MySQL的子查询一直以来都是性能差的著称，解决方法是以join代替。
 MySQL5.5版本中该查询先把accessLog表中版本为2.2的数据线过滤出来，然后每个符合条件的数据都与内表进行一次select id from accessLog_01，因此性能低下。MySQL5.5采取的解决方法是将in重写为exists。
 在MySQL5.6/5.7版本中，子查询执行计划是将in/exists重写为join，如下看执行计划：

点击(此处)折叠或打开

1. mysql> select version();
   
2. +------------+
   
3. | version() |
   
4. +------------+
   
5. | 5.7.18-log |
   
6. +------------+

点击(此处)折叠或打开

1. mysql> explain select * from accessLog ac where ac.id in (select id from accessLog_01);
   
2. +----+--------------+--------------+------------+--------+---------------+---------+---------+----------------+------+----------+-------+
   
3. | id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra |
   
4. +----+--------------+--------------+------------+--------+---------------+---------+---------+----------------+------+----------+-------+
   
5. | 1 | SIMPLE | <subquery2> | NULL | ALL | NULL | NULL | NULL | NULL | NULL | 100.00 | NULL |
   
6. | 1 | SIMPLE | ac | NULL | eq_ref | PRIMARY | PRIMARY | 8 | <subquery2>.id | 1 | 100.00 | NULL |
   
7. | 2 | MATERIALIZED | accessLog_01 | NULL | ALL | NULL | NULL | NULL | NULL | 1305 | 100.00 | NULL |
   
8. +----+--------------+--------------+------------+--------+---------------+---------+---------+----------------+------+----------+-------+

点击(此处)折叠或打开

1. mysql> explain select * from accessLog ac where exists (select * from accessLog_01);
   
2. +----+-------------+--------------+------------+------+---------------+------+---------+------+--------+----------+-------+
   
3. | id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra |
   
4. +----+-------------+--------------+------------+------+---------------+------+---------+------+--------+----------+-------+
   
5. | 1 | PRIMARY | ac | NULL | ALL | NULL | NULL | NULL | NULL | 586090 | 100.00 | NULL |
   
6. | 2 | SUBQUERY | accessLog_01 | NULL | ALL | NULL | NULL | NULL | NULL | 1305 | 100.00 | NULL |
   
7. +----+-------------+--------------+------------+------+---------------+------+---------+------+--------+----------+-------+

点击(此处)折叠或打开

1. mysql> explain select ac.* from accessLog ac join accessLog_01 b on ac.id=b.id;
   
2. +----+-------------+-------+------------+--------+---------------+---------+---------+--------------+------+----------+-------+
   
3. | id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra |
   
4. +----+-------------+-------+------------+--------+---------------+---------+---------+--------------+------+----------+-------+
   
5. | 1 | SIMPLE | b | NULL | ALL | NULL | NULL | NULL | NULL | 1305 | 100.00 | NULL |
   
6. | 1 | SIMPLE | ac | NULL | eq_ref | PRIMARY | PRIMARY | 8 | xinhost.b.id | 1 | 100.00 | NULL |
   
7. +----+-------------+-------+------------+--------+---------------+---------+---------+--------------+------+----------+-------+

  办连接查询优化默认开启，通过show variables like 'optimizer_switch' \G查询：

点击(此处)折叠或打开

1. mysql> show variables like 'optimizer_switch' \G
   
2. *************************** 1. row ***************************
   
3. Variable_name: optimizer_switch
   
4.         Value: index_merge=on,index_merge_union=on,index_merge_sort_union=on,index_merge_intersection=on,engine_condition_pushdown=on,index_condition_pushdown=on,mrr=on,mrr_cost_based=on,block_nested_loop=on,batched_key_access=off,materialization=on,semijoin=on,loosescan=on,firstmatch=on,duplicateweedout=on,subquery_materialization_cost_based=on,use_index_extensions=on,condition_fanout_filter=on,derived_merge=on
   
5. 1 row in set (0.01 sec)

  但是半连接优化只是针对查询，对于DML操作，性能依旧很差。