pt-variable-advisor - Analyze MySQL variables and advise on possible problems.
pt-variable-advisor [OPTIONS] [DSN]
pt-variable-advisor analyzes variables and advises on possible problems.
Get SHOW VARIABLES from localhost:
Get SHOW VARIABLES output saved in vars.txt:
pt-variable-advisor --source-of-variables vars.txt
Percona Toolkit is mature, proven in the real world, and well tested, but all database tools can pose a risk to the system and the database server. Before using this tool, please:
- Read the tool’s documentation
- Review the tool’s known “BUGS”
- Test the tool on a non-production server
- Backup your production server and verify the backups
SHOW VARIABLES for bad values and settings
according to the “RULES” described below. It reports on variables that
match the rules, so you can find bad settings in your MySQL server.
At the time of this release, pt-variable-advisor only examples
SHOW VARIABLES, but other input sources are planned like
SHOW SLAVE STATUS.
These are the rules that pt-variable-advisor will apply to SHOW VARIABLES. Each rule has three parts: an ID, a severity, and a description.
The rule’s ID is a short, unique name for the rule. It usually relates to the variable that the rule examines. If a variable is examined by several rules, then the rules’ IDs are numbered like “-1”, “-2”, “-N”.
The rule’s severity is an indication of how important it is that this rule matched a query. We use NOTE, WARN, and CRIT to denote these levels.
The rule’s description is a textual, human-readable explanation of what it means when a variable matches this rule. Depending on the verbosity of the report you generate, you will see more of the text in the description. By default, you’ll see only the first sentence, which is sort of a terse synopsis of the rule’s meaning. At a higher verbosity, you’ll see subsequent sentences.
Are you trying to write to more than one server in a dual-master or ring replication configuration? This is potentially very dangerous and in most cases is a serious mistake. Most people’s reasons for doing this are actually not valid at all.
Holes (spaces left by deletes) in MyISAM tables might never be reused.
A large value of this setting can create a denial of service vulnerability.
Servers built with debugging capability should not be used in production because of the large performance impact.
MyISAM index blocks are never flushed until necessary. If there is a server crash, data corruption on MyISAM tables can be much worse than usual.
This option might decrease performance greatly.
This option might decrease performance greatly.
The BDB engine is deprecated. If you aren’t using it, you should disable it with the skip_bdb option.
The init_connect option is enabled on this server.
The init_file option is enabled on this server.
The init_slave option is enabled on this server.
This variable generally doesn’t need to be larger than 20MB.
The InnoDB buffer pool size is unconfigured. In a production environment it should always be configured explicitly, and the default 10MB size is not good.
InnoDB checksums are disabled. Your data is not protected from hardware corruption or other errors!
InnoDB doublewrite is disabled. Unless you use a filesystem that protects against partial page writes, your data is not safe!
InnoDB’s shutdown behavior is not the default. This can lead to poor performance, or the need to perform crash recovery upon startup.
InnoDB is not configured in strictly ACID mode. If there is a crash, some transactions can be lost.
Setting innodb_flush_log_at_trx_commit to 0 has no performance benefits over setting it to 2, and more types of data loss are possible. If you are trying to change it from 1 for performance reasons, you should set it to 2 instead of 0.
InnoDB is in forced recovery mode! This should be used only temporarily when recovering from data corruption or other bugs, not for normal usage.
This option has an unusually long value, which can cause system overload if locks are not being released.
The InnoDB log buffer size generally should not be set larger than 16MB. If you are doing large BLOB operations, InnoDB is not really a good choice of engines anyway.
The InnoDB log file size is set to its default value, which is not usable on production systems.
The innodb_max_dirty_pages_pct is lower than the default. This can cause overly aggressive flushing and add load to the I/O system.
This setting is likely to cause very bad performance every flush_time seconds.
The key buffer size is set to its default value, which is not good for most production systems. In a production environment, key_buffer_size should be larger than the default 8MB size.
Large pages are enabled.
The server is locked in memory with –memlock.
Log_warnings is disabled, so unusual events such as statements unsafe for replication and aborted connections will not be logged to the error log.
Log_warnings must be set greater than 1 to log unusual events such as aborted connections.
The server is running with non-default lock priority for updates. This could cause update queries to wait unexpectedly for read queries.
The max_binlog_size is smaller than the default of 1GB.
max_connect_errors should probably be set as large as your platform allows.
If the server ever really has more than a thousand threads running, then the system is likely to spend more time scheduling threads than really doing useful work. This variable’s value should be considered in light of your workload.
myisam_repair_threads > 1 enables multi-threaded repair, which is relatively untested and is still listed as beta-quality code in the official documentation.
