问题语句

SELECT * FROMaWHERE `type` = 'appointment'
AND `event` = 14
AND EXISTS(SELECT * FROMbWHERE a.`sheet_id` =b.`id`AND `company_id` = 8
AND b.`deleted_at` IS NULL)ORDER BY a.id DESCLIMIT6;

解读执行计划

在exists类型的子查询的执行计划中，select_type一栏分别是PRIMARY和DEPENDENT SUBQUERY

DEPENDENT SUBQUERY的意思是：子查询，依赖于外层的查询;DEPEND SUBQUERY是依赖于SQL的主体部分，它的执行次数最大可能和SQL主体部分结果的行数一样多(这里因为有limit6，所以看起来主表扫描行数是6，如果去掉这个limit6的话，这个值是1500W)

上面这句话解释得通俗一点就是外连接先执行查询，然后把查询的结果集放入子查询内进行匹配；外查询每执行一次查询，就要来子查询匹配一次

join的执行计划中，select_type一栏都是simple

join的第一行的就是外表

从上面的对比可以看出，无论是 exists类型的子查询还是join，都基本可以看作遵循了第一行就是驱动表的规则(注意不是所有子查询都遵循这个规则，本篇只针对exists类型的dependent subquery)

分析

图一PRIMARY对应的表就是图二中的a表，DEPENDENT SUBQUERY表就是图二中的b表；a表有1500W行数据，b表有2W行数据

所以图一的SQL执行效率如此低下的原因就是大表驱动小表

优化

exists改写为join

1 SELECT a.*
2 FROM a join b on a.`sheet_id` =b.`id`3 WHERE a.`type` = 'appointment'
4 AND a.`event` = 14
5 AND b.`company_id` = 8
6 AND b.`deleted_at` IS NULL
7 ORDER BY a.`id` DESC
8 LIMIT 6;

由于a表作为内表，因此在a.`sheet_id`，a.`type`，a.`event`上创建联合索引；语句中出现了b表的本地谓词，所以b表的b.`company_id`，b.`deleted_at`上也要创建联合索引

优化结果，执行时间：117s→0.36s，性能提升了2000倍

这个语句有一个更极端的取值，在b.`company_id` = 2的时候，小表不会搜出任何满足条件的结果，在这种情况下，原语句执行时间在350s以上，而新语句仅需要0.03s，性能提升万倍

优化案例

今天优化的这批语句中，大多数是exists子查询的问题，可以看出这个研发小哥非常的喜欢用exists这种写法；前面的那个exists语句是泛用型，后面的exists语句加了些新花样

