MySQL使用profile查询性能的操作教程

MYSQL的profiling功能要在Mysql版本5.0.37以上才能使用。

查看profile是否开启

?

1	mysql> show variables like '%profil%';

?

1234567

+------------------------+-------+| Variable_name     | Value |+------------------------+-------+| profiling       | OFF  |     --开启SQL语句剖析功能                         | profiling_history_size | 15  |     --设置保留profiling的数目，缺省为15，范围为0至100，为0时将禁用profiling+------------------------+-------+2 rows in set (0.00 sec)

基于会话级别开启

?

1	mysql> set profiling = 1;     --关闭则用set profiling = off

?

1	Query OK, 0 rows affected (0.00 sec)

 

?

123456789

mysql> select distinct d.account,a.server_id from tab_appserver_user a  -> inner join tab_department_parent b on a.key_id = b.parent_id  -> inner join tab_department_member c on b.department_id = c.department_id and c.state=1  -> and c.isdefault=1 inner join tab_user_info d on c.user_id = d.user_id and d.state=1  -> where a.type=1   -> union           -> select distinct b.account,a.server_id from tab_appserver_user a  -> inner join tab_user_info b on a.key_id = b.user_id and b.state=1  -> where a.type=0;

查看是否设置生效：

?

1	select @@profiling;

默认是0，设置成功是1

运行SQL语句：

?

1	mysql> select * FROM hx_line WHERE id = '1455023';

查看profiles

?

1	mysql> show profiles;

?

12345

+----------+------------+---------------------------------------------+| Query_ID | Duration  | Query                    |+----------+------------+---------------------------------------------+|    1 | 0.00036150 | select * FROM hx_line WHERE id = '1455023' |+----------+------------+---------------------------------------------+

查看具体某条的profile

?

1	mysql> show profile FOR QUERY 1;

?

+--------------------------------+----------+| Status             | Duration |+--------------------------------+----------+| starting            | 0.000013 || Waiting for query cache lock  | 0.000014 || checking query cache for query | 0.000038 || checking permissions      | 0.000006 || Opening tables         | 0.000013 || System lock          | 0.000009 || Waiting for query cache lock  | 0.000024 || init              | 0.000060 || optimizing           | 0.000014 || statistics           | 0.000046 || preparing           | 0.000017 || executing           | 0.000004 || Sending data          | 0.000081 || end              | 0.000005 || query end           | 0.000004 || closing tables         | 0.000008 || freeing items         | 0.000009 || Waiting for query cache lock  | 0.000003 || freeing items         | 0.000013 || Waiting for query cache lock  | 0.000003 || freeing items         | 0.000003 || storing result in query cache | 0.000005 || logging slow query       | 0.000003 || cleaning up          | 0.000004 |+--------------------------------+----------+24 rows

我们看到了一个简单的查询，MYSQL内部做了24次操作。
另外，看到了一堆query cache的操作，试着把query_cache_size=0，把query_cache关闭，再次测试：

?

1	mysql> show profile FOR QUERY 1;

?

+----------------------+----------+| Status        | Duration |+----------------------+----------+| starting       | 0.000040 || checking permissions | 0.000007 || Opening tables    | 0.000015 || System lock     | 0.000010 || init         | 0.000061 || optimizing      | 0.000013 || statistics      | 0.000059 || preparing      | 0.000018 || executing      | 0.000004 || Sending data     | 0.000092 || end         | 0.000006 || query end      | 0.000004 || closing tables    | 0.000008 || freeing items    | 0.000020 || logging slow query  | 0.000003 || cleaning up     | 0.000004 |+----------------------+----------+16 rows in set (0.00 sec)

当开启了query_cache的情况下，需要多操作6次，在这个示例里面多化了0.000087s。

查询这条语句对CPU的使用情况：

?

1	mysql> show profile cpu FOR QUERY 1;

?

+----------------------+----------+----------+------------+| Status        | Duration | CPU_user | CPU_system |+----------------------+----------+----------+------------+| starting       | 0.000037 | 0.000000 |  0.000000 || checking permissions | 0.000009 | 0.000000 |  0.000000 || Opening tables    | 0.000014 | 0.000000 |  0.000000 || System lock     | 0.000009 | 0.000000 |  0.000000 || init         | 0.000059 | 0.000000 |  0.000000 || optimizing      | 0.000009 | 0.000000 |  0.000000 || statistics      | 0.000044 | 0.000000 |  0.000000 || preparing      | 0.000015 | 0.000000 |  0.000000 || executing      | 0.000004 | 0.000000 |  0.000000 || Sending data     | 0.000081 | 0.000000 |  0.000000 || end         | 0.000006 | 0.000000 |  0.000000 || query end      | 0.000004 | 0.000000 |  0.000000 || closing tables    | 0.000008 | 0.000000 |  0.000000 || freeing items    | 0.000021 | 0.000000 |  0.000000 || logging slow query  | 0.000004 | 0.000000 |  0.000000 || cleaning up     | 0.000004 | 0.000000 |  0.000000 |+----------------------+----------+----------+------------+

查看io及cpu的消耗

?

1	mysql> show profile block io,cpu for query 1;

?

