MySQL Profile对于分析执行计划的开销来说，还是有一定的帮助，至少在分析一些性能问题的时候有很多的参考依据。
我在5.6, 5.7版本中进行了测试，没发现差别，还是以5.7为例进行演示吧。
mysql> status;
--------------
MYSQL Ver 14.14 Distrib 5.7.17, for macos10.12 (x86_64) using EditLine wrapper
Connection id: 5
Current database: performance_schema
Current user: root@localhost
SSL: Not in use
Current pager: stdout
Using outfile: ''
Using delimiter: ;
Server version: 5.7.17 MySQL Community Server (GPL)
Protocol version: 10
Connection: Localhost via UNIX socket
Server characterset: latin1
Db characterset: utf8
Client characterset: utf8
Conn. characterset: utf8
UNIX socket: /tmp/mysql.sock
Uptime: 8 hours 26 min 58 sec
传统的使用Profile都是使用show profile这样的命令方式，这个功能默认是关闭的，通过 have_profiling 参数查看当前 MySQL是否支持 profile。
mysql> select @@have_profiling;
+------------------+
| @@have_profiling |
+------------------+
| YES |
+------------------+
1 row in set, 1 warning (0.00 sec)
可以看到有一个警告，我们看看是什么警告。
mysql> show warnings;
+---------+------+--------------------------------------------------------------------------------------------------------------+
| Level | Code | Message |
+---------+------+--------------------------------------------------------------------------------------------------------------+
| Warning | 1287 | 'SHOW PROFILES' is deprecated and will be removed in a future release. Please use Performance Schema instead |
+---------+------+--------------------------------------------------------------------------------------------------------------+
1 row in set (0.00 sec)
原来这种方式已经过期，新的功能是在performance_schema中开放。当然在5.6， 5.7版本中还是暂时可用，我们先简单了解一下，再来看performance_schema怎么用。
profile相关的几个参数如下：
mysql> show variables like '%profil%';
+------------------------+-------+
| Variable_name | Value |
+------------------------+-------+
| have_profiling | YES |
| profiling | OFF |
| profiling_history_size | 15 |
+------------------------+-------+
3 rows in set (0.00 sec)
可以看到Profileing为OFF,当前默认值为0,代表的是一个意思。
mysql> SELECT @@profiling;
+-------------+
| @@profiling |
+-------------+
| 0 |
+-------------+
1 row in set, 1 warning (0.00 sec)
have_profiling 用于控制是否开启或者禁用Profiling
profiling_history_size是保留Profiling的数目
当然本质上，Profile的内容还是来自于information_schema数据库的 profiling 表。
mysql> select * from information_schema.profiling\G
Empty set, 1 warning (0.00 sec)
这个地方还是有一个警告，还是过期的提示。
mysql> show warnings;
+---------+------+-----------------------------------------------------------------------------------------------------------------------------+
| Level | Code | Message |
+---------+------+-----------------------------------------------------------------------------------------------------------------------------+
| Warning | 1287 | 'INFORMATION_SCHEMA.PROFILING' is deprecated and will be removed in a future release. Please use Performance Schema instead |
+---------+------+-----------------------------------------------------------------------------------------------------------------------------+
1 row in set (0.00 sec)
接下来开启profiling
mysql> set profiling=1;
Query OK, 0 rows affected, 1 warning (0.00 sec)
查看所有的profiles
mysql> show profiles;
+----------+------------+---------------+
| Query_ID | Duration | Query |
+----------+------------+---------------+
| 1 | 0.00018200 | show warnings |
+----------+------------+---------------+
1 row in set, 1 warning (0.00 sec)
我们顺便运行一条SQL
mysql> select count(*)from information_schema.columns;
+----------+
| count(*) |
+----------+
| 3077 |
+----------+
1 row in set (0.07 sec)
然后再次查看，就会看到query_ID会得到刚刚运行的语句。
mysql> show profiles;
