分析诊断工具之三:使用show profiles分析SQL性能
分析SQL执行带来的开销是优化SQL的重要手段。在MySQL数据库中,可以通过配置profiling参数来启用SQL剖析。该参数可以在全局和session级别来设置。对于全局级别则作用于整个MySQL实例,而session级别紧影响当前session。该参数开启后,后续执行的SQL语句都将记录其资源开销,诸如IO,上下文切换,CPU,Memory等等。根据这些开销进一步分析当前SQL瓶颈从而进行优化与调整。本文描述了如何使用MySQL profile,不涉及具体的样例分析。
如何查看执行SQL的耗时的步骤:开启profile、发送sql、查看profile的资源开销结果、关闭profile。
使用show profiles分析sql性能。
Show profiles是5.0.37之后添加的,要想使用此功能,要确保版本在5.0.37之后。
查看数据库版本
mysql> select version();
profile默认是不打开的
mysql> show profiles;
Empty set (0.02 sec)
验证修改后的结果
mysql> show variables like "%pro%";
可以看到profiling 默认是OFF的。
开启profile,然后测试
开启profile
mysql> set profiling=1;
获取profile的帮助
help profile;
root@localhost[sakila]> help profile; Name: 'SHOW PROFILE' Description: Syntax: SHOW PROFILE [type [, type] ... ] [FOR QUERY n] [LIMIT row_count [OFFSET offset]] type: ALL --显示所有的开销信息 | BLOCK IO --显示块IO相关开销 | CONTEXT SWITCHES --上下文切换相关开销 | CPU --显示CPU相关开销信息 | IPC --显示发送和接收相关开销信息 | MEMORY --显示内存相关开销信息 | PAGE FAULTS --显示页面错误相关开销信息 | SOURCE --显示和Source_function,Source_file,Source_line相关的开销信息 | SWAPS --显示交换次数相关开销的信息 The SHOW PROFILE and SHOW PROFILES statements display profiling information that indicates resource usage for statements executed during the course of the current session. *Note*: These statements are deprecated as of MySQL 5.6.7 and will be removed in a future MySQL release. Use the Performance Schema instead; see http://dev.mysql.com/doc/refman/5.6/en/performance-schema.html. --上面描述从5.6.7开始该命令将会被移除,用Performance Schema instead代替 --在Oracle数据库中,是通过autotrace来剖析单条SQL并获取真实的执行计划以及其开销信息
执行业务SQL,并用profile分析示例:
--发布SQL查询 root@localhost[sakila]> select count(*) from customer; +----------+ | count(*) | +----------+ | 599 | +----------+ --查看当前session所有已产生的profile root@localhost[sakila]> show profiles; +----------+------------+--------------------------------+ | Query_ID | Duration | Query | +----------+------------+--------------------------------+ | 1 | 0.00253600 | show variables like '%profil%' | | 2 | 0.00138150 | select count(*) from customer | +----------+------------+--------------------------------+ 2 rows in set, 1 warning (0.01 sec) --我们看到有2个warning,之前一个,现在一个 root@localhost[sakila]> show warnings; --下面的结果表明SHOW PROFILES将来会被Performance Schema替换掉 +---------+------+--------------------------------------------------------------------------------------------------------------+ | Level | Code | Message | +---------+------+--------------------------------------------------------------------------------------------------------------+ | Warning | 1287 | 'SHOW PROFILES' is deprecated and will be removed in a future release. Please use Performance Schema instead | +---------+------+--------------------------------------------------------------------------------------------------------------+
获取SQL语句的开销信息
--可以直接使用show profile来查看上一条SQL语句的开销信息
--注,show profile之类的语句不会被profiling,即自身不会产生Profiling
--我们下面的这个show profile查看的是show warnings产生的相应开销
root@localhost[sakila]> show profile;
+----------------+----------+
| Status | Duration |
+----------------+----------+
| starting | 0.000141 |
| query end | 0.000058 |
| closing tables | 0.000014 |
