首页 > 数据库 > 文库 > 正文

模式配置like %XXX% 优化

2024-09-07 22:12:54
字体:
来源:转载
供稿:网友
  在MySQL里,like'XXX%可以用到索引,但like '%XXX%'却不行,比如,以下这个案例:
   查看测试表行数:
  点击(此处)折叠或打开
 
  mysql> select count(*) from test03;
  +----------+
  | count(*) |
  +----------+
  | 117584   |
  +----------+
    两次like匹配对比:
  点击(此处)折叠或打开
 
  mysql> explain select count(*) from test03 where username like '1%';
  +----+-------------+--------+-------+-----------------+-----------------+---------+------+-------+--------------------------+
  | id | select_type | table  | type  | possible_keys   | key             | key_len | ref  | rows  | Extra                    |
  +----+-------------+--------+-------+-----------------+-----------------+---------+------+-------+--------------------------+
  | 1  | SIMPLE      | test03 | range | idx_test03_name | idx_test03_name | 302     | NULL | 58250 | Using where; Using index |
  +----+-------------+--------+-------+-----------------+-----------------+---------+------+-------+--------------------------+
  1 row in set (0.03 sec)
 
  mysql> explain select count(*) from test03 where username like '%1%';
  +----+-------------+--------+-------+---------------+-----------------+---------+------+--------+--------------------------+
  | id | select_type | table | type   | possible_keys | key             | key_len | ref  | rows   | Extra                    |
  +----+-------------+--------+-------+---------------+-----------------+---------+------+--------+--------------------------+
  | 1  | SIMPLE      | test03| index  | NULL          | idx_test03_name | 302     | NULL | 116500 | Using where; Using index |
  +----+-------------+--------+-------+---------------+-----------------+---------+------+--------+--------------------------+
  1 row in set (0.00 sec)
  优化思路:
   这个测试表中,id是主键,叶子节点上保存了数据,从索引中就可以去到select的的id的列,不必读取数据行(只有select字段正好就是索引,那么就用到了覆盖索引),通过覆盖索引,减少I/O,提高性能。
   优化之前的执行计划:
  点击(此处)折叠或打开
 
  mysql> explain select count(*) from test03 where username like '%1%';
  +----+-------------+--------+------+---------------+------+---------+------+------+-------------+
  | id | select_type | table  | type | possible_keys | key  | key_len | ref  | rows | Extra       |
  +----+-------------+--------+------+---------------+------+---------+------+------+-------------+
  | 1  | SIMPLE      | test03 | ALL  | NULL          | NULL | NULL    | NULL | 7164 | Using where |
  +----+-------------+--------+------+---------------+------+---------+------+------+-------------+
    优化之后的执行计划:
  点击(此处)折叠或打开
 
  mysql> explain select count(*) from test03 a join (select id from test03 where username like '%1%') b on a.id=b.id;
  +----+-------------+------------+--------+---------------+---------+---------+------+------+-------------+
  | id | select_type | table      | type   | possible_keys | key     | key_len | ref  | rows | Extra       |
  +----+-------------+------------+--------+---------------+---------+---------+------+------+-------------+
  | 1  | PRIMAR      | <derived2> | ALL    | NULL          | NULL    | NULL    | NULL | 7164 | NULL        |
  | 1  | PRIMARY     | a          | eq_ref | PRIMARY       | PRIMARY | 8       | b.id | 1    | Using index |
  | 2  | DERIVED     | test03     | ALL    | NULL          | NULL    | NULL    | NULL | 7164 | Using where |
  +----+-------------+------------+--------+---------------+---------+---------+------+------+-------------+

(编辑:武林网)

发表评论 共有条评论
用户名: 密码:
验证码: 匿名发表