+----------+------------+------------------------------------------------+
| Query_ID | Duration | Query |
+----------+------------+------------------------------------------------+
| 1 | 0.00018200 | show warnings |
| 2 | 0.06627200 | select count(*)from information_schema.columns |
+----------+------------+------------------------------------------------+
2 rows in set, 1 warning (0.00 sec)
可以使用如下的方式来查看profile的信息，比如涉及CPU的明细信息。
mysql> show profile cpu for query 2;
+----------------------+----------+----------+------------+
| Status | Duration | CPU_user | CPU_system |
+----------------------+----------+----------+------------+
| checking permissions | 0.000004 | 0.000000 | 0.000000 |
| checking permissions | 0.000053 | 0.000999 | 0.000000 |
| checking permissions | 0.000014 | 0.000000 | 0.000000 |
| checking permissions | 0.000006 | 0.000000 | 0.000000 |
。。。。。
| closing tables | 0.000005 | 0.000000 | 0.000000 |
| freeing items | 0.000052 | 0.000000 | 0.000000 |
| cleaning up | 0.000023 | 0.000000 | 0.000000 |
+----------------------+----------+----------+------------+
100 rows in set, 1 warning (0.00 sec)
除此之外，还有其他选项all、cpu、block、io、context、switch、page faults 等用来查看 MySQL在使用什么资源时浪费了过高的时间。
上面的内容已经过期即将在未来版本中删除，那么我们来看看新版本中推荐的performace_schema如何使用。先切换到performance_schema下，这是MySQL新增的性能优化引擎，在5.6以前是关闭的，5.6, 5.7中是默认开启的，5.7切换的时候还会有一句提示：
mysql> use performance_schema;
Reading table information for completion of table and column names
You can turn off this feature to get a quicker startup with -A
Database changed
提示我们在连接 MySQL 的时候可以使用-A 参数预读数据库。
使用profile涉及几个表，setup_actors、setup_instruments、setup_consumers。默认表setup_actors的内容如下：
mysql> SELECT * FROM setup_actors;
+------+------+------+---------+---------+
| HOST | USER | ROLE | ENABLED | HISTORY |
+------+------+------+---------+---------+
| % | % | % | YES | YES |
+------+------+------+---------+---------+
1 row in set (0.00 sec)
按照官方的建议，默认是启用，可以根据需求禁用。
mysql> UPDATE performance_schema.setup_actors
-> SET enabled = 'no',history = 'no'
-> WHERE host = '%' AND user = '%';
Query OK, 1 row affected (0.04 sec)
Rows matched: 1 Changed: 1 Warnings: 0
禁用后的内容如下：
mysql> select * from setup_actors;
+------+------+------+---------+---------+
| HOST | USER | ROLE | ENABLED | HISTORY |
+------+------+------+---------+---------+
| % | % | % | NO | NO |
+------+------+------+---------+---------+
1 row in set (0.00 sec)
然后加入指定的用户
mysql> INSERT INTO performance_schema.setup_actors
-> (HOST,USER,ROLE,ENABLED,HISTORY)
-> VALUES('localhost','root','%','YES','YES');
Query OK, 1 row affected (0.03 sec)
加入成功后的数据内容如下:
mysql> select * from setup_actors;
+-----------+------+------+---------+---------+
| HOST | USER | ROLE | ENABLED | HISTORY |
+-----------+------+------+---------+---------+
| % | % | % | NO | NO |
| localhost | root | % | YES | YES |
+-----------+------+------+---------+---------+
2 rows in set (0.00 sec)
好了，setup_actors 的配置就这样，另外两个表的内容修改也是大同小异。
表 setup_consumers 描述各种事件，
表setup_instruments 描述这个数据库下的表名以及是否开启监控。
我统计了一下，两个表的默认数据还不少。
setup_instruments 1006 rows
setup_consumers 15 rows
我们按照官方的建议来修改，可以看到修改的不是一行，而是相关的很多行。
mysql> UPDATE performance_schema.setup_instruments
-> SET ENABLED = 'YES', TIMED = 'YES'
-> WHERE NAME LIKE '%statement/%';
Query OK, 0 rows affected (0.00 sec)
Rows matched: 193 Changed: 0 Warnings: 0
mysql> UPDATE performance_schema.setup_instruments
-> SET ENABLED = 'YES', TIMED = 'YES'
-> WHERE NAME LIKE '%stage/%';
Query OK, 120 rows affected (0.00 sec)
Rows matched: 129 Changed: 120 Warnings: 0
mysql> UPDATE performance_schema.setup_consumers
