對MySQL索引知識的總結筆記。

普通索引

索引是一種數據結構，主要用于性能的提高。

比如我們有一個表t_users，有4個字段：

1234567

create table t_users ( id bigint(20) not null auto_increment, name varchar(255) not null,age bigint(20) not null, num bigint(20) not null,primary key (id)) ENGINE=InnoDB DEFAULT CHARSET=utf8;

往這個表中插入100w條數據，name字段是format-1到100w，age字段也是1-100w，num字段是個隨機的10以內的數字。然后執行sql語句進行查詢：

1234567

mysql> select * from t_users where name = 'format-500000';+--------+---------------+--------+-----+| id | name | age | num |+--------+---------------+--------+-----+| 500000 | format-500000 | 500000 | 38 |+--------+---------------+--------+-----+1 row in set (0.47 sec)

explain一下這條語句：

123456

mysql> explain select * from t_users where name = 'format-500000';+----+-------------+---------+------+---------------+------+---------+------+--------+-------------+| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |+----+-------------+---------+------+---------------+------+---------+------+--------+-------------+| 1 | SIMPLE | t_users | ALL | NULL | NULL | NULL | NULL | 996677 | Using where |+----+-------------+---------+------+---------------+------+---------+------+--------+-------------+

發現查詢了996677條數據，進行了全表的掃描。

我們給name字段加上索引：

1	mysql> create index IDX_FOR_NAME on t_users(name);

再次執行sql語句：

1234567

mysql> select * from t_users where name = 'format-500000';+--------+---------------+--------+-----+| id | name | age | num |+--------+---------------+--------+-----+| 500000 | format-500000 | 500000 | 38 |+--------+---------------+--------+-----+1 row in set (0.00 sec)

explain一下：

1234567

mysql> explain select * from t_users where name = 'format-500000';+----+-------------+---------+------+---------------+--------------+---------+-------+------+-----------------------+| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |+----+-------------+---------+------+---------------+--------------+---------+-------+------+-----------------------+| 1 | SIMPLE | t_users | ref | IDX_FOR_NAME | IDX_FOR_NAME | 767 | const | 1 | Using index condition |+----+-------------+---------+------+---------------+--------------+---------+-------+------+-----------------------+1 row in set (0.00 sec)

只查詢了1條數據，因為我們加了btree索引，可以快速定位到具體的值。

接下來我們查詢num字段：

12345678910

mysql> select * from t_users where num = 66;+--------+---------------+--------+-----+....| 965109 | format-965109 | 965109 | 66 || 965172 | format-965172 | 965172 | 66 || 965182 | format-965182 | 965182 | 66 || 965213 | format-965213 | 965213 | 66 |....+--------+---------------+--------+-----+10029 rows in set (0.30 sec)

由于num字段也是沒有加索引的，查詢的時候也進行全表的掃描，查詢耗時。接下來我們給num字段加上索引。

1	mysql> create index IDX_FOR_NUM on t_users(num);

然后進行查詢：

12345678910

mysql> select * from t_users where num = 5;+--------+---------------+--------+-----+....| 965109 | format-965109 | 965109 | 5 || 965172 | format-965172 | 965172 | 5 || 965182 | format-965182 | 965182 | 5 || 965213 | format-965213 | 965213 | 5 |....+--------+---------------+--------+-----+10029 rows in set (0.04 sec)

explain一下：

123456

mysql> explain select * from t_users where num = 5;+----+-------------+---------+------+---------------+-------------+---------+-------+-------+-------+| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |+----+-------------+---------+------+---------------+-------------+---------+-------+-------+-------+| 1 | SIMPLE | t_users | ref | IDX_FOR_NUM | IDX_FOR_NUM | 8 | const | 206712 | NULL |+----+-------------+---------+------+---------------+-------------+---------+-------+-------+-------+

雖然我們在num字段上加了索引，但是由于num值在這100w條數據中有很多重復的數據，這個時候索引對查詢速度的提高就沒有那么明顯了。

因為不論是hash類型的索引還是btree類型的索引，他們對于重復的數據的查詢并沒有提高多少。相反，由于添加了索引，導致數據寫入性能變差，而查詢性能又沒有增強多少。所以說不能盲目地添加索引。

