(arg3),strjoin("@",copyinstr(arg4)));
self->db = copyinstr(arg2);
self->connid = arg1;
self->querystart = timestamp;
self->netwrite = 0;
self->netwritecum = 0;
self->netwritebase = 0;
self->netread = 0;
self->netreadcum = 0;
self->netreadbase = 0;
}
mysql*:::net-write-start
{
self->netwrite += arg0;
self->netwritebase = timestamp;
}
mysql*:::net-write-done
{
self->netwritecum += (timestamp - self->netwritebase);
self->netwritebase = 0;
}
mysql*:::net-read-start
{
self->netreadbase = timestamp;
}
mysql*:::net-read-done
{
self->netread += arg1;
self->netreadcum += (timestamp - self->netreadbase);
self->netreadbase = 0;
}
mysql*:::query-done
{
this->elapsed = (timestamp - self->querystart) /1000000;
printf("%2d %-30s %-10s %9d %18d %s\n",
arg0, self->who, self->db,
self->connid, this->elapsed, self->query);
printf("Net read: %d bytes (%d ms) write: %d bytes (%d ms)\n",
self->netread, (self->netreadcum/1000000),
self->netwrite, (self->netwritecum/1000000));
}When executing the above script on a machine with a remote client, you can see that approximately a third of the time spent executing the query is related to writing the query results back to the client.
St Who DB ConnID Dur microsec Query 0 root@::ffff:192.168.0.108 test 31 3495 select * from t1 limit 1000000 Net read: 0 bytes (0 ms) write: 10000075 bytes (1220 ms)
The keycache probes are triggered when using the index key cache used with the MyISAM storage engine. Probes exist to monitor when data is read into the keycache, cached key data is written from the cache into a cached file, or when accessing the keycache.
Keycache usage indicates when data is read or written from the index files into the cache, and can be used to monitor how efficient the memory allocated to the keycache is being used. A high number of keycache reads across a range of queries may indicate that the keycache is too small for size of data being accessed.
keycache-read-start(filepath, bytes, mem_used, mem_free) keycache-read-block(bytes) keycache-read-hit() keycache-read-miss() keycache-read-done(mem_used, mem_free) keycache-write-start(filepath, bytes, mem_used, mem_free) keycache-write-block(bytes) keycache-write-done(mem_used, mem_free)
When reading data from the index files into the keycache, the
process first initializes the read operation (indicated by
keycache-read-start), then loads blocks of
data (keycache-read-block), and then the read
block is either matches the data being identified
(keycache-read-hit) or more data needs to be
read (keycache-read-miss). Once the read
operation has completed, reading stops with the
keycache-read-done.
Data will be read from the index file into the keycache only when the specified key is not already within the keycache.
keycache-read-start: Triggered when the keycache read operation is started. Data is read from the specifiedfilepath, reading the specified number ofbytes. Themem_usedandmem_availindicate memory currently used by the keycache and the amount of memory available within the keycache.keycache-read-block: Triggered when the keycache reads a block of data, of the specified number ofbytes, from the index file into the keycache.keycache-read-hit: Triggered when the block of data read from the index file matches the key data requested.keycache-read-miss: Triggered when the block of data read from the index file does not match the key data needed.keycache-read-done: Triggered when the keycache read operation has completed. Themem_usedandmem_availindicate memory currently used by the keycache and the amount of memory available within the keycache.
Keycache writes occur when the index information is updated
during an INSERT, UPDATE,
or DELETE operation, and the cached key
information is flushed back to the index file.
keycache-write-start: Triggered when the keycache write operation is started. Data is written to the specifiedfilepath, reading the specified number ofbytes. Themem_usedandmem_availindicate memory currently used by the keycache and the amount of memory available within the keycache.keycache-write-block: Triggered when the keycache writes a block of data, of the specified number ofbytes, to the index file from the keycache.keycache-write-done: Triggered when the keycache write operation has completed. Themem_usedandmem_availindicate memory currently used by the keycache and the amount of memory available within the keycache.