Mysql索引

Posted 2021-03-13 Books&Coffee

 

高性能的索引策略

1 独立的列

　　“独立的列”：索引列不能是表达式的一部分，也不能是参数的参数，必须将索引列单独放在比较符号的一侧

mysql> select actor_id from sakila.actor where actore_id + 1 = 5;【X】

mysql> select ... where to_days(current_date) - to_days(date_col) <= 10;【X】

　　

2 前缀索引和索引选择性

有时候需要索引很长的字符串列，但这样会让索引变得很大且慢。这时候需要索引字符串列的一部分字符串，这样可以大大节约索引空间，从而提高索引效率，但是这样有可能降低索引的选择性。

【索引选择性：不重复的索引值（基数）和数据表的记录总数（#T）的比值，范围从1/#T 到 1 之间，索引选择性越高，查询效率越高，因为选择性高的索引可以让mysql过滤掉更多的行】　　

我们需要通过特定的方法确认，当前缀索引长度为N时，这N个字符的选择性已经接近整个字段的选择性，如下面例子所示，当N为7时，再增加前缀长度，选择性的增加已经很小。

 

select
count(distinct left(city,3))/count(*) as sel3,
count(distinct left(city,4))/count(*) as sel4,
count(distinct left(city,5))/count(*) as sel5,
count(distinct left(city,6))/count(*) as sel6,
count(distinct left(city,7))/count(*) as sel7
from sakila.city;

+--------+--------+--------+--------+--------+

| sel3   | sel4   | sel5   | sel6   | sel7   |

+--------+--------+--------+--------+--------+

| 0.7633 | 0.9383 | 0.9750 | 0.9900 | 0.9933 |

+--------+--------+--------+--------+--------+

　　

增加前缀索引：mysql> alter table sakila.city add key(city(7));

-- 原始ddl，city字段上没有索引
 CREATE TABLE `city` (
  `city_id` smallint(5) unsigned NOT NULL AUTO_INCREMENT,
  `city` varchar(50) NOT NULL,
  `country_id` smallint(5) unsigned NOT NULL,
  `last_update` timestamp NOT NULL DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP,
  PRIMARY KEY (`city_id`),
  KEY `idx_fk_country_id` (`country_id`),
  KEY `city` (`city`(7)),
  CONSTRAINT `fk_city_country` FOREIGN KEY (`country_id`) REFERENCES `country` (`country_id`) ON UPDATE CASCADE
) ENGINE=InnoDB AUTO_INCREMENT=601 DEFAULT CHARSET=utf8mb4

-- 查询的执行计划
mysql> explain select * from city where city = ‘Ahmadnagar‘;
+----+-------------+-------+------------+------+---------------+------+---------+------+------+----------+-------------+
| id | select_type | table | partitions | type | possible_keys | key  | key_len | ref  | rows | filtered | Extra       |
+----+-------------+-------+------------+------+---------------+------+---------+------+------+----------+-------------+
|  1 | SIMPLE      | city  | NULL       | ALL  | NULL          | NULL | NULL    | NULL |  600 |    10.00 | Using where |
+----+-------------+-------+------------+------+---------------+------+---------+------+------+----------+-------------+
1 row in set, 1 warning (0.00 sec)

-- 增加city字段索引后的ddl
 CREATE TABLE `city` (
  `city_id` smallint(5) unsigned NOT NULL AUTO_INCREMENT,
  `city` varchar(50) NOT NULL,
  `country_id` smallint(5) unsigned NOT NULL,
  `last_update` timestamp NOT NULL DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP,
  PRIMARY KEY (`city_id`),
  KEY `idx_fk_country_id` (`country_id`),
  KEY `city` (`city`(7)),
  CONSTRAINT `fk_city_country` FOREIGN KEY (`country_id`) REFERENCES `country` (`country_id`) ON UPDATE CASCADE
) ENGINE=InnoDB AUTO_INCREMENT=601 DEFAULT CHARSET=utf8mb4

-- 查询的执行计划，查询已经可以走上索引了。
mysql> explain select * from city where city = ‘Ahmadnagar‘;
+----+-------------+-------+------------+------+---------------+------+---------+-------+------+----------+-------------+
| id | select_type | table | partitions | type | possible_keys | key  | key_len | ref   | rows | filtered | Extra       |
+----+-------------+-------+------------+------+---------------+------+---------+-------+------+----------+-------------+
|  1 | SIMPLE      | city  | NULL       | ref  | city          | city | 30      | const |    1 |   100.00 | Using where |
+----+-------------+-------+------------+------+---------------+------+---------+-------+------+----------+-------------+
1 row in set, 1 warning (0.00 sec)

　前缀索引是一种能让索引更小、更快的方法，但是也有缺点：

1. 无法使用前缀索引做order by 和 group by
2. 无法使用前缀索引做覆盖扫描

 3 多列索引

 

CREATE TABLE `film_actor_test` (
  `actor_id` smallint(5) unsigned NOT NULL,
  `film_id` smallint(5) unsigned NOT NULL,
  `last_update` timestamp NOT NULL DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP,
  KEY `actor_id` (`actor_id`),
  KEY `film_id` (`film_id`)
) ENGINE=InnoDB DEFAULT CHARSET=latin1


mysql> explain select film_id,actor_id from sakila.film_actor_test where actor_id = 1 or film_id = 1 G;
*************************** 1. row ***************************
           id: 1
  select_type: SIMPLE
        table: film_actor_test
   partitions: NULL
         type: index_merge
possible_keys: actor_id,film_id
          key: actor_id,film_id
      key_len: 2,2
          ref: NULL
         rows: 29
     filtered: 100.00
        Extra: Using union(actor_id,film_id); Using where
1 row in set, 1 warning (0.00 sec)