復合索引

復合索引也叫聯合索引，表示索引建立在多個列上。

我們刪除t_users上的索引，然后創建一個復合索引。

1234	mysql> drop index IDX_FOR_NUM on t_users;mysql> drop index IDX_FOR_NAME on t_users;mysql> create index INDEX_FOR_NAME_AGE on t_users(name, age);

復合索引支持最左原則，也就是說INDEX_FOR_NAME_AGE索引支持name字段的查找，支持name，age字段的查找，但是不支持age，name字段的查找以及age字段的查找。

1234567891011121314151617181920

mysql> explain select * from t_users where name = 'format-100000';+----+-------------+---------+------+--------------------+--------------------+---------+-------+------+-----------------------+| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |+----+-------------+---------+------+--------------------+--------------------+---------+-------+------+-----------------------+| 1 | SIMPLE | t_users | ref | INDEX_FOR_NAME_AGE | INDEX_FOR_NAME_AGE | 767 | const | 1 | Using index condition |+----+-------------+---------+------+--------------------+--------------------+---------+-------+------+-----------------------+mysql> explain select * from t_users where name = 'format-100000' and age = 100000;+----+-------------+---------+------+--------------------+--------------------+---------+-------------+------+-----------------------+| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |+----+-------------+---------+------+--------------------+--------------------+---------+-------------+------+-----------------------+| 1 | SIMPLE | t_users | ref | INDEX_FOR_NAME_AGE | INDEX_FOR_NAME_AGE | 775 | const,const | 1 | Using index condition |+----+-------------+---------+------+--------------------+--------------------+---------+-------------+------+-----------------------+mysql> explain select * from t_users where age = 100000;+----+-------------+---------+------+---------------+------+---------+------+--------+-------------+| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |+----+-------------+---------+------+---------------+------+---------+------+--------+-------------+| 1 | SIMPLE | t_users | ALL | NULL | NULL | NULL | NULL | 996677 | Using where |+----+-------------+---------+------+---------------+------+---------+------+--------+-------------+

使用age，name字段查找：

123456

mysql> explain select * from t_users where age = 100000 and name = 'format-100000';+----+-------------+---------+------+--------------------+--------------------+---------+-------------+------+-----------------------+| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |+----+-------------+---------+------+--------------------+--------------------+---------+-------------+------+-----------------------+| 1 | SIMPLE | t_users | ref | INDEX_FOR_NAME_AGE | INDEX_FOR_NAME_AGE | 775 | const,const | 1 | Using index condition |+----+-------------+---------+------+--------------------+--------------------+---------+-------------+------+-----------------------+

我們發現使用age，name字段的查找使用了復合索引，這是因為MySQL內部有個查詢優化器幫我們進行了優化。

索引的生效

索引創建之后，查詢的過程并不一定會使用索引。整理如下(t_users表中name和num字段都有單獨的索引)：

1.當使用索引查詢比全表掃描更慢。比如下面這句sql中num字段的值分布在1-10之間

123456

mysql> explain select * from t_users where num > 1 and num < 8;+----+-------------+----------+------+---------------+------+---------+------+--------+-------------+| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |+----+-------------+----------+------+---------------+------+---------+------+--------+-------------+| 1 | SIMPLE | t_users | ALL | IDX_ON_NUM | NULL | NULL | NULL | 996504 | Using where |+----+-------------+----------+------+---------------+------+---------+------+--------+-------------+

2.使用or進行查詢，并且or左右兩邊的列有不存在索引的列

123456

mysql> explain select * from t_users where name = 'format-4000' or age = 50;+----+-------------+---------+------+---------------+------+---------+------+--------+-------------+| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |+----+-------------+---------+------+---------------+------+---------+------+--------+-------------+| 1 | SIMPLE | t_users | ALL | IDX_FOR_NAME | NULL | NULL | NULL | 996677 | Using where |+----+-------------+---------+------+---------------+------+---------+------+--------+-------------+

or兩邊的列都有索引：

123456

mysql> explain select * from t_users where name = 'format-4000' or num = 50;+----+-------------+----------+-------------+---------------------------------+---------------------------------+---------+------+------+-----------------------------------------------------------+| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |+----+-------------+----------+-------------+---------------------------------+---------------------------------+---------+------+------+-----------------------------------------------------------+| 1 | SIMPLE | t_users | index_merge | IDX_ON_NUM,IDX_FOR_NAME | IDX_FOR_NAME,IDX_ON_NUM | 767,8 | NULL | 2 | Using union(IDX_FOR_NAME,IDX_ON_NUM); Using where |+----+-------------+----------+-------------+---------------------------------+---------------------------------+---------+------+------+-----------------------------------------------------------+