Old-style passwords are insecure. They are sent in plain text across the wire.
The optimizer will use an exhaustive search when planning complex queries, which can cause the planning process to take a long time.
The server is listening on a non-default port.
The query cache does not scale to large sizes and can cause unstable performance when larger than 128MB, especially on multi-core machines.
The query cache can cause severe performance problems when it is larger than 256MB, especially on multi-core machines.
The read_buffer_size variable should generally be left at its default unless an expert determines it is necessary to change it.
The read_buffer_size variable should not be larger than 8MB. It should generally be left at its default unless an expert determines it is necessary to change it. Making it larger than 2MB can hurt performance significantly, and can make the server crash, swap to death, or just become extremely unstable.
The read_rnd_buffer_size variable should generally be left at its default unless an expert determines it is necessary to change it.
The read_rnd_buffer_size variable should not be larger than 4M. It should generally be left at its default unless an expert determines it is necessary to change it.
Setting relay_log_space_limit can cause replicas to stop fetching binary logs from their master immediately. This could increase the risk that your data will be lost if the master crashes. If the replicas have encountered a limit on relay log space, then it is possible that the latest transactions exist only on the master and no replica has retrieved them.
This variable is set too high. This is too long to wait before noticing that the connection to the master has failed and retrying. This should probably be set to 60 seconds or less. It is also a good idea to use pt-heartbeat to ensure that the connection does not appear to time out when the master is simply idle.
You should not set this option. If replication is having errors, you need to find and resolve the cause of that; it is likely that your slave’s data is different from the master. You can find out with pt-table-checksum.
The sort_buffer_size variable should generally be left at its default unless an expert determines it is necessary to change it.
The sort_buffer_size variable should generally be left at its default unless an expert determines it is necessary to change it. Making it larger than a few MB can hurt performance significantly, and can make the server crash, swap to death, or just become extremely unstable.
This server is configured not to log Note level warnings to the error log.
It is best to set sync_frm so that .frm files are flushed safely to disk in case of a server crash.
This server’s transaction isolation level is non-default.
Most applications should use the default REPEATABLE-READ transaction isolation level, or in a few cases READ-COMMITTED.
Binary logs are enabled, but automatic purging is not enabled. If you do not purge binary logs, your disk will fill up. If you delete binary logs externally to MySQL, you will cause unwanted behaviors. Always ask MySQL to purge obsolete logs, never delete them externally.
This option is useless except on Windows.
Auto-extending InnoDB files can consume a lot of disk space that is very difficult to reclaim later. Some people prefer to set innodb_file_per_table and allocate a fixed-size file for ibdata1.
Most production database servers that use InnoDB should set innodb_flush_method to O_DIRECT to avoid double-buffering, unless the I/O system is very low performance.
This option makes point-in-time recovery from binary logs, and replication, untrustworthy if statement-based logging is used.
MySQL’s internal XA transaction support between InnoDB and the binary log is disabled. The binary log might not match InnoDB’s state after crash recovery, and replication might drift out of sync due to out-of-order statements in the binary log.
Binary logging is disabled, so point-in-time recovery and replication are not possible.
Directing log output to tables has a high performance impact.
A custom max_relay_log_size is defined.
myisam_recover_options should be set to some value such as BACKUP,FORCE to ensure that table corruption is noticed.
The server is using a non-standard storage engine as default.
Binary logging is enabled, but sync_binlog isn’t configured so that every transaction is flushed to the binary log for durability.
The effective minimum size of in-memory implicit temporary tables used internally during query execution is min(tmp_table_size, max_heap_table_size), so max_heap_table_size should be at least as large as tmp_table_size.
old mysql version
These are the recommended minimum version for each major release: 3.23, 4.1.20, 5.0.37, 5.1.30.
end-of-life mysql version
Every release older than 5.1 is now officially end-of-life.
This tool accepts additional command-line arguments. Refer to the “SYNOPSIS” and usage information for details.
Prompt for a password when connecting to MySQL.
short form: -A; type: string
Default character set. If the value is utf8, sets Perl’s binmode on STDOUT to utf8, passes the mysql_enable_utf8 option to DBD::mysql, and runs SET NAMES UTF8 after connecting to MySQL. Any other value sets binmode on STDOUT without the utf8 layer, and runs SET NAMES after connecting to MySQL.
Read this comma-separated list of config files; if specified, this must be the first option on the command line.
Fork to the background and detach from the shell. POSIX operating systems only.
short form: -D; type: string
Connect to this database.
short form: -F; type: string
Only read mysql options from the given file. You must give an absolute pathname.
