MySQL 通过多个示例学习索引

　　最近在准备面试，关于索引这一块，发现很多以前忽略的点，这里好好整理一下

　　

首先为什么要建立索引

　　一本书，有章、节、段、行这种单位。

　　如果现在需要找一个内容：第9章>第2节>第3段>第4行>第5个字。

　　如果没有索引（目录），那么，就必须从第一页开始，按照这个顺序：第1章>第1节>第1段>第1行>第1个字开始，顺序查找，也就是说，如果运气足够坏，那么，找到目标的时候，已经将整本书都差不多扫了一遍。

　　但是现在，如果建立了索引（目录），只是建立了 “章” 的索引：每一章的第一页，那么，我们开始查找的时候，并不是从第1章开始，我们可以根据目录，查找到第9章的页数，然后直接翻到第9章的那章内容，初始就是第9章>第1节>第1段>第1行>第1个字，然后仍旧开始顺序查找，那么这个时候，虽然仍然有点慢，但是已经快了很多很多。

　　现在，如果建立（章-节）的索引，记录每一章的每一个小节的第1页，那么我们开始查找的时候，可以通过读一下索引，直接找到第9章的第2小节，然后从这1小节的第1段的第1行的第1个字开始查找。仍旧顺序查找。

　　很好，你会想，可以建立（章-节-段）的索引，那么，一开始就可以直接定位到第9章>第2小节>第3段，然后从这段的第1行的第1个字开始顺序查找。

　　那么，有没有必要建立（章-节-段-行）的索引呢？可以根据情况而定。

索引类型

　　有很多中索引，比如：

　　1、普通索引：最基本的索引，没有任何约束限制

　　2、唯一索引：与主键索引类似，但是不允许出现相同的值（唯一性约束）。

　　3、主键索引：特殊的唯一索引，不允许有空值。

　　4、复盖索引（组合索引）：将多个列组合在一起创建索引，可以覆盖多个列，比如“章-节-段”建立索引

　　5、外键索引：只有Innodb支持，保证数据一致性、完整性，实现级联操作。

　　6、全文索引：只有MyISAM引擎支持，并且只支持对英文进行全文检索。

建立索引

　　假设有一个user表（用户表）

mysql> desc user;
+-------+----------+------+-----+---------+----------------+
| Field | Type     | Null | Key | Default | Extra          |
+-------+----------+------+-----+---------+----------------+
| uid   | int(11)  | NO   | PRI | NULL    | auto_increment |
| uname | char(20) | NO   | MUL | NULL    |                |
| addr  | char(20) | NO   |     | NULL    |                |
| token | char(20) | NO   |     | NULL    |                |
+-------+----------+------+-----+---------+----------------+
5 rows in set (0.05 sec)

　　uid因为是主键，所以已经默认创建了主键索引　　

　　为uname字段建立一个唯一索引，为addr字段建立普通索引；

　　token字段不建立索引

mysql> create unique index uname_index on user(uname);
mysql> create index addr_index on user(addr);

　　

查看索引

mysql> show index from user;
*************************** 1. row ***************************
        Table: user
   Non_unique: 0
     Key_name: PRIMARY
 Seq_in_index: 1
  Column_name: uid
    Collation: A
  Cardinality: 0
     Sub_part: NULL
       Packed: NULL
         Null:
   Index_type: BTREE
      Comment:
Index_comment:
*************************** 2. row ***************************
        Table: user
   Non_unique: 0
     Key_name: uname_index
 Seq_in_index: 1
  Column_name: uname
    Collation: A
  Cardinality: 0
     Sub_part: NULL
       Packed: NULL
         Null:
   Index_type: BTREE
      Comment:
Index_comment:
*************************** 3. row ***************************
        Table: user
   Non_unique: 1
     Key_name: addr_index
 Seq_in_index: 1
  Column_name: addr
    Collation: A
  Cardinality: NULL
     Sub_part: NULL
       Packed: NULL
         Null:
   Index_type: BTREE
      Comment:
Index_comment:

