mysql5.7关于使用到OR是否会用到索引并提高查询效率的探讨

相信很多人在mysql中看到了where条件中使用到了or就会以为这样是不会走索引的，通常会使用union all或者in 来进行优化，事实并不是想象的这样具体问题具体分析。

下面我们来看看

首先我们用sysbench生成两个100w行的表

表结构如下

mysql> show create table sbtest1 \g;
*************************** 1. row ***************************
       table: sbtest1
create table: create table `sbtest1` (
  `id` int(11) not null auto_increment,
  `k` int(11) not null default '0',
  `c` char(120) not null default '',
  `pad` char(60) not null default '',
  primary key (`id`),
  key `k_1` (`k`),
  key `c_1` (`c`)
) engine=innodb auto_increment=1000001 default charset=latin1
1 row in set (0.00 sec)

error: 
no query specified

mysql> show create table sbtest2 \g;
*************************** 1. row ***************************
       table: sbtest2
create table: create table `sbtest2` (
  `id` int(11) not null auto_increment,
  `k` int(11) not null default '0',
  `c` char(120) not null default '',
  `pad` char(60) not null default '',
  primary key (`id`),
  key `k_2` (`k`),
  key `c_2` (`c`)
) engine=innodb auto_increment=1000001 default charset=latin1
1 row in set (0.00 sec)

error: 
no query specified

1.首先我们使用同一列带索引字段的进行查询。

mysql> explain select * from sbtest1 where k='501462' or k='502480';   
+----+-------------+---------+------------+-------+---------------+------+---------+------+------+----------+-----------------------+
| id | select_type | table   | partitions | type  | possible_keys | key  | key_len | ref  | rows | filtered | extra                 |
+----+-------------+---------+------------+-------+---------------+------+---------+------+------+----------+-----------------------+
|  1 | simple      | sbtest1 | null       | range | k_1           | k_1  | 4       | null |  214 |   100.00 | using index condition |
+----+-------------+---------+------------+-------+---------------+------+---------+------+------+----------+-----------------------+

从执行计划中看出这样是可以使用到索引的，另外我们使用in 或者union all来看。

mysql> explain select pad from sbtest1 where k in ('501462','502480');
+----+-------------+---------+------------+-------+---------------+------+---------+------+------+----------+-----------------------+
| id | select_type | table   | partitions | type  | possible_keys | key  | key_len | ref  | rows | filtered | extra                 |
+----+-------------+---------+------------+-------+---------------+------+---------+------+------+----------+-----------------------+
|  1 | simple      | sbtest1 | null       | range | k_1           | k_1  | 4       | null |  214 |   100.00 | using index condition |
+----+-------------+---------+------------+-------+---------------+------+---------+------+------+----------+-----------------------+

in的执行计划和or相同。

mysql>  explain select pad from sbtest1 where k='501462' union all select pad from sbtest1 where k='502480';
+----+-------------+---------+------------+------+---------------+------+---------+-------+------+----------+-------+
| id | select_type | table   | partitions | type | possible_keys | key  | key_len | ref   | rows | filtered | extra |
+----+-------------+---------+------------+------+---------------+------+---------+-------+------+----------+-------+
|  1 | primary     | sbtest1 | null       | ref  | k_1           | k_1  | 4       | const |  113 |   100.00 | null  |
|  2 | union       | sbtest1 | null       | ref  | k_1           | k_1  | 4       | const |  101 |   100.00 | null  |

虽然执行计划不通但union all估计的查询行数和上面相同。

2.我们再来看看不同列带索引字段的进行查询

mysql> explain select pad from sbtest1 where  k='501462' or c='68487932199-96439406143-93774651418-41631865787-96406072701-20604855487-25459966574-28203206787-41238978918-19503783441';
+----+-------------+---------+------------+-------------+---------------+---------+---------+------+------+----------+-----------------------------------+
| id | select_type | table   | partitions | type        | possible_keys | key     | key_len | ref  | rows | filtered | extra                             |
+----+-------------+---------+------------+-------------+---------------+---------+---------+------+------+----------+-----------------------------------+
|  1 | simple      | sbtest1 | null       | index_merge | k_1,c_1       | k_1,c_1 | 4,120   | null |  114 |   100.00 | using union(k_1,c_1); using where |
+----+-------------+---------+------------+-------------+---------------+---------+---------+------+------+----------+-----------------------------------

