优化sql语句是优化数据库的一个很重要的方面，那么怎么发现那些耗时耗资源的sql语句呢，开启Mysql慢查询!

1.查看是否开启慢查询，默认情况下是关闭的。你的mysql最好在5.1版本以上

mysql> show variables like "%slow%";

1. +---------------------+---------------------------------------------------------
2. ------------------+
3. |Variable_name|Value
4. |
5. +---------------------+---------------------------------------------------------
6. ------------------+
7. | log_slow_queries | OFF
8. |
9. | slow_launch_time |2
10. |
11. | slow_query_log | OFF
12. |
13. | slow_query_log_file | D:\software\wamp\wamp\bin\mysql\mysql5.5.20\data\WIN-23J
14. S4C29BV4-slow.log |
15. +---------------------+---------------------------------------------------------
16. ------------------+

2.命令行开启mysql慢查询：

1. mysql>setglobal slow_query_log=on;
2. Query OK,0 rows affected (0.09 sec)

或者修改mysql配置文件开启mysql慢查询：(记得重启mysql)

Windows下开启MySQL慢查询
MySQL在Windows系统中的配置文件一般是是my.ini找到[mysqld]下面加上：
log-slow-queries = F:\MySQL\log\mysqlslowquery.log
long_query_time = 2

Linux下启用MySQL慢查询
MySQL在Windows系统中的配置文件一般是是my.cnf找到[mysqld]下面加上：
log-slow-queries=/data/mysqldata/slowquery.log
long_query_time=2

3.分析mysql慢查询日志：
mysql慢查询日志类似这样的：

1. /usr/local/mysql/libexec/mysqld,Version:5.1.26-rc-log (Source distribution). started with:
2. Tcp port:3306Unix socket:/tmp/mysql.sock
3. TimeIdCommandArgument
4. # Time: 100814 13:28:30
5. # User@Host: root[root] @ localhost []
6. # Query_time: 10.096500 Lock_time: 0.045791 Rows_sent: 1 Rows_examined: 2374192
7. SET timestamp=1281763710;
8. select count(distinct ad_code)as x from ad_visit_history where ad_code in(select ad_code from ad_list where media_id=15);
9. # Time: 100814 13:37:02
10. # User@Host: root[root] @ localhost []
11. # Query_time: 10.394134 Lock_time: 0.000091 Rows_sent: 1 Rows_examined: 2374192
12. SET timestamp=1281764222;
13. select count(distinct ad_code)as x from ad_visit_history where ad_code in(select ad_code from ad_list where media_id=15);
14. # Time: 100814 13:37:16
15. # User@Host: root[root] @ localhost []
16. # Query_time: 4.608920 Lock_time: 0.000078 Rows_sent: 1 Rows_examined: 1260544
17. SET timestamp=1281764236;
18. select count(*)as cou from ad_visit_history where ad_code in(select ad_code from ad_list where id=41) order by id desc;

看日志可以看出它记录了mysql慢查询语句的执行时间、锁定时间等等。
(1)用mysql自带的mysql自带的慢查询分析工具mysqldumpslow

(2)第三方分析工具mysqlsla
安装：