　　

插入数据

　　使用pdo循环插入10000条记录，各个字段除了uid自增之外，都是随机值。

<?php
    $pdo = new PDO("mysql:host=127.0.0.1;dbname=MyDb;charset=utf8", "root", "root");

    $stmt = $pdo->prepare("insert into user (uname, addr, token) values (?, ?, ?)");

    $str = "abcdefghijklmnopqrstuvwxzy";

    for ($j = 0; $j < 10000; $j++) {
        $name = "";
        $addr = "";
        $token = "";
        for ($i = 0; $i < 7; $i++) {
            $name .= $str[mt_rand(0, 25)];
            $addr .= $str[mt_rand(0, 25)];
            $token .= $str[mt_rand(0, 25)];
        }
        $stmt->execute(array($name, $addr, $token));
    }

?>

　　

是否会使用索引？

　　示例1：

mysql> explain select count(*) from user;
+----+-------------+-------+------+---------------+------+---------+------+------+------------------------------+
| id | select_type | table | type | possible_keys | key  | key_len | ref  | rows | Extra                        |
+----+-------------+-------+------+---------------+------+---------+------+------+------------------------------+
|  1 | SIMPLE      | NULL  | NULL | NULL          | NULL | NULL    | NULL | NULL | Select tables optimized away |
+----+-------------+-------+------+---------------+------+---------+------+------+------------------------------+
1 row in set (0.00 sec)

　　未使用索引，因为没有使用where子句，索引是在查找的时候（有where的时候才使用索引）。

　　示例2：

mysql> explain select * from user where uname='outalwa';
+----+-------------+-------+-------+---------------+-------------+---------+-------+------+-------+
| id | select_type | table | type  | possible_keys | key         | key_len | ref   | rows | Extra |
+----+-------------+-------+-------+---------------+-------------+---------+-------+------+-------+
|  1 | SIMPLE      | user  | const | uname_index   | uname_index | 60      | const |    1 |       |
+----+-------------+-------+-------+---------------+-------------+---------+-------+------+-------+
1 row in set (0.00 sec)

mysql> explain select * from user where uname!='outalwa';
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+
| id | select_type | table | type | possible_keys | key  | key_len | ref  | rows  | Extra       |
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+
|  1 | SIMPLE      | user  | ALL  | uname_index   | NULL | NULL    | NULL | 10000 | Using where |
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+
1 row in set (0.00 sec)

mysql> explain select * from user where uname<>'outalwa';
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+
| id | select_type | table | type | possible_keys | key  | key_len | ref  | rows  | Extra       |
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+
|  1 | SIMPLE      | user  | ALL  | uname_index   | NULL | NULL    | NULL | 10000 | Using where |
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+
1 row in set (0.00 sec)

　　对于一个建了索引的列来说，如果使用=、!=、<>来判断字段值是否与一个值相等的时候，只有使用=的时候会使用索引，而!=和<>不会使用索引。

　　示例3：

mysql> explain select * from user where uid='9999' and uname='muqehaq';                              
+----+-------------+-------+-------+---------------------+---------+---------+-------+------+-------+
| id | select_type | table | type  | possible_keys       | key     | key_len | ref   | rows | Extra |
+----+-------------+-------+-------+---------------------+---------+---------+-------+------+-------+
|  1 | SIMPLE      | user  | const | PRIMARY,uname_index | PRIMARY | 4       | const |    1 |       |
+----+-------------+-------+-------+---------------------+---------+---------+-------+------+-------+
1 row in set (0.06 sec)                                                                              

　　因为uid建立了主键索引，uname建立了唯一索引，使用and表示（并且）的关系时，会使用各自的索引。只扫描了1行数据就找到了

　　