Show help and exit.
short form: -h; type: string
Connect to host.
Ignore these rule IDs.
Specify a comma-separated list of rule IDs (e.g. LIT.001,RES.002,etc.) to ignore.
short form: -p; type: string
Password to use when connecting. If password contains commas they must be escaped with a backslash: “exam,ple”
Create the given PID file. The tool won’t start if the PID file already exists and the PID it contains is different than the current PID. However, if the PID file exists and the PID it contains is no longer running, the tool will overwrite the PID file with the current PID. The PID file is removed automatically when the tool exits.
short form: -P; type: int
Port number to use for connection.
Set the MySQL variables in this comma-separated list of
By default, the tool sets:
Variables specified on the command line override these defaults. For example, specifying
--set-vars wait_timeout=500overrides the defaultvalue of
The tool prints a warning and continues if a variable cannot be set.
short form: -S; type: string
Socket file to use for connection.
type: string; default: mysql
SHOW VARIABLESfrom this source. Possible values are “mysql”, “none” or a file name. If “mysql” is specified then you must also specify a DSN on the command line.
short form: -u; type: string
User for login if not current user.
short form: -v; cumulative: yes; default: 1
Increase verbosity of output. At the default level of verbosity, the program prints only the first sentence of each rule’s description. At higher levels, the program prints more of the description.
Show version and exit.
Check for the latest version of Percona Toolkit, MySQL, and other programs.
This is a standard “check for updates automatically” feature, with two additional features. First, the tool checks the version of other programs on the local system in addition to its own version. For example, it checks the version of every MySQL server it connects to, Perl, and the Perl module DBD::mysql. Second, it checks for and warns about versions with known problems. For example, MySQL 5.5.25 had a critical bug and was re-released as 5.5.25a.
Any updates or known problems are printed to STDOUT before the tool’s normal output. This feature should never interfere with the normal operation of the tool.
For more information, visit https://www.percona.com/version-check.
These DSN options are used to create a DSN. Each option is given like
option=value. The options are case-sensitive, so P and p are not the
same option. There cannot be whitespace before or after the
if the value contains whitespace it must be quoted. DSN options are
comma-separated. See the percona-toolkit manpage for full details.
dsn: charset; copy: yes
Default character set.
dsn: database; copy: yes
dsn: mysql_read_default_file; copy: yes
Only read default options from the given file
dsn: host; copy: yes
Connect to host.
dsn: password; copy: yes
dsn: port; copy: yes
Port number to use for connection.
dsn: mysql_socket; copy: yes
Socket file to use for connection.
dsn: user; copy: yes
User for login if not current user.
The environment variable
PTDEBUG enables verbose debugging output to STDERR.
To enable debugging and capture all output to a file, run the tool like:
PTDEBUG=1 pt-variable-advisor ... > FILE 2>&1
Be careful: debugging output is voluminous and can generate several megabytes of output.
You need Perl, DBI, DBD::mysql, and some core packages that ought to be installed in any reasonably new version of Perl.
For a list of known bugs, see http://www.percona.com/bugs/pt-variable-advisor.
Please report bugs at https://bugs.launchpad.net/percona-toolkit. Include the following information in your bug report:
- Complete command-line used to run the tool
- MySQL version of all servers involved
- Output from the tool including STDERR
- Input files (log/dump/config files, etc.)
If possible, include debugging output by running the tool with
Visit http://www.percona.com/software/percona-toolkit/ to download the latest release of Percona Toolkit. Or, get the latest release from the command line:
wget percona.com/get/percona-toolkit.tar.gz wget percona.com/get/percona-toolkit.rpm wget percona.com/get/percona-toolkit.deb
You can also get individual tools from the latest release:
TOOL with the name of any tool.
ABOUT PERCONA TOOLKIT¶
This tool is part of Percona Toolkit, a collection of advanced command-line tools for MySQL developed by Percona. Percona Toolkit was forked from two projects in June, 2011: Maatkit and Aspersa. Those projects were created by Baron Schwartz and primarily developed by him and Daniel Nichter. Visit http://www.percona.com/software/ to learn about other free, open-source software from Percona.
COPYRIGHT, LICENSE, AND WARRANTY¶
This program is copyright 2010-2016 Percona LLC and/or its affiliates.
THIS PROGRAM IS PROVIDED “AS IS” AND WITHOUT ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.
This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 2; OR the Perl Artistic License. On UNIX and similar systems, you can issue `man perlgpl’ or `man perlartistic’ to read these licenses.
You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA.