您好,登录后才能下订单哦!
这篇文章给大家介绍如何理解MySQL执行计划中的各个参数及含义,内容非常详细,感兴趣的小伙伴们可以参考借鉴,希望对大家能有所帮助。
本文是对于MySQL执行计划的解析,主要解释了MySQL执行计划中的各个参数及含义。
显示可能应用在这张表中的索引,一个或者多个;
查询涉及到的字段上若存在索引,则该索引将被列出,但不一定被查询实际使用(可能自己创建了4个索引,在执行的时候,
可能根据内部的自动判断,只使用了3个)。
先创建索引 CREATE INDEX IDX_EMP_01 ON employee(dep_id); CREATE INDEX IDX_EMP_02 ON employee(dep_id,NAME); 查看索引 mysql> show index from employee; +----------+------------+------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+---------+------------+ | Table | Non_unique | Key_name | Seq_in_index | Column_name | Collation | Cardinality | Sub_part | Packed | Null | Index_type | Comment | Index_comment | Visible | Expression | +----------+------------+------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+---------+------------+ | employee | 0 | PRIMARY | 1 | id | A | 8 | NULL | NULL | | BTREE | | | YES | NULL | | employee | 1 | IDX_EMP_01 | 1 | dep_id | A | 4 | NULL | NULL | YES | BTREE | | | YES | NULL | | employee | 1 | IDX_EMP_02 | 1 | dep_id | A | 4 | NULL | NULL | YES | BTREE | | | YES | NULL | | employee | 1 | IDX_EMP_02 | 2 | name | A | 8 | NULL | NULL | YES | BTREE | | | YES | NULL | +----------+------------+------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+---------+------------+ 4 rows in set (0.06 sec)
可以看到表上有3个索引。
进行查询,并且查看执行计划:
EXPLAIN SELECT ID FROM EMPLOYEE WHERE DEP_ID = 1 AND NAME = '鲁班'; +----+-------------+----------+------------+------+-----------------------+------------+---------+-------------+------+----------+-------------+ | id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra | +----+-------------+----------+------------+------+-----------------------+------------+---------+-------------+------+----------+-------------+ | 1 | SIMPLE | EMPLOYEE | NULL | ref | IDX_EMP_01,IDX_EMP_02 | IDX_EMP_02 | 68 | const,const | 1 | 100.00 | Using index | +----+-------------+----------+------------+------+-----------------------+------------+---------+-------------+------+----------+-------------+ 1 row in set, 1 warning (0.05 sec)
表上有三个索引,其中索引IDX_EMP_01是dep_id的单列索引,IDX_EMP_02是列dep_id和列name的复合索引,
在进行查询时,可能会用到索引,因为有过滤条件dep_id=1,所以会考虑使用这两个索引,因为索引的第一列都是dep_id,
此时的possiable_keys为IDX_EMP_01和IDX_EMP_02。
实际使用的索引,如果为NULL,则没有使用索引 ;
查询中若使用了覆盖索引 ,则该索引仅出现在key列表中。
possible_keys与key关系:possiable_keys表示理论应该用到哪些索引,key表示实际用到了哪些索引。
还是和上面一样:
EXPLAIN SELECT ID FROM EMPLOYEE WHERE DEP_ID = 1 AND NAME = '鲁班'; +----+-------------+----------+------------+------+-----------------------+------------+---------+-------------+------+----------+-------------+ | id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra | +----+-------------+----------+------------+------+-----------------------+------------+---------+-------------+------+----------+-------------+ | 1 | SIMPLE | EMPLOYEE | NULL | ref | IDX_EMP_01,IDX_EMP_02 | IDX_EMP_02 | 68 | const,const | 1 | 100.00 | Using index | +----+-------------+----------+------------+------+-----------------------+------------+---------+-------------+------+----------+-------------+ 1 row in set, 1 warning (0.05 sec)
在possiable_keys里显示了两个索引,但是呢,并不是这两个索引都会使用,用哪一个需要由优化器自己决定。
通过key可以发现,使用的是IDX_EMP_02这个索引,因为where里同时有dep_id和name条件,
因此使用这个索引最为合理,效率最高。
通过key可以发现,对该表的真实的索引的使用情况。
表示索引中使用的字节数,可通过该列计算查询中使用的索引长度。
EXPLAIN SELECT ID FROM EMPLOYEE WHERE DEP_ID = 1; +----+-------------+----------+------------+------+-----------------------+------------+---------+-------+------+----------+-------------+ | id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra | +----+-------------+----------+------------+------+-----------------------+------------+---------+-------+------+----------+-------------+ | 1 | SIMPLE | EMPLOYEE | NULL | ref | IDX_EMP_01,IDX_EMP_02 | IDX_EMP_01 | 5 | const | 3 | 100.00 | Using index | +----+-------------+----------+------------+------+-----------------------+------------+---------+-------+------+----------+-------------+ 1 row in set, 1 warning (0.00 sec)
可以看到只使用索引的第一列时,长度为5个字节。
EXPLAIN SELECT ID FROM EMPLOYEE WHERE DEP_ID = 1 AND NAME = '鲁班'; +----+-------------+----------+------------+------+-----------------------+------------+---------+-------------+------+----------+-------------+ | id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra | +----+-------------+----------+------------+------+-----------------------+------------+---------+-------------+------+----------+-------------+ | 1 | SIMPLE | EMPLOYEE | NULL | ref | IDX_EMP_01,IDX_EMP_02 | IDX_EMP_02 | 68 | const,const | 1 | 100.00 | Using index | +----+-------------+----------+------------+------+-----------------------+------------+---------+-------------+------+----------+-------------+ 1 row in set, 1 warning (0.05 sec)
对于使用了整个索引时,显示的是该索引的大小,key_len由5变成了68,说明name这个字段在索引里占用的字节数为63。
并且,key_len显示的值是where里的索引的具体大小,不包括order by或者group by时使用的索引。
索引是否被引入到, 到底引用到了哪几个索引。
ref列用来显示使用哪个列或常数与key一起从表中选择相应的行。它显示的列的名字(或const),此列多数时候为null。
EXPLAIN SELECT D.ID ,D.ADDRESS FROM DEPARTMENT D ,EMPLOYEE E WHERE D.ID = E.DEP_ID; +----+-------------+-------+------------+------+-----------------------+------------+---------+-----------+------+----------+-------------+ | id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra | +----+-------------+-------+------------+------+-----------------------+------------+---------+-----------+------+----------+-------------+ | 1 | SIMPLE | D | NULL | ALL | PRIMARY | NULL | NULL | NULL | 5 | 100.00 | NULL | | 1 | SIMPLE | E | NULL | ref | IDX_EMP_01,IDX_EMP_02 | IDX_EMP_01 | 5 | demo.D.id | 2 | 100.00 | Using index | +----+-------------+-------+------------+------+-----------------------+------------+---------+-----------+------+----------+-------------+ 2 rows in set, 1 warning (0.01 sec)
Ref是用来表示索引是否被使用,到底用了哪个索引,从执行计划可看到,对表d进行扫描时,没有使用索引,type为all,
是一个全表扫描,因此ref值为null。对表e的扫描时,type为ref,表示使用了索引,并且通过key可以看到确实使用了索引,
因此在ref列需要被标记,索引被引用了,并且把该索引引用了到了d表的id字段上。
rows列显示的是mysql解析器认为执行此SQL时必须扫描的行数。此数值为一个预估值,不是具体值,通常比实际值小。
没有建索引前:
EXPLAIN SELECT D.ID ,E.DEP_ID FROM DEPARTMENT D ,EMPLOYEE E WHERE D.ID = E.DEP_ID; +----+-------------+-------+------------+-------+---------------+---------+---------+------+------+----------+----------------------------------------------------+ | id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra | +----+-------------+-------+------------+-------+---------------+---------+---------+------+------+----------+----------------------------------------------------+ | 1 | SIMPLE | D | NULL | index | PRIMARY | PRIMARY | 4 | NULL | 5 | 100.00 | Using index | | 1 | SIMPLE | E | NULL | ALL | NULL | NULL | NULL | NULL | 8 | 12.50 | Using where; Using join buffer (Block Nested Loop) | +----+-------------+-------+------------+-------+---------------+---------+---------+------+------+----------+----------------------------------------------------+ 2 rows in set, 1 warning (0.00 sec)
通过执行计划可以看到对表d采用索引全扫的扫描方式,需要对整个索引进行扫描;对表e采用全表扫描的方式,
因此需要对整个数据集进行扫描,d表有5行,e表有8行,因此在rows里显示为5和8.