　　示例4：

mysql> explain select * from user where uid='9999' or uname='muqehaq';
+----+-------------+-------+-------------+---------------------+---------------------+---------+------+------+-----------------------------------------------+
| id | select_type | table | type        | possible_keys       | key                 | key_len | ref  | rows | Extra                                         |
+----+-------------+-------+-------------+---------------------+---------------------+---------+------+------+-----------------------------------------------+
|  1 | SIMPLE      | user  | index_merge | PRIMARY,uname_index | PRIMARY,uname_index | 4,60    | NULL |    2 | Using union(PRIMARY,uname_index); Using where |
+----+-------------+-------+-------------+---------------------+---------------------+---------+------+------+-----------------------------------------------+
1 row in set (0.00 sec)

　　因为uid建立了主键索引，uname建立了唯一索引，使用or表示（或）的关系时，使用了索引。

　　示例5

mysql> explain select * from user where uid='9999' and token='dakghzz';
+----+-------------+-------+-------+---------------+---------+---------+-------+------+-------+
| id | select_type | table | type  | possible_keys | key     | key_len | ref   | rows | Extra |
+----+-------------+-------+-------+---------------+---------+---------+-------+------+-------+
|  1 | SIMPLE      | user  | const | PRIMARY       | PRIMARY | 4       | const |    1 |       |
+----+-------------+-------+-------+---------------+---------+---------+-------+------+-------+

　　uid有主键索引，token没有建立索引，使用and来连接条件，可以使用索引。

　　如果条件是token='dakghzz' and uid=9999 ，同样会使用索引。

　　示例6

mysql> explain select * from user where uid='9999' or token='dakghzz';
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+
| id | select_type | table | type | possible_keys | key  | key_len | ref  | rows  | Extra       |
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+
|  1 | SIMPLE      | user  | ALL  | PRIMARY       | NULL | NULL    | NULL | 10000 | Using where |
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+
1 row in set (0.00 sec)

　　uid有主键索引，但是token没有建立索引，使用or来连接的时候，可以看到，查询并没有使用索引。

　　示例7

mysql> explain select count(*) from user where uid < 1000;
+----+-------------+-------+-------+---------------+---------+---------+------+------+--------------------------+
| id | select_type | table | type  | possible_keys | key     | key_len | ref  | rows | Extra                    |
+----+-------------+-------+-------+---------------+---------+---------+------+------+--------------------------+
|  1 | SIMPLE      | user  | range | PRIMARY       | PRIMARY | 4       | NULL | 1401 | Using where; Using index |
+----+-------------+-------+-------+---------------+---------+---------+------+------+--------------------------+
1 row in set (0.00 sec)

mysql> explain select count(*) from user where uid between 10 and 1000;
+----+-------------+-------+-------+---------------+---------+---------+------+------+--------------------------+
| id | select_type | table | type  | possible_keys | key     | key_len | ref  | rows | Extra                    |
+----+-------------+-------+-------+---------------+---------+---------+------+------+--------------------------+
|  1 | SIMPLE      | user  | range | PRIMARY       | PRIMARY | 4       | NULL | 1388 | Using where; Using index |
+----+-------------+-------+-------+---------------+---------+---------+------+------+--------------------------+
1 row in set (0.00 sec)

mysql> explain select count(*) from user where uid >10 and uid < 1000;
+----+-------------+-------+-------+---------------+---------+---------+------+------+--------------------------+
| id | select_type | table | type  | possible_keys | key     | key_len | ref  | rows | Extra                    |
+----+-------------+-------+-------+---------------+---------+---------+------+------+--------------------------+
|  1 | SIMPLE      | user  | range | PRIMARY       | PRIMARY | 4       | NULL | 1386 | Using where; Using index |
+----+-------------+-------+-------+---------------+---------+---------+------+------+--------------------------+
1 row in set (0.00 sec)