3.使用like，并以 % 開頭

12345678910111213

mysql> explain select * from t_users where name like "%format-200";+----+-------------+---------+------+---------------+------+---------+------+--------+-------------+| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |+----+-------------+---------+------+---------------+------+---------+------+--------+-------------+| 1 | SIMPLE | t_users | ALL | NULL | NULL | NULL | NULL | 996677 | Using where |+----+-------------+---------+------+---------------+------+---------+------+--------+-------------+mysql> explain select * from t_users where name like "%format-200%";+----+-------------+---------+------+---------------+------+---------+------+--------+-------------+| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |+----+-------------+---------+------+---------------+------+---------+------+--------+-------------+| 1 | SIMPLE | t_users | ALL | NULL | NULL | NULL | NULL | 996677 | Using where |+----+-------------+---------+------+---------------+------+---------+------+--------+-------------+

以 % 結尾的查詢會使用索引：

123456

mysql> explain select * from t_users where name like "format-200%";+----+-------------+---------+-------+---------------+--------------+---------+------+------+-----------------------+| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |+----+-------------+---------+-------+---------------+--------------+---------+------+------+-----------------------+| 1 | SIMPLE | t_users | range | IDX_FOR_NAME | IDX_FOR_NAME | 767 | NULL | 1110 | Using index condition |+----+-------------+---------+-------+---------------+--------------+---------+------+------+-----------------------+

4.復合索引

不是復合索引的最左字段(t_users表有(name,age)復合索引)。

123456

mysql> explain select * from t_users where age = 100000;+----+-------------+---------+------+---------------+------+---------+------+--------+-------------+| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |+----+-------------+---------+------+---------------+------+---------+------+--------+-------------+| 1 | SIMPLE | t_users | ALL | NULL | NULL | NULL | NULL | 996677 | Using where |+----+-------------+---------+------+---------------+------+---------+------+--------+-------------+

5.對varchar類型的字段進行查詢的時候，沒有加上引號

12345678910111213

mysql> explain select * from t_users where name = 111;+----+-------------+---------+------+---------------+------+---------+------+--------+-------------+| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |+----+-------------+---------+------+---------------+------+---------+------+--------+-------------+| 1 | SIMPLE | t_users | ALL | IDX_FOR_NAME | NULL | NULL | NULL | 996677 | Using where |+----+-------------+---------+------+---------------+------+---------+------+--------+-------------+mysql> explain select * from t_users where name = "111";+----+-------------+---------+------+---------------+--------------+---------+-------+------+-----------------------+| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |+----+-------------+---------+------+---------------+--------------+---------+-------+------+-----------------------+| 1 | SIMPLE | t_users | ref | IDX_FOR_NAME | IDX_FOR_NAME | 767 | const | 1 | Using index condition |+----+-------------+---------+------+---------------+--------------+---------+-------+------+-----------------------+

hash索引和btree索引的區別

hash索引不能使用范圍查詢，只能使用一些比如 “=”, “<>”, “in”查詢。因為hash索引會計算索引列的hash值，計算出后的hash值經過了hash算法與原先的值完全不一樣，只能進行等值的過濾，不能基于范圍的過濾

hash索引遇到大量hash值相同的情況下，性能比btree要差

hash索引并不一定一次可以定位到數據。因為基于索引列計算出的hash值會有重復，重復的話需要掃描hash表進行比較

由于hash索引中存放的是經過hash計算之后的hash值，而且hash值的大小關系并不一定和hash運算前的鍵值完全一樣，所以數據庫無法利用索引的數據來避免任何排序運算

對于組合索引，hash索引在計算hash值的時候是組合索引鍵合并后再一起計算hash值，而不是單獨計算hash值，所以通過組合索引的前面一個或幾個索引鍵進行查詢的時候，hash索引也無法被利用

InnoDB和MyISAM引擎不支持hash索引

本文作者：Format
原文鏈接： http://fangjian0423.github.io/2017/07/05/mysql-index-summary/

總結

以上是生活随笔為你收集整理的MySQL索引知识总结的全部內容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網站內容還不錯，歡迎將生活随笔推薦給好友。