MySQL8.0新特性--函数索引

很多开发人员在使用MySQL时经常会在部分列上进行函数计算等，导致无法走索引，在数据量大的时候，查询效率低下。针对此种情况本文从MySQL5.7 及MySQL8.0中分别进行不同方式的优化。

1、 MySQL5.7

MySQL5.7版本中不支持函数索引，因此 遇到函数索引的时候需要进行修改，否则即使查询的字段上有索引，执行时也无法使用索引而进行全表扫描，数据量大的表查询时间会比较长。具体案例如下：

1.1 创建测试表及数据

mysql> use testdb;
Database changed
mysql> create table   tb_function(id int primary key auto_increment,name  varchar(100),create_time datetime);
Query OK, 0 rows affected (0.01 sec)


mysql> insert into  tb_function(name,creatE_time) values('anniuadaOAIFAPUHIA','2020-07-01 12:00:00');
Query OK, 1 row affected (0.02 sec)


mysql> insert into  tb_function(name,creatE_time) values('CWQSsar3qcssg','2020-07-01 15:00:00');
Query OK, 1 row affected (0.01 sec)


mysql> insert into  tb_function(name,creatE_time) values('vxfqrt2adafz','2020-07-01 21:30:00');
Query OK, 1 row affected (0.01 sec)


mysql> insert into  tb_function(name,creatE_time) values('etxzwrwbdhegqgaheqhag','2020-07-02 01:30:00');
Query OK, 1 row affected (0.01 sec)


mysql> insert into  tb_function(name,creatE_time) values('awrs433fsgvsfwtwg','2020-07-02 03:30:00');
Query OK, 1 row affected (0.00 sec)


mysql> insert into  tb_function(name,creatE_time) values('awrs433fsgvsfwtwg','2020-07-02 07:32:00');
Query OK, 1 row affected (0.00 sec)


mysql> insert into  tb_function(name,creatE_time) values('awrs433fsgvsfwtwg','2020-07-02 10:32:00');
Query OK, 1 row affected (0.00 sec)


mysql> insert into  tb_function(name,creatE_time) values('tuilklmdadq','2020-07-02 15:32:00');
Query OK, 1 row affected (0.00 sec)


mysql> insert into  tb_function(name,creatE_time) values('wesv2wqdshehq','2020-07-02 20:32:00');
Query OK, 1 row affected (0.00 sec)


mysql> insert into  tb_function(name,creatE_time) values('89yoijnlkwr1','2020-07-03 02:56:00');
Query OK, 1 row affected (0.00 sec)


mysql> insert into  tb_function(name,creatE_time) values('olj;nsaaq','2020-07-03 08:41:00');
Query OK, 1 row affected (0.01 sec)


mysql> insert into  tb_function(name,creatE_time) values('ygo;jkdsaq','2020-07-03 16:20:00');
Query OK, 1 row affected (0.01 sec)


mysql> select  * from tb_function;
+----+-----------------------+---------------------+
| id | name                  | create_time         |
+----+-----------------------+---------------------+
|  1 | anniuadaOAIFAPUHIA    | 2020-07-01 12:00:00 |
|  2 | CWQSsar3qcssg         | 2020-07-01 15:00:00 |
|  3 | vxfqrt2adafz          | 2020-07-01 21:30:00 |
|  4 | etxzwrwbdhegqgaheqhag | 2020-07-02 01:30:00 |
|  5 | awrs433fsgvsfwtwg     | 2020-07-02 03:30:00 |
|  6 | awrs433fsgvsfwtwg     | 2020-07-02 07:32:00 |
|  7 | awrs433fsgvsfwtwg     | 2020-07-02 10:32:00 |
|  8 | tuilklmdadq           | 2020-07-02 15:32:00 |
|  9 | wesv2wqdshehq         | 2020-07-02 20:32:00 |
| 10 | 89yoijnlkwr1          | 2020-07-03 02:56:00 |
| 11 | olj;nsaaq             | 2020-07-03 08:41:00 |
| 12 | ygo;jkdsaq            | 2020-07-03 16:20:00 |
+----+-----------------------+---------------------+
12 rows in set (0.00 sec)

1.2  创建索引

在create_time字段上创建索引

mysql> alter  table tb_function add key idx_create_time(create_time);
Query OK, 0 rows affected (0.13 sec)
Records: 0  Duplicates: 0  Warnings: 0

1.3  按时间查询

查询创建时间是2020-07-01那天的所有记录

mysql> select  *  from  tb_function  where   date(create_time)='2020-07-01';
+----+--------------------+---------------------+
| id | name               | create_time         |
+----+--------------------+---------------------+
|  1 | anniuadaOAIFAPUHIA | 2020-07-01 12:00:00 |
|  2 | CWQSsar3qcssg      | 2020-07-01 15:00:00 |
|  3 | vxfqrt2adafz       | 2020-07-01 21:30:00 |
+----+--------------------+---------------------+
3 rows in set (0.00 sec)

执行计划如下：

mysql> explain select  *  from  tb_function  where   date(create_time)='2020-07-01';
+----+-------------+-------------+------------+------+---------------+------+---------+------+------+----------+-------------+
| id | select_type | table       | partitions | type | possible_keys | key  | key_len | ref  | rows | filtered | Extra       |
+----+-------------+-------------+------------+------+---------------+------+---------+------+------+----------+-------------+
|  1 | SIMPLE      | tb_function | NULL       | ALL  | NULL          | NULL | NULL    | NULL |   12 |   100.00 | Using where |
+----+-------------+-------------+------------+------+---------------+------+---------+------+------+----------+-------------+
1 row in set, 1 warning (0.00 sec)