对e表的dep_id列创建索引:
create index idx_employee_01 on employee(dep_id); 再次查询: +----+-------------+-------+------------+-------+-----------------+-----------------+---------+-----------+------+----------+-------------+ | id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra | +----+-------------+-------+------------+-------+-----------------+-----------------+---------+-----------+------+----------+-------------+ | 1 | SIMPLE | D | NULL | index | PRIMARY | PRIMARY | 4 | NULL | 5 | 100.00 | Using index | | 1 | SIMPLE | E | NULL | ref | idx_employee_01 | idx_employee_01 | 5 | demo.D.id | 2 | 100.00 | Using index | +----+-------------+-------+------------+-------+-----------------+-----------------+---------+-----------+------+----------+-------------+ 2 rows in set, 1 warning (0.00 sec)
通过执行计划可以看到,对d表采用索引全扫的方式,需要对整个索引进行扫描;对表e采用ref的非唯一性索引扫描,
只需要扫描索引的部分数据,d表有5行数据,所以rows为5,对于e表,扫描满足条件的数据,rows为2。
表d的数据: +----+----------------+---------+ | id | deptName | address | +----+----------------+---------+ | 1 | 研发部(RD) | 2层 | | 2 | 人事部(HR) | 3层 | | 3 | 市场部(MK) | 4层 | | 4 | 后勤部(MIS) | 5层 | | 5 | 财务部(FD) | 6层 | +----+----------------+---------+ 5 rows in set (0.07 sec) 表e的数据: +----+-----------+--------+------+---------+--------+ | id | name | dep_id | age | salary | cus_id | +----+-----------+--------+------+---------+--------+ | 1 | 鲁班 | 1 | 10 | 1000.00 | 1 | | 2 | 后裔 | 1 | 20 | 2000.00 | 1 | | 3 | 孙尚香 | 1 | 20 | 2500.00 | 1 | | 4 | 凯 | 4 | 20 | 3000.00 | 1 | | 5 | 典韦 | 4 | 40 | 3500.00 | 2 | | 6 | 貂蝉 | 6 | 20 | 5000.00 | 1 | | 7 | 孙膑 | 6 | 50 | 5000.00 | 1 | | 8 | 蔡文姬 | 30 | 35 | 4000.00 | 1 | +----+-----------+--------+------+---------+--------+ 8 rows in set (0.00 sec) 对d表的id和e表的dep_id查询: +----+--------+ | ID | DEP_ID | +----+--------+ | 1 | 1 | | 1 | 1 | | 1 | 1 | | 4 | 4 | | 4 | 4 | +----+--------+ 5 rows in set (0.00 sec)
Ref是对传过来的数据进行索引的扫描,d表的id有5行,e的dep_id值有多个,在进行等值匹配时,只有1和4满足条件。
对表d先进性操作,先扫描这5行数据,然后把1和4传给e表,所以对于e表只需要对1和4进行索引的扫描,只需要扫描两行。
满足查询的记录数量的比例,注意是百分比,不是具体记录数;值越大越好,filtered列的值依赖统计信息,并不十分准确。
对全表进行扫描时:
EXPLAIN SELECT * FROM EMPLOYEE; +----+-------------+----------+------------+------+---------------+------+---------+------+------+----------+-------+ | id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra | +----+-------------+----------+------------+------+---------------+------+---------+------+------+----------+-------+ | 1 | SIMPLE | EMPLOYEE | NULL | ALL | NULL | NULL | NULL | NULL | 8 | 100.00 | NULL | +----+-------------+----------+------------+------+---------------+------+---------+------+------+----------+-------+ 1 row in set, 1 warning (0.00 sec)
进行全表扫描时,会对所有的数据进行扫描,此时filtered为100.
当进行范围查询时:
EXPLAIN SELECT * FROM EMPLOYEE WHERE DEP_ID > 4; +----+-------------+----------+------------+------+---------------+------+---------+------+------+----------+-------------+ | id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra | +----+-------------+----------+------------+------+---------------+------+---------+------+------+----------+-------------+ | 1 | SIMPLE | EMPLOYEE | NULL | ALL | NULL | NULL | NULL | NULL | 8 | 33.33 | Using where | +----+-------------+----------+------------+------+---------------+------+---------+------+------+----------+-------------+ 1 row in set, 1 warning (0.00 sec)
对dep_id大于4的数据进行查询,也是全表扫描,返回的数据为3行,优化器估计返回了三分之一的数据,标记为33.33。
再一次进行查询:
EXPLAIN SELECT * FROM EMPLOYEE WHERE DEP_ID > 1; +----+-------------+----------+------------+------+---------------+------+---------+------+------+----------+-------------+ | id | select_type | table | partitions | type | possible_keys | key | key_len | ref | rows | filtered | Extra | +----+-------------+----------+------------+------+---------------+------+---------+------+------+----------+-------------+ | 1 | SIMPLE | EMPLOYEE | NULL | ALL | NULL | NULL | NULL | NULL | 8 | 33.33 | Using where | +----+-------------+----------+------------+------+---------------+------+---------+------+------+----------+-------------+ 1 row in set, 1 warning (0.00 sec)
对dep_id大于1的数据进行查询,也是全表扫描,返回的数据为5行,但是还是被认定filterd为33.33。
关于如何理解MySQL执行计划中的各个参数及含义就分享到这里了,希望以上内容可以对大家有一定的帮助,可以学到更多知识。如果觉得文章不错,可以把它分享出去让更多的人看到。
免责声明:本站发布的内容(图片、视频和文字)以原创、转载和分享为主,文章观点不代表本网站立场,如果涉及侵权请联系站长邮箱:is@yisu.com进行举报,并提供相关证据,一经查实,将立刻删除涉嫌侵权内容。