　　如果存储引擎是innodb的话，对于主键索引列，使用 大于、小于、大于等于、小于等于、between and 作为筛选条件，都会使用索引。

　　如果存储引擎是MyISAM的话，即使是索引列，使用是大于、小于、大于等于、小于等于、between and作为筛选条件，都不会使用索引。

　　无论是哪种存储引擎，即使某一列创建了索引，如果使用大于、小于、大于等于、小于等于、between and作为筛选条件，都不会使用索引。

　　示例8

mysql> explain select * from user where uname=null;
+----+-------------+-------+------+---------------+------+---------+------+------+-----------------------------------------------------+
| id | select_type | table | type | possible_keys | key  | key_len | ref  | rows | Extra                                               |
+----+-------------+-------+------+---------------+------+---------+------+------+-----------------------------------------------------+
|  1 | SIMPLE      | NULL  | NULL | NULL          | NULL | NULL    | NULL | NULL | Impossible WHERE noticed after reading const tables |
+----+-------------+-------+------+---------------+------+---------+------+------+-----------------------------------------------------+
1 row in set (0.00 sec)

mysql> explain select * from user where uname!=null;
+----+-------------+-------+------+---------------+------+---------+------+------+-----------------------------------------------------+
| id | select_type | table | type | possible_keys | key  | key_len | ref  | rows | Extra                                               |
+----+-------------+-------+------+---------------+------+---------+------+------+-----------------------------------------------------+
|  1 | SIMPLE      | NULL  | NULL | NULL          | NULL | NULL    | NULL | NULL | Impossible WHERE noticed after reading const tables |
+----+-------------+-------+------+---------------+------+---------+------+------+-----------------------------------------------------+
1 row in set (0.00 sec)

mysql> explain select * from user where uname is null;
+----+-------------+-------+------+---------------+------+---------+------+------+------------------+
| id | select_type | table | type | possible_keys | key  | key_len | ref  | rows | Extra            |
+----+-------------+-------+------+---------------+------+---------+------+------+------------------+
|  1 | SIMPLE      | NULL  | NULL | NULL          | NULL | NULL    | NULL | NULL | Impossible WHERE |
+----+-------------+-------+------+---------------+------+---------+------+------+------------------+
1 row in set (0.00 sec)

mysql> explain select * from user where uname is not null;
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+
| id | select_type | table | type | possible_keys | key  | key_len | ref  | rows  | Extra       |
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+
|  1 | SIMPLE      | user  | ALL  | uname_index   | NULL | NULL    | NULL | 10000 | Using where |
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+
1 row in set (0.00 sec)

　　　　对于一个建立了索引的列来说，判断字段是否为0时，使用=，!=，is null，is not null这几种方式都不会使用索引。

　　示例9

mysql> explain select * from user where uname like '%wolwgrc';
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+
| id | select_type | table | type | possible_keys | key  | key_len | ref  | rows  | Extra       |
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+
|  1 | SIMPLE      | user  | ALL  | NULL          | NULL | NULL    | NULL | 10000 | Using where |
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+
1 row in set (0.00 sec)

mysql> explain select * from user where uname like '%wolwgrc%';
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+
| id | select_type | table | type | possible_keys | key  | key_len | ref  | rows  | Extra       |
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+
|  1 | SIMPLE      | user  | ALL  | NULL          | NULL | NULL    | NULL | 10000 | Using where |
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+

mysql> explain select * from user where uname like 'wolwgrc';
+----+-------------+-------+-------+---------------+-------------+---------+------+------+-------------+
| id | select_type | table | type  | possible_keys | key         | key_len | ref  | rows | Extra       |
+----+-------------+-------+-------+---------------+-------------+---------+------+------+-------------+
|  1 | SIMPLE      | user  | range | uname_index   | uname_index | 60      | NULL |    1 | Using where |
+----+-------------+-------+-------+---------------+-------------+---------+------+------+-------------+
1 row in set (0.00 sec)

mysql> explain select * from user where uname like 'wolwgrc%';
+----+-------------+-------+-------+---------------+-------------+---------+------+------+-------------+
| id | select_type | table | type  | possible_keys | key         | key_len | ref  | rows | Extra       |
+----+-------------+-------+-------+---------------+-------------+---------+------+------+-------------+
|  1 | SIMPLE      | user  | range | uname_index   | uname_index | 60      | NULL |    1 | Using where |
+----+-------------+-------+-------+---------------+-------------+---------+------+------+-------------+
1 row in set (0.00 sec)

