MySQL使用索引优化DISTINCT操作

MySQL通常使用GROUPBY(本质上是排序动作)完成DISTINCT操作,如果DISTINCT操作和ORDERBY操作组合使用,通常会用到临时表.这样会影响性能. 在一些情况下,MySQL可以使用索引优化DISTINCT操作,但需要活学活用.本文涉及一个不能利用索引完成DISTINCT操作的实例.
实例1 使用索引优化DISTINCT操作create table m11 (a int, b int, c int, d int, primary key(a)) engine=INNODB;insert into m11 values (1,1,1,1),(2,2,2,2),(3,3,3,3),(4,4,4,4),(5,5,5,5),(6,6,6,6),(7,7,7,7),(8,8,8,8);explain select distinct(a) from m11;

mysql> explain select distinct(a) from m11;+----+-------------+-------+------------+-------+---------------+---------+---------+------+------+----------+-------------+| id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra |+----+-------------+-------+------------+-------+---------------+---------+---------+------+------+----------+-------------+| 1 | SIMPLE | m11 | NULL | index | PRIMARY | PRIMARY | 4 | NULL | 1 | 100.00 | Using index |+----+-------------+-------+------------+-------+---------------+---------+---------+------+------+----------+-------------+

说明:

1 'a'列上存在主键索引,MySQL可以利用索引(key列值表明使用了主键索引)完成了DISTINCT操作.2 这是使用索引优化DISTINCT操作的典型实例.

实例2 使用索引不能优化DISTINCT操作create table m31 (a int, b int, c int, d int, primary key(a)) engine=MEMORY;insert into m31 values (1,1,1,1),(2,2,2,2),(3,3,3,3),(4,4,4,4),(5,5,5,5),(6,6,6,6),(7,7,7,7),(8,8,8,8);explain select distinct(a) from m31;

mysql> explain select distinct(a) from m31;+----+-------------+-------+------------+------+---------------+------+---------+------+------+----------+-------+| id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra |+----+-------------+-------+------------+------+---------------+------+---------+------+------+----------+-------+| 1 | SIMPLE | m31 | NULL | ALL | NULL | NULL | NULL | NULL | 8 | 100.00 | NULL |+----+-------------+-------+------------+------+---------------+------+---------+------+------+----------+-------+

说明:

1 从查询执行计划看,索引没有被使用.2 对比实例1的建表语句,只是存储引擎不同.3 为什么主键索引没有起作用? 难道MEMORY存储引擎上的索引不可使用?
实例3 使用索引可以优化DISTINCT操作的Memory表create table m33 (a int, b int, c int, d int, INDEX USING BTREE (a)) engine=MEMORY;insert into m33 values (1,1,1,1),(2,2,2,2),(3,3,3,3),(4,4,4,4),(5,5,5,5),(6,6,6,6),(7,7,7,7),(8,8,8,8);explain select distinct(a) from m33;

mysql> explain select distinct(a) from m33;+----+-------------+-------+------------+-------+---------------+------+---------+------+------+----------+-------+| id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra |+----+-------------+-------+------------+-------+---------------+------+---------+------+------+----------+-------+| 1 | SIMPLE | m33 | NULL | index | NULL | a | 5 | NULL | 8 | 100.00 | NULL |+----+-------------+-------+------------+-------+---------------+------+---------+------+------+----------+-------+

说明:

1 'a'列上存在主键索引,MySQL可以利用索引(key列值表明使用了主键索引)完成了DISTINCT操作.2 对比实例2,可以发现,二者都使用了Memory引擎. 但实例3指名使用Btree类型的索引.3 实例2没有指定使用什么类型的索引,MySQL将采用默认值. MySQL手册上说:As indicated by the engine name, MEMORY tables are stored in memory. They use hash indexes by default, which makes them very fast for single-value lookups, and very useful for creating temporary tables.
结论:1 看索引对查询的影响,要注意索引的类型.2 HASH索引适合等值查找,但不适合需要有序的场景,而Btree却适合有序的场景.3 看查询执行计划,发现索引没有被使用,需要进一步考察索引的类型.