| freeing items | 0.001802 |
| cleaning up | 0.000272 |
+----------------+----------+
--如下面的查询show warnings被添加到profiles
root@localhost[sakila]> show profiles;
+----------+------------+--------------------------------+
| Query_ID | Duration | Query |
+----------+------------+--------------------------------+
| 1 | 0.00253600 | show variables like '%profil%' |
| 2 | 0.00138150 | select count(*) from customer |
| 3 | 0.00228600 | show warnings |
+----------+------------+--------------------------------+
--获取指定查询的开销(第二条查询的开销明细)
root@localhost[sakila]> show profile for query 2;
+----------------------+----------+
| Status | Duration |
+----------------------+----------+
| starting | 0.000148 |
| checking permissions | 0.000014 |
| Opening tables | 0.000047 |
| init | 0.000023 |
| System lock | 0.000035 |
| optimizing | 0.000012 |
| statistics | 0.000019 |
| preparing | 0.000014 |
| executing | 0.000006 |
| Sending data | 0.000990 |
| end | 0.000010 |
| query end | 0.000011 |
| closing tables | 0.000010 |
| freeing items | 0.000016 |
| cleaning up | 0.000029 |
+----------------------+----------+
--查看特定部分的开销,如下为CPU部分的开销
root@localhost[sakila]> show profile cpu for query 2 ;
+----------------------+----------+----------+------------+
| Status | Duration | CPU_user | CPU_system |
+----------------------+----------+----------+------------+
| starting | 0.000148 | 0.000000 | 0.000000 |
| checking permissions | 0.000014 | 0.000000 | 0.000000 |
| Opening tables | 0.000047 | 0.000000 | 0.000000 |
| init | 0.000023 | 0.000000 | 0.000000 |
| System lock | 0.000035 | 0.000000 | 0.000000 |
| optimizing | 0.000012 | 0.000000 | 0.000000 |
| statistics | 0.000019 | 0.000000 | 0.000000 |
| preparing | 0.000014 | 0.000000 | 0.000000 |
| executing | 0.000006 | 0.000000 | 0.000000 |
| Sending data | 0.000990 | 0.001000 | 0.000000 |
| end | 0.000010 | 0.000000 | 0.000000 |
| query end | 0.000011 | 0.000000 | 0.000000 |
| closing tables | 0.000010 | 0.000000 | 0.000000 |
| freeing items | 0.000016 | 0.000000 | 0.000000 |
| cleaning up | 0.000029 | 0.000000 | 0.000000 |
+----------------------+----------+----------+------------+
--如下为MEMORY部分的开销
root@localhost[sakila]> show profile memory for query 2 ;
+----------------------+----------+
| Status | Duration |
+----------------------+----------+
| starting | 0.000148 |
| checking permissions | 0.000014 |
| Opening tables | 0.000047 |
| init | 0.000023 |
| System lock | 0.000035 |
| optimizing | 0.000012 |
| statistics | 0.000019 |
| preparing | 0.000014 |
| executing | 0.000006 |
| Sending data | 0.000990 |
| end | 0.000010 |
| query end | 0.000011 |
| closing tables | 0.000010 |
| freeing items | 0.000016 |
| cleaning up | 0.000029 |
+----------------------+----------+
--同时查看不同资源开销
root@localhost[sakila]> show profile block io,cpu for query 2;
+----------------------+----------+----------+------------+--------------+---------------+
| Status | Duration | CPU_user | CPU_system | Block_ops_in | Block_ops_out |
+----------------------+----------+----------+------------+--------------+---------------+
| starting | 0.000148 | 0.000000 | 0.000000 | 0 | 0 |