-> SET ENABLED = 'YES'
-> WHERE NAME LIKE '%events_statements_%';
Query OK, 1 row affected (0.00 sec)
Rows matched: 3 Changed: 1 Warnings: 0
mysql> UPDATE performance_schema.setup_consumers
-> SET ENABLED = 'YES'
-> WHERE NAME LIKE '%events_stages_%';
Query OK, 3 rows affected (0.00 sec)
Rows matched: 3 Changed: 3 Warnings: 0
OK，配置完成，我们来看一下怎么用。创建一个test数据库。
mysql> create database test;
Query OK, 1 row affected (0.00 sec)
切换到test数据库
mysql> use test
Database changed
创建一个测试表test_profile，插入几行数据。
mysql> create table test_profile as select * from information_schema.columns limit 1,5;
Query OK, 5 rows affected (0.10 sec)
Records: 5 Duplicates: 0 Warnings: 0
运行语句来得到一些详细的统计信息。
mysql> select * from test.test_profile limit 1,2 \G
*************************** 1. row ***************************
TABLE_CATALOG: def
TABLE_SCHEMA: information_schema
TABLE_NAME: CHARACTER_SETS
COLUMN_NAME: DESCRIPTION
ORDINAL_POSITION: 3
COLUMN_DEFAULT:
IS_NULLABLE: NO
DATA_TYPE: varchar
CHARACTER_MAXIMUM_LENGTH: 60
CHARACTER_OCTET_LENGTH: 180
NUMERIC_PRECISION: NULL
NUMERIC_SCALE: NULL
DATETIME_PRECISION: NULL
CHARACTER_SET_NAME: utf8
COLLATION_NAME: utf8_general_ci
COLUMN_TYPE: varchar(60)
COLUMN_KEY:
EXTRA:
PRIVILEGES: select
COLUMN_COMMENT:
GENERATION_EXPRESSION:
*************************** 2. row ***************************
TABLE_CATALOG: def
TABLE_SCHEMA: information_schema
TABLE_NAME: CHARACTER_SETS
COLUMN_NAME: MAXLEN
ORDINAL_POSITION: 4
COLUMN_DEFAULT: 0
IS_NULLABLE: NO
DATA_TYPE: bigint
CHARACTER_MAXIMUM_LENGTH: NULL
CHARACTER_OCTET_LENGTH: NULL
NUMERIC_PRECISION: 19
NUMERIC_SCALE: 0
DATETIME_PRECISION: NULL
CHARACTER_SET_NAME: NULL
COLLATION_NAME: NULL
COLUMN_TYPE: bigint(3)
COLUMN_KEY:
EXTRA:
PRIVILEGES: select
COLUMN_COMMENT:
GENERATION_EXPRESSION:
2 rows in set (0.00 sec)
根据下面的语句查询一个历史表，从表名可以看出是和事件相关的，感觉越来越像Oracle了。。。
mysql> SELECT EVENT_ID, TRUNCATE(TIMER_WAIT/1000000000000,6) as Duration, SQL_TEXT
-> FROM performance_schema.events_statements_history_long
-> WHERE SQL_TEXT like '%limit 1,2%';
+----------+----------+-------------------------------------------+
| EVENT_ID | Duration | SQL_TEXT |
+----------+----------+-------------------------------------------+
| 4571 | 0.000196 | select * from test.test_profile limit 1,2 |
| 4587 | 0.000214 | select * from test.test_profile limit 1,2 |
| 4603 | 0.000211 | select * from test.test_profile limit 1,2 |
+----------+----------+-------------------------------------------+
3 rows in set (0.04 sec)
我们通过上面的语句可以得到一个概览，对应的事件和执行时间。然后到stage相关的历史表中查看事件的详细信息，这就是我们期望的性能数据。如此一来应该就明白上面的配置表中所要做的工作是什么意思了。
mysql> SELECT event_name AS Stage, TRUNCATE(TIMER_WAIT/1000000000000,6) AS Duration
-> FROM performance_schema.events_stages_history_long
-> WHERE NESTING_EVENT_ID=4603;
+--------------------------------+----------+
| Stage | Duration |
+--------------------------------+----------+
| stage/sql/starting | 0.000064 |
| stage/sql/checking permissions | 0.000004 |
| stage/sql/Opening tables | 0.000023 |
| stage/sql/init | 0.000023 |
| stage/sql/System lock | 0.000005 |
| stage/sql/optimizing | 0.000002 |
| stage/sql/statistics | 0.000007 |
| stage/sql/preparing | 0.000006 |
| stage/sql/executing | 0.000001 |
| stage/sql/Sending data | 0.000046 |
| stage/sql/end | 0.000002 |
| stage/sql/query end | 0.000003 |
| stage/sql/closing tables | 0.000005 |
| stage/sql/freeing items | 0.000012 |
| stage/sql/cleaning up | 0.000000 |
+--------------------------------+----------+
15 rows in set (0.01 sec)
整体来看，看到这个特性的输出，十分类似于Oracle中的Datapump，输出实在是太像了，很有条理。