　　对一个已经建立了索引的列使用模式匹配的时候，%keyword、%keyword%不会使用索引，而keyword和keyword%会使用索引。

普通索引的注意事项

　　1、 对于一个没有where的查询，不会使用索引，所以，在创建索引的时候，应该以where子句的字段为准，而不是select后面的字段为准。

　　2、对于一个建了索引的列来说，如果使用=、!=、<>来判断字段值是否与一个值相等的时候，只有使用=的时候会使用索引，而!=和<>不会使用索引。

　　3、where包含多个字段，如果每个字段都建了索引，那么使用and和or连接的时候，查询会使用索引。

　　4、where包含多个字段，如果一部分字段建立了索引，而有一部分字段没有建立索引，那么使用and连接的时候，查询的时候，建立了索引的字段会使用索引，而没有建立索引的字段不会使用索引。

　　5、where包含多个字段，如果一部分字段建立了索引，而有一部分字段没有建立索引，那么使用or连接的时候，查询的时候，不会使用索引。

　　6、如果存储引擎是innodb的话，对于主键索引列，使用 大于、小于、大于等于、小于等于、between and 作为筛选条件，都会使用索引。

　　7、如果存储引擎是MyISAM的话，即使是索引列，使用 大于、小于、大于等于、小于等于、between and作为筛选条件，都不会使用索引。

　　8、无论是哪种存储引擎，即使某一列创建了索引，如果使用大于、小于、大于等于、小于等于、between and作为筛选条件，都不会使用索引。

　　9、对于一个建立了索引的列来说，判断字段是否为0时，使用=，!=，is null，is not nulll这几种方式都不会使用索引。

　　10、对一个已经建立了索引的列使用模式匹配的时候，%keyword、%keyword%不会使用索引，而keyword和keyword%会使用索引。

复合索引

　　复合索引有前缀原则，后面示例会解释。

　　删除之前为uname字段和addr字段创建的索引。

　　创建一个复合索引，包含uname、addr、token，命令如下：

mysql> drop index uname_index on user;                                                                                                                               
Query OK, 10000 rows affected (0.27 sec)                                                                                                                             
Records: 10000  Duplicates: 0  Warnings: 0                                                                                                                           
                                                                                                                                                                     
mysql> drop index addr_index on user;                                                                                                                                
Query OK, 10000 rows affected (0.24 sec)                                                                                                                             
Records: 10000  Duplicates: 0  Warnings: 0                                                                                                                           
                                                                                                                                                                     
mysql> create index three_fields_index on user(uname, addr, token);                                                                                                  
Query OK, 10000 rows affected (0.25 sec)                                                                                                                             
Records: 10000  Duplicates: 0  Warnings: 0                                                                                                                           
                                                                                                                                                                     
mysql> show index from user;                                                                                                                                         
+-------+------------+--------------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+ 
| Table | Non_unique | Key_name           | Seq_in_index | Column_name | Collation | Cardinality | Sub_part | Packed | Null | Index_type | Comment | Index_comment | 
+-------+------------+--------------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+ 
| user  |          0 | PRIMARY            |            1 | uid         | A         |       10000 |     NULL | NULL   |      | BTREE      |         |               | 
| user  |          1 | three_fields_index |            1 | uname       | A         |       10000 |     NULL | NULL   |      | BTREE      |         |               | 
| user  |          1 | three_fields_index |            2 | addr        | A         |       10000 |     NULL | NULL   |      | BTREE      |         |               | 
| user  |          1 | three_fields_index |            3 | token       | A         |       10000 |     NULL | NULL   |      | BTREE      |         |               | 
+-------+------------+--------------------+--------------+-------------+-----------+-------------+----------+--------+------+------------+---------+---------------+ 