| checking permissions | 0.000014 | 0.000000 | 0.000000 | 0 | 0 |
| Opening tables | 0.000047 | 0.000000 | 0.000000 | 0 | 0 |
| init | 0.000023 | 0.000000 | 0.000000 | 0 | 0 |
| System lock | 0.000035 | 0.000000 | 0.000000 | 0 | 0 |
| optimizing | 0.000012 | 0.000000 | 0.000000 | 0 | 0 |
| statistics | 0.000019 | 0.000000 | 0.000000 | 0 | 0 |
| preparing | 0.000014 | 0.000000 | 0.000000 | 0 | 0 |
| executing | 0.000006 | 0.000000 | 0.000000 | 0 | 0 |
| Sending data | 0.000990 | 0.001000 | 0.000000 | 0 | 0 |
| end | 0.000010 | 0.000000 | 0.000000 | 0 | 0 |
| query end | 0.000011 | 0.000000 | 0.000000 | 0 | 0 |
| closing tables | 0.000010 | 0.000000 | 0.000000 | 0 | 0 |
| freeing items | 0.000016 | 0.000000 | 0.000000 | 0 | 0 |
| cleaning up | 0.000029 | 0.000000 | 0.000000 | 0 | 0 |
+----------------------+----------+----------+------------+--------------+---------------+
--下面的SQL语句用于查询query_id为2的SQL开销,且按最大耗用时间倒序排列
root@localhost[sakila]> set @query_id=2;
root@localhost[sakila]> SELECT STATE, SUM(DURATION) AS Total_R,
-> ROUND(
-> 100 * SUM(DURATION) /
-> (SELECT SUM(DURATION)
-> FROM INFORMATION_SCHEMA.PROFILING
-> WHERE QUERY_ID = @query_id
-> ), 2) AS Pct_R,
-> COUNT(*) AS Calls,
-> SUM(DURATION) / COUNT(*) AS "R/Call"
-> FROM INFORMATION_SCHEMA.PROFILING
-> WHERE QUERY_ID = @query_id
-> GROUP BY STATE
-> ORDER BY Total_R DESC;
+----------------------+----------+-------+-------+--------------+
| STATE | Total_R | Pct_R | Calls | R/Call |
+----------------------+----------+-------+-------+--------------+
| Sending data | 0.000990 | 71.53 | 1 | 0.0009900000 |--最大耗用时间部分为发送数据
| starting | 0.000148 | 10.69 | 1 | 0.0001480000 |
| Opening tables | 0.000047 | 3.40 | 1 | 0.0000470000 |
| System lock | 0.000035 | 2.53 | 1 | 0.0000350000 |
| cleaning up | 0.000029 | 2.10 | 1 | 0.0000290000 |
| init | 0.000023 | 1.66 | 1 | 0.0000230000 |
| statistics | 0.000019 | 1.37 | 1 | 0.0000190000 |
| freeing items | 0.000016 | 1.16 | 1 | 0.0000160000 |
| preparing | 0.000014 | 1.01 | 1 | 0.0000140000 |
| checking permissions | 0.000014 | 1.01 | 1 | 0.0000140000 |
| optimizing | 0.000012 | 0.87 | 1 | 0.0000120000 |
| query end | 0.000011 | 0.79 | 1 | 0.0000110000 |
| end | 0.000010 | 0.72 | 1 | 0.0000100000 |
| closing tables | 0.000010 | 0.72 | 1 | 0.0000100000 |
| executing | 0.000006 | 0.43 | 1 | 0.0000060000 |
+----------------------+----------+-------+-------+--------------+
--开启profiling后,我们可以通过show profile等方式查看,其实质是这些开销信息被记录到information_schema.profiling表
--如下面的查询,部分信息省略
profiling
root@localhost[information_schema]> select * from profiling limit 3,3\G;
*************************** 1. row ***************************
QUERY_ID: 1
SEQ: 5
STATE: init
DURATION: 0.000020
CPU_USER: 0.000000
CPU_SYSTEM: 0.000000
CONTEXT_VOLUNTARY: 0
CONTEXT_INVOLUNTARY: 0
BLOCK_OPS_IN: 0
BLOCK_OPS_OUT: 0
MESSAGES_SENT: 0
MESSAGES_RECEIVED: 0
PAGE_FAULTS_MAJOR: 0
PAGE_FAULTS_MINOR: 0
SWAPS: 0
SOURCE_FUNCTION: mysql_prepare_select
SOURCE_FILE: sql_select.cc
SOURCE_LINE: 1050
--停止profile,可以设置profiling参数,或者在session退出之后,profiling会被自动关闭
root@localhost[sakila]> set profiling=off;
Query OK, 0 rows affected, 1 warning (0.00 sec)