根据 MySQL 状态优化 ---- 1. 慢查询

查看 MySQL 服务器运行的各种状态值：

mysql> show global status；

1. 慢查询

mysql> show variables like '%slow%';
+---------------------------+-------------------------------+
| Variable_name             | Value                         |
+---------------------------+-------------------------------+
| log_slow_admin_statements | OFF                           |
| log_slow_slave_statements | OFF                           |
| slow_launch_time          | 2                             |
| slow_query_log            | ON                            |
| slow_query_log_file       | /var/db/my3306/logs/slows.log |
+---------------------------+-------------------------------+
5 rows in set (0.00 sec)

mysql> show global status like '%slow%';
+---------------------+-------+
| Variable_name       | Value |
+---------------------+-------+
| Slow_launch_threads | 0     |
| Slow_queries        | 18384 |
+---------------------+-------+

打开慢查询日志可能会对系统性能有一点儿影响，如果你的 MySQL 是主-从结构，可以考虑打开其中一条从服务器的慢查询日志，这样既可以监控慢查询，对系统性能的影响也会很小。

显示哪些线程正在运行：

mysql> show processlist;
+------------+-----------------+--------------------+------------+-------------+----------+-----------------------------------------------------------------------+------------------+
| Id         | User            | Host               | db         | Command     | Time     | State                                                                 | Info             |
+------------+-----------------+--------------------+------------+-------------+----------+-----------------------------------------------------------------------+------------------+
|          1 | event_scheduler | localhost          | NULL       | Daemon      |       42 | Waiting for next activation                                           | NULL             |
| 1060236184 | my_sync         | 172.16.88.33:60726 | NULL       | Binlog Dump | 21244918 | Master has sent all binlog to slave; waiting for binlog to be updated | NULL             |
| 4874532774 | athena          | 172.16.88.12:46808 | NULL       | Sleep       |     6945 |                                                                       | NULL             |
| 4875812388 | california      | 172.16.88.11:59687 | california | Sleep       |        8 |                                                                       | NULL             |
| 4875812768 | california      | 172.16.88.11:59698 | california | Sleep       |        5 |                                                                       | NULL             |
| 4875813123 | athena          | 172.16.88.12:45683 | NULL       | Query       |        0 | init                                                                  | show processlist |
| 4875813758 | california      | 172.16.88.11:59711 | california | Sleep       |        1 |                                                                       | NULL             |
| 4875813759 | california      | 172.16.88.11:59713 | california | Sleep       |        1 |                                                                       | NULL             |
| 4875813760 | california      | 172.16.88.11:59712 | california | Sleep       |        1 |                                                                       | NULL             |
| 4875813915 | athena          | 172.16.88.12:50003 | athena     | Sleep       |        0 |                                                                       | NULL             |
| 4875813916 | athena          | 172.16.88.12:50004 | athena     | Sleep       |        0 |                                                                       | NULL             |
| 4875813917 | athena          | 172.16.88.12:50005 | athena     | Sleep       |        0 |                                                                       | NULL             |
+------------+-----------------+--------------------+------------+-------------+----------+-----------------------------------------------------------------------+------------------+
12 rows in set (0.00 sec)

若发现大量的 unauthenticated user 连接，数据库每次肯定都要响应，索引速速越来越慢，在 mysql.conf 里添加 skip-name-resolve 即可，表示不启用 DNS 反应解析。

再看 State 这列，若长期处于 Sending data (正在处理SELECT查询的记录，同时正在把结果发送给客户端)状态，则可怀疑是磁盘 I/O 压力过大，可用 explain 来分析语句。

mysql> explain SELECT count(new_cheat_id) FROM new_cheat WHERE account_id='14348612' AND offer_id='689'\G;
*************************** 1. row ***************************
id: 1
select_type: SIMPLE
table: new_cheat
type: ALL
possible_keys: NULL
key:NULL
key_len: NULL
ref: NULL
rows: 2529529
Extra: Using where
1 row in set (0.00 sec)

从结果可看出来此表没建索引，导致一次查询扫描了2529529行记录。

建立好索引后，用 show index 查看表索引：

mysql> show index from xk_order_ready_5;
+-----------+------------+-----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| Table     | Non_unique | Key_name  | Seq_in_index | Column_name | Collation | Cardinality | Sub_part | Packed | Null | Index_type | Comment | Index_comment |
+-----------+------------+-----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
| new_cheat |          0 | PRIMARY   |            1 | new_cheat_id| A         |     2577704 |     NULL | NULL   |      | BTREE      |         |               |
| new_cheat |          1 | ip        |            1 | ip          | A         |     1288852 |     NULL | NULL   |      | BTREE      |         |               |
| new_cheat |          1 | account_id|            1 | account_id  | A         |     1288852 |     NULL | NULL   |      | BTREE      |         |               |
+-----------+------------+-----------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+
3 rows in set (0.00 sec)

再次查看 explain 结果：

mysql> explain SELECT count(new_cheat_id) FROM new_cheat WHERE account_id='14348612' AND offer_id='689'\G;
*************************** 1. row ***************************
id: 1
select_type: SIMPLE
table: new_cheat
type: ref
possible_keys: account_id
key: account_id
key_len: 4
ref: const
rows: 6
Extra: Using where
1 row in set (0.00 sec)

可以发现直接读取了6条记录就获取了查询结果。