　　可以从上面索引查询结果中看出，uname、addr、token都包含在一个索引中（Key_name相同）。

覆盖索引的使用示例

　　下面的实例，简化一下，创建的索引是key(uname, addr, token)，那么对应key(A, B, C)

　　示例1

mysql> explain select * from user where uname='muqehaq' and addr='hgommlp' and token='dakghzz';
+----+-------------+-------+------+--------------------+--------------------+---------+-------------------+------+-------------+
| id | select_type | table | type | possible_keys      | key                | key_len | ref               | rows | Extra       |
+----+-------------+-------+------+--------------------+--------------------+---------+-------------------+------+-------------+
|  1 | SIMPLE      | user  | ref  | three_fields_index | three_fields_index | 180     | const,const,const |    1 | Using where |
+----+-------------+-------+------+--------------------+--------------------+---------+-------------------+------+-------------+
1 row in set (0.00 sec)

mysql> explain select * from user where addr='hgommlp' and token='dakghzz' and uname='muqehaq';
+----+-------------+-------+------+--------------------+--------------------+---------+-------------------+------+-------------+
| id | select_type | table | type | possible_keys      | key                | key_len | ref               | rows | Extra       |
+----+-------------+-------+------+--------------------+--------------------+---------+-------------------+------+-------------+
|  1 | SIMPLE      | user  | ref  | three_fields_index | three_fields_index | 180     | const,const,const |    1 | Using where |
+----+-------------+-------+------+--------------------+--------------------+---------+-------------------+------+-------------+
1 row in set (0.00 sec)

　　因为建立了uname-addr-token的联合索引，在where中，三个字段都出现了（和建立索引时使用的字段列相同时），此时，和每个字段的顺序无关，都会使用索引。

　　结论：where A='xx' and B='xx' and 或者where C='xx'  and B='xx' and A='xx'会使用索引。

　　示例2

mysql> explain select * from user where uname='muqehaq' and addr='hgommlp';
+----+-------------+-------+------+--------------------+--------------------+---------+-------------+------+-------------+
| id | select_type | table | type | possible_keys      | key                | key_len | ref         | rows | Extra       |
+----+-------------+-------+------+--------------------+--------------------+---------+-------------+------+-------------+
|  1 | SIMPLE      | user  | ref  | three_fields_index | three_fields_index | 120     | const,const |    1 | Using where |
+----+-------------+-------+------+--------------------+--------------------+---------+-------------+------+-------------+
1 row in set (0.00 sec)

mysql> explain select * from user where addr='hgommlp' and uname='muqehaq' ;
+----+-------------+-------+------+--------------------+--------------------+---------+-------------+------+-------------+
| id | select_type | table | type | possible_keys      | key                | key_len | ref         | rows | Extra       |
+----+-------------+-------+------+--------------------+--------------------+---------+-------------+------+-------------+
|  1 | SIMPLE      | user  | ref  | three_fields_index | three_fields_index | 120     | const,const |    1 | Using where |
+----+-------------+-------+------+--------------------+--------------------+---------+-------------+------+-------------+
1 row in set (0.00 sec)

　　示例：where A='xx' and B='xx' 或者 where B='xx' and A='xx'

