MySQL性能优化-in和exists

一直听说exists性能比in快，但为啥快，一直不明白，乘着今天重点研究mysql，看到底是不是这么回事，原因又是在哪里。

1、我们先准备2张表和数据，人员表插入100W条数据，部门表插入5条数据。

DROP TABLE IF EXISTS `bd_dept`;
CREATE TABLE `bd_dept`  (
  `id` int(0) NOT NULL,
  `dept_name` varchar(20) CHARACTER SET utf8 COLLATE utf8_general_ci NULL DEFAULT NULL COMMENT '部门名称',PRIMARY KEY (`id`) USING BTREE
) ENGINE = InnoDB CHARACTER SET = utf8 COLLATE = utf8_general_ci COMMENT = '部门' ROW_FORMAT = Dynamic;

DROP TABLE IF EXISTS `bd_user`;
CREATE TABLE `bd_user`  (
  `id` bigint(0) NOT NULL,
  `dept_id` int(0) NULL DEFAULT NULL COMMENT '部门ID',
  `user_name` varchar(20) CHARACTER SET utf8 COLLATE utf8_general_ci NULL DEFAULT NULL COMMENT '用户名称',PRIMARY KEY (`id`) USING BTREE,INDEX `dept_id`(`dept_id`) USING BTREE,
  CONSTRAINT `bd_user_ibfk_1` FOREIGN KEY (`dept_id`) REFERENCES `bd_dept` (`id`) ON DELETE RESTRICT ON UPDATE RESTRICT
) ENGINE = InnoDB CHARACTER SET = utf8 COLLATE = utf8_general_ci COMMENT = '人员' ROW_FORMAT = DYNAMIC;

2、我们测试一下结果

  1）根据部门名称找所有该部门下的人员信息：

select * from bd_user a where exists (select 1 from bd_dept b where b.dept_name='研发中心' and b.id=a.dept_id)
> OK
> Time: 0.814s

select * from bd_user a where a.dept_id in (select id from bd_dept b where b.dept_name='研发中心')
> OK
> Time: 0.8s

  从结果上看，in和exists查询效率基本一致，多次执行执行查询，看不出来谁快谁慢。我们看下他们的执行计划：

mysql> explain select * from bd_user a where exists (select 1 from bd_dept b where b.dept_name='研发中心' and b.id=a.dept_id);
+----+-------------+-------+------------+------+---------------+---------+---------+---------------+--------+----------+-------------+
| id | select_type | table | partitions | type | possible_keys | key     | key_len | ref           | rows   | filtered | Extra       |
+----+-------------+-------+------------+------+---------------+---------+---------+---------------+--------+----------+-------------+
|  1 | SIMPLE      | b     | NULL       | ALL  | PRIMARY       | NULL    | NULL    | NULL          |      4 |    25.00 | Using where |
|  1 | SIMPLE      | a     | NULL       | ref  | dept_id       | dept_id | 5       | zhi_test.b.id | 332010 |   100.00 | NULL        |
+----+-------------+-------+------------+------+---------------+---------+---------+---------------+--------+----------+-------------+

mysql> explain select * from bd_user a where a.dept_id in (select id from bd_dept b where b.dept_name='研发中心');
+----+-------------+-------+------------+------+---------------+---------+---------+---------------+--------+----------+-------------+
| id | select_type | table | partitions | type | possible_keys | key     | key_len | ref           | rows   | filtered | Extra       |
+----+-------------+-------+------------+------+---------------+---------+---------+---------------+--------+----------+-------------+
|  1 | SIMPLE      | b     | NULL       | ALL  | PRIMARY       | NULL    | NULL    | NULL          |      4 |    25.00 | Using where |
|  1 | SIMPLE      | a     | NULL       | ref  | dept_id       | dept_id | 5       | zhi_test.b.id | 332010 |   100.00 | NULL        |
+----+-------------+-------+------------+------+---------------+---------+---------+---------------+--------+----------+-------------+

  他们的执行计划完全相同。

  2）根据人员名称找他所在部门的信息：

select * from bd_dept a where exists (select 1 from bd_user b where b.user_name='测试0000000031' and a.id=b.dept_id)
> OK
> Time: 0.19s

select * from bd_dept a where a.id in (select b.dept_id from bd_user b where b.user_name='测试0000000031')
> OK
> Time: 0.19s

  从结果上看in和exists效率还是一样的，我们再看看执行计划

mysql> explain select * from bd_dept a where exists (select 1 from bd_user b where b.user_name='测试0000000031' and a.id=b.dept_id);
+----+--------------+-------------+------------+--------+---------------------+---------------------+---------+---------------+--------+----------+-------------+
| id | select_type  | table       | partitions | type   | possible_keys       | key                 | key_len | ref           | rows   | filtered | Extra       |
+----+--------------+-------------+------------+--------+---------------------+---------------------+---------+---------------+--------+----------+-------------+
|  1 | SIMPLE       | a           | NULL       | ALL    | PRIMARY             | NULL                | NULL    | NULL          |      4 |   100.00 | Using where |
|  1 | SIMPLE       | <subquery2> | NULL       | eq_ref | <auto_distinct_key> | <auto_distinct_key> | 5       | zhi_test.a.id |      1 |   100.00 | NULL        |
|  2 | MATERIALIZED | b           | NULL       | ALL    | dept_id             | NULL                | NULL    | NULL          | 996030 |    10.00 | Using where |
+----+--------------+-------------+------------+--------+---------------------+---------------------+---------+---------------+--------+----------+-------------+

mysql> explain select * from bd_dept a where a.id in (select b.dept_id from bd_user b where b.user_name='测试0000000031');
+----+--------------+-------------+------------+--------+---------------------+---------------------+---------+---------------+--------+----------+-------------+
| id | select_type  | table       | partitions | type   | possible_keys       | key                 | key_len | ref           | rows   | filtered | Extra       |
+----+--------------+-------------+------------+--------+---------------------+---------------------+---------+---------------+--------+----------+-------------+
|  1 | SIMPLE       | a           | NULL       | ALL    | PRIMARY             | NULL                | NULL    | NULL          |      4 |   100.00 | Using where |
|  1 | SIMPLE       | <subquery2> | NULL       | eq_ref | <auto_distinct_key> | <auto_distinct_key> | 5       | zhi_test.a.id |      1 |   100.00 | NULL        |
|  2 | MATERIALIZED | b           | NULL       | ALL    | dept_id             | NULL                | NULL    | NULL          | 996030 |    10.00 | Using where |
+----+--------------+-------------+------------+--------+---------------------+---------------------+---------+---------------+--------+----------+-------------+

  这2个SQL的执行计划也完全相同。

3、结论

  实践出真知，in和exists的性能在MySQL8.x中是一样的。