这样的情况也会使用索引

如果or的条件中有个条件不带索引的话，那这条sql就不会使用到索引了，如下。

mysql> explain select pad from sbtest1 where  k='501462' or pad='00592560354-80393027097-78244247549-39135306455-88936868384';
+----+-------------+---------+------------+------+---------------+------+---------+------+--------+----------+-------------+
| id | select_type | table   | partitions | type | possible_keys | key  | key_len | ref  | rows   | filtered | extra       |
+----+-------------+---------+------------+------+---------------+------+---------+------+--------+----------+-------------+
|  1 | simple      | sbtest1 | null       | all  | k_1           | null | null    | null | 986400 |    19.00 | using where |
+----+-------------+---------+------------+------+---------------+------+---------+------+--------+----------+-------------+

pad列没索引所以整条的sql就不会使用到索引

假设使用union all来改写一样需要全表扫描所以意义也不大，如下

mysql>  explain select pad from sbtest1 where  k='501462' union all select pad from sbtest1 where pad='00592560354-80393027097-78244247549-39135306455-88936868384';
+----+-------------+---------+------------+------+---------------+------+---------+-------+--------+----------+-------------+
| id | select_type | table   | partitions | type | possible_keys | key  | key_len | ref   | rows   | filtered | extra       |
+----+-------------+---------+------------+------+---------------+------+---------+-------+--------+----------+-------------+
|  1 | primary     | sbtest1 | null       | ref  | k_1           | k_1  | 4       | const |    113 |   100.00 | null        |
|  2 | union       | sbtest1 | null       | all  | null          | null | null    | null  | 986400 |    10.00 | using where |
+----+-------------+---------+------------+------+---------------+------+---------+-------+--------+----------+-------------+

3.接下来我们看看多表关联查询

 

mysql> explain select a.pad,b.pad from sbtest1 a,sbtest2 b where a.id=b.id and (a.c='123' or b.c='1234');                 
+----+-------------+-------+------------+--------+---------------+---------+---------+-----------+--------+----------+-------------+
| id | select_type | table | partitions | type   | possible_keys | key     | key_len | ref       | rows   | filtered | extra       |
+----+-------------+-------+------------+--------+---------------+---------+---------+-----------+--------+----------+-------------+
|  1 | simple      | a     | null       | all    | primary,c_1   | null    | null    | null      | 986400 |   100.00 | null        |
|  1 | simple      | b     | null       | eq_ref | primary,c_2   | primary | 4       | test.a.id |      1 |   100.00 | using where |
+----+-------------+-------+------------+--------+---------------+---------+---------+-----------+--------+----------+-------------+
2 rows in set, 1 warning (0.00 sec)

mysql> explain select a.pad,b.pad from sbtest1 a,sbtest2 b where a.id=b.id and (a.c='123' or a.c='1234'); 
+----+-------------+-------+------------+--------+---------------+---------+---------+-----------+------+----------+-----------------------+
| id | select_type | table | partitions | type   | possible_keys | key     | key_len | ref       | rows | filtered | extra                 |
+----+-------------+-------+------------+--------+---------------+---------+---------+-----------+------+----------+-----------------------+
|  1 | simple      | a     | null       | range  | primary,c_1   | c_1     | 120     | null      |    2 |   100.00 | using index condition |
|  1 | simple      | b     | null       | eq_ref | primary       | primary | 4       | test.a.id |    1 |   100.00 | null                  |
+----+-------------+-------+------------+--------+---------------+---------+---------+-----------+------+----------+-----------------------+
2 rows in set, 1 warning (0.00 sec)

mysql> 

可以看出在多表查询的情况下or条件如果不在同一个表内执行计划表a的查询不走索引。

我们试试看用union all来进行改写

mysql> explain select a.pad,a.c from sbtest1 a,sbtest2 b where a.id=b.id and a.c='123' union all select a.pad,a.c from sbtest1 a,sbtest2 b where a.id=b.id and b.c='1234';
+----+-------------+-------+------------+--------+---------------+---------+---------+-----------+------+----------+-------------+
| id | select_type | table | partitions | type   | possible_keys | key     | key_len | ref       | rows | filtered | extra       |
+----+-------------+-------+------------+--------+---------------+---------+---------+-----------+------+----------+-------------+
|  1 | primary     | a     | null       | ref    | primary,c_1   | c_1     | 120     | const     |    1 |   100.00 | null        |
|  1 | primary     | b     | null       | eq_ref | primary       | primary | 4       | test.a.id |    1 |   100.00 | using index |
|  2 | union       | b     | null       | ref    | primary,c_2   | c_2     | 120     | const     |    1 |   100.00 | using index |
|  2 | union       | a     | null       | eq_ref | primary       | primary | 4       | test.b.id |    1 |   100.00 | null        |
+----+-------------+-------+------------+--------+---------------+---------+---------+-----------+------+----------+-------------+

在or的条件不在同一个表的情况下 使用union all来改写扫描行数减少且会走索引。