执行计划中可以看出是进行了全面扫描

1.4  优化

因MySQL5.7不支持函数索引，所以需要修改SQL写法来实现走索引（或者使用虚拟列的方式），上述SQL可以修改为

mysql> select  *  from  tb_function  where   create_time>='2020-07-01' and create_time<date_add('2020-07-01',INTERVAL 1 day);
+----+--------------------+---------------------+
| id | name               | create_time         |
+----+--------------------+---------------------+
|  1 | anniuadaOAIFAPUHIA | 2020-07-01 12:00:00 |
|  2 | CWQSsar3qcssg      | 2020-07-01 15:00:00 |
|  3 | vxfqrt2adafz       | 2020-07-01 21:30:00 |
+----+--------------------+---------------------+
3 rows in set (0.00 sec)

执行计划如下：

mysql> explain select  *  from  tb_function  where   create_time>='2020-07-01' and create_time<date_add('2020-07-01',INTERVAL 1 day);
+----+-------------+-------------+------------+-------+-----------------+-----------------+---------+------+------+----------+-----------------------+
| id | select_type | table       | partitions | type  | possible_keys   | key             | key_len | ref  | rows | filtered | Extra                 |
+----+-------------+-------------+------------+-------+-----------------+-----------------+---------+------+------+----------+-----------------------+
|  1 | SIMPLE      | tb_function | NULL       | range | idx_create_time | idx_create_time | 6       | NULL |    3 |   100.00 | Using index condition |
+----+-------------+-------------+------------+-------+-----------------+-----------------+---------+------+------+----------+-----------------------+
1 row in set, 1 warning (0.00 sec)

可见，修改后，使用了索引。

2、 MySQL8.0

MySQL8.0的索引特性增加了函数索引。其实MySQL5.7中推出了虚拟列的功能，而MySQL8.0的函数索引也是依据虚拟列来实现的。将上述的案例在MySQL8.0中实现情况如下文所述。

2.1  创建函数索引

在将上述的表及数据在MySQL8.0的实例上创建，然后创建create_time的函数索引，SQL如下

mysql> alter  table tb_function add key idx_create_time((date(create_time))); --   注意里面字段的括号
Query OK, 0 rows affected (0.10 sec)
Records: 0  Duplicates: 0  Warnings: 0

2.2  按时间查询

mysql> select  *  from  tb_function  where   date(create_time)='2020-07-01';
+----+--------------------+---------------------+
| id | name               | create_time         |
+----+--------------------+---------------------+
|  1 | anniuadaOAIFAPUHIA | 2020-07-01 12:00:00 |
|  2 | CWQSsar3qcssg      | 2020-07-01 15:00:00 |
|  3 | vxfqrt2adafz       | 2020-07-01 21:30:00 |
+----+--------------------+---------------------+
3 rows in set (0.00 sec)

执行计划如下

mysql> explain select  *  from  tb_function  where   date(create_time)='2020-07-01';
+----+-------------+-------------+------------+------+-----------------+-----------------+---------+-------+------+----------+-------+
| id | select_type | table       | partitions | type | possible_keys   | key             | key_len | ref   | rows | filtered | Extra |
+----+-------------+-------------+------------+------+-----------------+-----------------+---------+-------+------+----------+-------+
|  1 | SIMPLE      | tb_function | NULL       | ref  | idx_create_time | idx_create_time | 4       | const |    3 |   100.00 | NULL  |
+----+-------------+-------------+------------+------+-----------------+-----------------+---------+-------+------+----------+-------+
1 row in set, 1 warning (0.00 sec)

可见，在MySQL8.0 创建对应的函数索引后，不改变SQL写法的前提下，查询的列上进行对应的函数计算后也可以走索引。

关于MySQL函数索引的优化及MySQL8.0函数索引还可以有更多的场景进行测试，建议大家多动手试试，提高SQL改写及优化的能力。

想了解更多内容或参与技术交流可以关注微信公众号【数据库干货铺】或进技术交流群沟通。

>>>>

近期活动

2020年，Gdevops全球敏捷运维峰会开启了传播前沿技术、助力技术人成长进阶、促进跨界交流合作的第五个年头。依托迄今成功举办的17场大会在分享议题上的精心打磨、在技术圈子里的口碑传播，Gdevops在展开新一年技术巡演中邀请到更顶级的嘉宾阵容，将带来更重磅的科技成果与独家实践。点击链接了解更多情况

2020 Gdevops全球敏捷运维峰会

往期精彩回顾

1.  MySQL高可用之MHA集群部署

2.  mysql8.0新增用户及加密规则修改的那些事

3.  比hive快10倍的大数据查询利器-- presto

4.  监控利器出鞘：Prometheus+Grafana监控MySQL、Redis数据库

5.  PostgreSQL主从复制--物理复制

6.  MySQL传统点位复制在线转为GTID模式复制

7.  MySQL敏感数据加密及解密

8.  MySQL数据备份及还原（一）

9.  MySQL数据备份及还原（二）

扫

码

关

注