　　结论：会使用索引，使用复合索引中的前两个字段的索引。

　　示例3：

mysql> explain select * from user where uname='muqehaq' and token='dakghzz';
+----+-------------+-------+------+--------------------+--------------------+---------+-------+------+-------------+
| id | select_type | table | type | possible_keys      | key                | key_len | ref   | rows | Extra       |
+----+-------------+-------+------+--------------------+--------------------+---------+-------+------+-------------+
|  1 | SIMPLE      | user  | ref  | three_fields_index | three_fields_index | 60      | const |    1 | Using where |
+----+-------------+-------+------+--------------------+--------------------+---------+-------+------+-------------+
1 row in set (0.00 sec)

mysql> explain select * from user where token='dakghzz' and uname='muqehaq';
+----+-------------+-------+------+--------------------+--------------------+---------+-------+------+-------------+
| id | select_type | table | type | possible_keys      | key                | key_len | ref   | rows | Extra       |
+----+-------------+-------+------+--------------------+--------------------+---------+-------+------+-------------+
|  1 | SIMPLE      | user  | ref  | three_fields_index | three_fields_index | 60      | const |    1 | Using where |
+----+-------------+-------+------+--------------------+--------------------+---------+-------+------+-------------+
1 row in set (0.00 sec)

　　示例：where A='xxx' and C='xx'   或者 where C='xx' and A='xxx'

　　结论：查询A字段的时候，会使用复合索引中的A那一部分，但是查询C字段的时候，并不会使用复合索引。

　　原因：因为没有使用复合索引中的B字段。

　　示例4

mysql> explain select * from user where addr='hgommlp' and token='dakghzz';
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+
| id | select_type | table | type | possible_keys | key  | key_len | ref  | rows  | Extra       |
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+
|  1 | SIMPLE      | user  | ALL  | NULL          | NULL | NULL    | NULL | 10000 | Using where |
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+
1 row in set (0.00 sec)

　　示例：where B='xxx' and C='xxx' 

　　结论：不会使用索引。

　　原因：没有使用复合索引中B前面的A字段索引。

　　示例5

mysql> explain select * from user where uname='muqehaq';                                                              
+----+-------------+-------+------+--------------------+--------------------+---------+-------+------+-------------+  
| id | select_type | table | type | possible_keys      | key                | key_len | ref   | rows | Extra       |  
+----+-------------+-------+------+--------------------+--------------------+---------+-------+------+-------------+  
|  1 | SIMPLE      | user  | ref  | three_fields_index | three_fields_index | 60      | const |    1 | Using where |  
+----+-------------+-------+------+--------------------+--------------------+---------+-------+------+-------------+  
1 row in set (0.00 sec)                                                                                               
                                                                                                 
mysql> explain select * from user where addr='hgommlp';                                                               
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+                     
| id | select_type | table | type | possible_keys | key  | key_len | ref  | rows  | Extra       |                     
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+                     
|  1 | SIMPLE      | user  | ALL  | NULL          | NULL | NULL    | NULL | 10000 | Using where |                     
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+                     
1 row in set (0.00 sec)                                                                                               
                                                                                                                      
mysql> explain select * from user where token='dakghzz';                                                              
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+                     
| id | select_type | table | type | possible_keys | key  | key_len | ref  | rows  | Extra       |                     
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+                     
|  1 | SIMPLE      | user  | ALL  | NULL          | NULL | NULL    | NULL | 10000 | Using where |                     
+----+-------------+-------+------+---------------+------+---------+------+-------+-------------+                     

　　示例：where A='xxx'    结论：会使用索引。

　　示例：where B='xxx'    结论：不会使用索引。

　　示例：where C='xxx'    结论：不会使用索引。

复合索引的使用总结

　　假设创建了key（a，b，c）复合索引，那么：

　　where a = 1 and b = 2  and  c = 3;       会使用索引

　　where a = 1 and c = 3 and b = 2;   会使用索引

　　where a = 1 and b = 2 ；      会使用索引

　　where b = 2 and a = 1;      注意，这也是会使用复合索引的ab

　　where a = 1;会使用索引

　　where a = 1 and c = 3;   不会使用索引，因为c前面的b没有使用

　　where b = 2 and c = 3;   不会使用索引，因为b前面的a没有使用