eg.
1 SELECT SUM(`xxxx`) ASag2 FROMa3 WHERE EXISTS(4 SELECT * FROMb5 WHERE a.`delivery_sheet_id` =b.`id`6 AND (`status` = 4
7 OR `is_rejected` = '1')8 AND `company_id` = 8
9 AND b.`deleted_at` IS NULL
10 )11 AND `status` IN (0, 4)12 AND `collection_type` IN (2, 3)13 AND a.`deleted_at` IS NULL;
or的优化通常改写union，但这里是求sum不能这么改，需要改写成2个语句然后求和；对应的列要建好索引
1 select c.ag+d.ag as ag from
2 (SELECT SUM(a.`xxxx`) ASag3 FROM a joinb4 on a.`delivery_sheet_id` =b.`id`5 where
6 b.`status` = 4
7 AND b.`company_id` = 8
8 AND b.`deleted_at` IS NULL
9 AND a.`status` IN (0, 4)10 AND a.`collection_type` IN (2, 3)11 AND a.`deleted_at` IS NULL) c,12 (13 SELECT SUM(a.`xxxx`) ASag14 FROM a joinb15 on a.`delivery_sheet_id` =b.`id`16 where
17 b.`is_rejected` = '1'
18 AND b.`company_id` = 8
19 AND b.`deleted_at` IS NULL
20 AND a.`status` IN (0, 4)21 AND a.`collection_type` IN (2, 3)22 AND a.`deleted_at` IS NULL) d;
优化结果，执行时间：18s→0.2s
in改写join的思路和exists差不多
这里没有现成的例子，粘贴一篇郑松华老师公众号的分析过来
原语句
1 SELECT
2
3 COUNT( * ) AStotalNum,4
5 sum( CASE WHEN F.ALARM_LEVEL = 1 THEN 1 ELSE 0 END ) ASLEVELS1,6
7 sum( CASE WHEN F.ALARM_LEVEL = 2 THEN 1 ELSE 0 END ) ASLEVELS2,8
9 sum( CASE WHEN F.ALARM_LEVEL = 3 THEN 1 ELSE 0 END ) ASLEVELS3,10
11 sum( CASE WHEN F.DEAL_STATE = 0 THEN 1 ELSE 0 END ) ASDESTS12
13 FROM
14
15 F16
17 LEFT JOIN DC ON DC.ID =F.CONST_ID18
19 LEFT JOIN V ON V.ID =F.VEHICLE_ID20
21 LEFT JOIN AREA ON AREA.ID =V.SYS_DIVISION_ID22
23 WHERE
24
25 DC.ID IS NOT NULL
26
27 AND V.ID IS NOT NULL
28
29 AND F.DEAL_STATE = 0
30
31 AND ALARM_LEVEL IN ( 1, 2, 3)32
33 AND F.VEHICLE_ID IN(34
35 SELECT
36
37 VEHICLE_ID38
39 FROM
40
41 GVLK42
43 WHERE
44
45 GROUP_ID IN ( SELECT GROUP_ID FROM GULK WHERE USER_ID = 'ff8080816091b09c0161f9b825750a9a')46
47 UNION
48
49 SELECT
50
51 VEHICLE_ID52
53 FROM
54
55 UVLK56
57 WHERE
58
59 USER_ID = 'ff8080816091b09c0161f9b825750a9a'
60
61 )62
63 AND date( F.ALARM_TIME ) BETWEEN '2000-01-01'
64
65 AND '2018-08-14'
66
67 AND AREA.PATH LIKE CONCAT( ( SELECT ARE.PATH FROM ARE WHERE ARE.ID = '0' ), '%' )
执行计划如下
改写如下(in改join)
1 explain extended2
3 SELECT
4
5 COUNT( * ) AStotalNum,6
7 sum( CASE WHEN F.ALARM_LEVEL = 1 THEN 1 ELSE 0 END ) ASLEVELS1,8
9 sum( CASE WHEN F.ALARM_LEVEL = 2 THEN 1 ELSE 0 END ) ASLEVELS2,10
11 sum( CASE WHEN F.ALARM_LEVEL = 3 THEN 1 ELSE 0 END ) ASLEVELS3,12
13 sum( CASE WHEN F.DEAL_STATE = 0 THEN 1 ELSE 0 END ) ASDESTS14
15 FROM
16
17 F18
19 straight_join (20
21 SELECT
22
23 VEHICLE_ID24
25 FROM
26
27 GVLK28
29 WHERE
30
31 GROUP_ID IN ( SELECT GROUP_ID FROM GULK WHERE USER_ID = 'ff8080816091b09c0161f9b825750a9a')32
33 UNION
34
35 SELECT
36
37 VEHICLE_ID38
39 FROM
40
41 UVLK42
43 WHERE
44
45 USER_ID = 'ff8080816091b09c0161f9b825750a9a'
46
47 ) s on F.VEHICLE_ID =s.VEHICLE_ID48
49 straight_join DC ON DC.ID =F.CONST_ID50
51 straight_join V ON V.ID =F.VEHICLE_ID52
53 straight_join AREA ON AREA.ID =V.SYS_DIVISION_ID54
55 WHERE
56
57 DC.ID IS NOT NULL
58
59 AND V.ID IS NOT NULL
60
61 AND F.DEAL_STATE = 0
62
63 AND ALARM_LEVEL IN ( 1, 2, 3)64
65 AND date( F.ALARM_TIME ) BETWEEN '2000-01-01'
66
67 AND '2018-08-14'
68
69 AND AREA.PATH LIKE CONCAT( ( SELECT ARE.PATH FROM ARE WHERE ARE.ID = '0' ), '%' )