+--------------------------------+----------+----------+------------+--------------+---------------+| Status             | Duration | CPU_user | CPU_system | Block_ops_in | Block_ops_out |+--------------------------------+----------+----------+------------+--------------+---------------+| starting            | 0.000018 |   NULL |    NULL |     NULL |     NULL || checking query cache for query | 0.000099 |   NULL |    NULL |     NULL |     NULL || Opening tables         | 0.000963 |   NULL |    NULL |     NULL |     NULL || System lock          | 0.000015 |   NULL |    NULL |     NULL |     NULL || Table lock           | 0.000169 |   NULL |    NULL |     NULL |     NULL || optimizing           | 0.000020 |   NULL |    NULL |     NULL |     NULL || statistics           | 0.000027 |   NULL |    NULL |     NULL |     NULL || preparing           | 0.000018 |   NULL |    NULL |     NULL |     NULL || Creating tmp table       | 0.000055 |   NULL |    NULL |     NULL |     NULL || executing           | 0.000003 |   NULL |    NULL |     NULL |     NULL || Copying to tmp table      | 0.704845 |   NULL |    NULL |     NULL |     NULL || Sending data          | 0.130039 |   NULL |    NULL |     NULL |     NULL || optimizing           | 0.000029 |   NULL |    NULL |     NULL |     NULL || statistics           | 0.000029 |   NULL |    NULL |     NULL |     NULL || preparing           | 0.000020 |   NULL |    NULL |     NULL |     NULL || Creating tmp table       | 0.000142 |   NULL |    NULL |     NULL |     NULL || executing           | 0.000003 |   NULL |    NULL |     NULL |     NULL || Copying to tmp table      | 0.000086 |   NULL |    NULL |     NULL |     NULL || Sending data          | 0.000067 |   NULL |    NULL |     NULL |     NULL || optimizing           | 0.000004 |   NULL |    NULL |     NULL |     NULL || statistics           | 0.000005 |   NULL |    NULL |     NULL |     NULL || preparing           | 0.000005 |   NULL |    NULL |     NULL |     NULL || executing           | 0.000002 |   NULL |    NULL |     NULL |     NULL || Sending data          | 0.023963 |   NULL |    NULL |     NULL |     NULL || removing tmp table       | 0.003420 |   NULL |    NULL |     NULL |     NULL || Sending data          | 0.000005 |   NULL |    NULL |     NULL |     NULL || removing tmp table       | 0.003308 |   NULL |    NULL |     NULL |     NULL || Sending data          | 0.000006 |   NULL |    NULL |     NULL |     NULL || removing tmp table       | 0.000007 |   NULL |    NULL |     NULL |     NULL || Sending data          | 0.000009 |   NULL |    NULL |     NULL |     NULL || query end           | 0.000003 |   NULL |    NULL |     NULL |     NULL || freeing items         | 0.000144 |   NULL |    NULL |     NULL |     NULL || storing result in query cache | 0.000011 |   NULL |    NULL |     NULL |     NULL || logging slow query       | 0.000003 |   NULL |    NULL |     NULL |     NULL || cleaning up          | 0.000006 |   NULL |    NULL |     NULL |     NULL |+--------------------------------+----------+----------+------------+--------------+---------------+35 rows in set (0.00 sec)

使用查询语句对消耗进行排序

?

123456

mysql> SELECT STATE, SUM(DURATION) AS Total_R,ROUND( 100 * SUM(DURATION) / (SECT SUM(DURATION)  -> FROM INFORMATION_SCHEMA.PROFILING WHERE QUERY_ID = 1), 2) AS Pct_R, COT(*) AS Calls,SUM(DURATION) / COUNT(*) AS "R/Call"  -> FROM INFORMATION_SCHEMA.PROFILING WHERE QUERY_ID = 1 GROUP BY STATE OER BY Total_R DESC;

?

+--------------------------------+----------+-------+-------+--------------+| STATE             | Total_R | Pct_R | Calls | R/Call    |+--------------------------------+----------+-------+-------+--------------+| Copying to tmp table      | 0.704931 | 81.26 |   2 | 0.3524655000 || Sending data          | 0.154089 | 17.76 |   6 | 0.0256815000 || removing tmp table       | 0.006735 | 0.78 |   3 | 0.0022450000 || Opening tables         | 0.000963 | 0.11 |   1 | 0.0009630000 || Creating tmp table       | 0.000197 | 0.02 |   2 | 0.0000985000 || Table lock           | 0.000169 | 0.02 |   1 | 0.0001690000 || freeing items         | 0.000144 | 0.02 |   1 | 0.0001440000 || checking query cache for query | 0.000099 | 0.01 |   1 | 0.0000990000 || statistics           | 0.000061 | 0.01 |   3 | 0.0000203333 || optimizing           | 0.000053 | 0.01 |   3 | 0.0000176667 || preparing           | 0.000043 | 0.00 |   3 | 0.0000143333 || starting            | 0.000018 | 0.00 |   1 | 0.0000180000 || System lock          | 0.000015 | 0.00 |   1 | 0.0000150000 || storing result in query cache | 0.000011 | 0.00 |   1 | 0.0000110000 || executing           | 0.000008 | 0.00 |   3 | 0.0000026667 || cleaning up          | 0.000006 | 0.00 |   1 | 0.0000060000 || logging slow query       | 0.000003 | 0.00 |   1 | 0.0000030000 || query end           | 0.000003 | 0.00 |   1 | 0.0000030000 |+--------------------------------+----------+-------+-------+--------------+18 rows in set (0.01 sec)

show profile额外一些命令：
* ALL - displays all information
* BLOCK IO - displays counts for block input and output Operations
* CONTEXT SWITCHES - displays counts for voluntary and involuntary context switches
* ipC - displays counts for messages sent and received
* MEMORY - is not currently implemented
* PAGE FAULTS - displays counts for major and minor page faults
* SOURCE - displays the names of functions from the source code, together with the name and line number of the file in which the function occurs
* SWAPS - displays swap counts

最后说明：profile是一个非常量化的子标，可以根据这些量化指标来比较各项资源的消耗，有利于我们对该语句的整体把控！