一百万数据索引实例測试--mysql

推荐书籍：http://pan.baidu.com/s/1sjJIyRV

任务描写叙述：

如果一高频查询例如以下 

SELECT * FROM user WHERE area='amoy' AND sex=0 ORDER BY last_login DESC limit 30; 

怎样建立索引？描写叙述考虑的过程

user表例如以下： 

初始化100W条数据，当中。area要通过IP查询生成，sex为 0,1 随机

CREATE TABLE user ( 

id int(10) NOT NULL AUTOINCREMENT COMMENT '自增编号', 

username varchar(30) NOT NULL DEFAULT '0' COMMENT '用户名', 

password varchar(30) NOT NULL DEFAULT '0' COMMENT '密码', 

area varchar(30) NOT NULL COMMENT '地址', 

sex int(10) NOT NULL COMMENT '性别0，男；1，女。

', 

last_login int(10) NOT NULL COMMENT '近期一次登录时间戳', 

PRIMARY KEY (id) 

) ENGINE=InnoDB AUTOINCREMENT=892013 DEFAULT CHARSET=latin1

终于我的索引 

(last_login,area)

数据例如以下：http://pan.baidu.com/s/1eQy0eQI

測试结果：http://pan.baidu.com/s/1jGn2AcY

watermark/2/text/aHR0cDovL2Jsb2cuY3Nkbi5uZXQvbGFtcF93YXRlcg==/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70/gravity/SouthEast" alt="">

索引原则：

1.where和order by等的字段建立索引

2.使用唯一索引：对于last_login,area等字段反复的次数比較少，能够使用索引；而sex无非就两个值：性别1。男；2。不值得索引

3.多列索引：不要为每个列单独建立索引。这样并不能将mysql索引的效率最大化。使用“索引合并策略”

4.选择合理的索引列顺序：索引列的顺序意味着索引首先依照最左列进行排序。然后是第二列，以此类推。如(lastlogin,area)会先依照 lastlogin 进行排序。然后才是area。

5.将选择性最高的索引放到前面。也就是会所依照这个条件搜索到的数据最少，选择性就越高。比方选择性：last_login> area> sex。

6.索引不是越多越好。适合的索引能够提高查询效率。可是会减少写入效率。依据项目保持两者的平衡性最好了。

总结上面，首先sex不适合建立索引，有没有索引对于效率的提升意义不大，其次索引会依照最左列进行排序，因此将last_login放到最前面

測试过程：

user表 

没有不论什么索引的查询相关日志： 

SELECT * FROM user WHERE area='美国ATT用户' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.57s 

SELECT * FROM user WHERE area='泰国' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.56s 

SELECT * FROM user WHERE area='台湾省台湾大宽频' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.55s 

SELECT * FROM user WHERE area='美国弗吉尼亚州' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.59s 

SELECT * FROM user WHERE area='德国奔驰汽车' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.55s 

SELECT * FROM user WHERE area='台湾省中华电信' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.55s 

SELECT * FROM user WHERE area='韩国' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.57s 

SELECT * FROM user WHERE area='拉美地区' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.58s 

SELECT * FROM user WHERE area='美国纽约(Prudential)' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.57s 

SELECT * FROM user WHERE area='印度尼西亚' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.57s 

共花费时间：5.66s

建立索引area: 

ALTER TABLE user ADD INDEX index_area (area)
; 

SELECT * FROM user WHERE area='美国ATT用户' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.06s 

SELECT * FROM user WHERE area='泰国' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.02s 

SELECT * FROM user WHERE area='台湾省台湾大宽频' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.00s 

SELECT * FROM user WHERE area='美国弗吉尼亚州' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.10s 

SELECT * FROM user WHERE area='德国奔驰汽车' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.04s 

SELECT * FROM user WHERE area='台湾省中华电信' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.02s 

SELECT * FROM user WHERE area='韩国' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.11s 

SELECT * FROM user WHERE area='拉美地区' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.20s 

SELECT * FROM user WHERE area='美国纽约(Prudential)' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.07s 

SELECT * FROM user WHERE area='印度尼西亚' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.04s 

共花费时间：0.66s 

可见。建立area以后对性能的影响是巨大的(5.66/0.66 约为8.5758倍) 

删除索引：ALTER TABLE user DROP INDEX index_area; 

删除area索引发现时间又变成了0.57s

建立lastlogin索引： 

SELECT * FROM user WHERE area='美国ATT用户' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.03s 

SELECT * FROM user WHERE area='泰国' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.09s 

SELECT * FROM user WHERE area='台湾省台湾大宽频' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.51s 

SELECT * FROM user WHERE area='美国弗吉尼亚州' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.01s 

SELECT * FROM user WHERE area='德国奔驰汽车' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.04s 

SELECT * FROM user WHERE area='台湾省中华电信' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.07s 

SELECT * FROM user WHERE area='韩国' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.01s 

SELECT * FROM user WHERE area='拉美地区' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.01s 

SELECT * FROM user WHERE area='美国纽约(Prudential)' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.04s 

SELECT * FROM user WHERE area='印度尼西亚' AND sex=0 ORDER BY last_login DESC limit 30; 0.06s 

共花费时间：0.87s 

相同可以提升性能（5.66/0.87 约为6.5057倍）

建立sex索引： 

ALTER TABLE user ADD INDEX index_sex (sex)
; 

SELECT * FROM user WHERE area='美国ATT用户' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.87s 

SELECT * FROM user WHERE area='泰国' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.87s 

SELECT * FROM user WHERE area='台湾省台湾大宽频' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.87s 

SELECT * FROM user WHERE area='美国弗吉尼亚州' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.89s 

SELECT * FROM user WHERE area='德国奔驰汽车' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.88s 

SELECT * FROM user WHERE area='台湾省中华电信' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.87s 

SELECT * FROM user WHERE area='韩国' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.86s 

SELECT * FROM user WHERE area='拉美地区' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.88s 

SELECT * FROM user WHERE area='美国纽约(Prudential)' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.87s 

SELECT * FROM user WHERE area='印度尼西亚' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.87s 

共花费时间：8.73s 

相同可以提升性能（5.66s/8.73 约为0.6483倍）效率反而减少了？？求解？ 

建立这个sex索引还不如不建。

删除索引： 

ALTER TABLE user DROP INDEX index_sex; 

发现时间又变成了0.57s左右，

建立两个单独的索引： 

ALTER TABLE user 

ADD INDEX index_area (area)
, 

ADD INDEX index_last_login (last_login)
;

SELECT * FROM user WHERE area='美国ATT用户' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.09s 

SELECT * FROM user WHERE area='泰国' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.33s 

SELECT * FROM user WHERE area='台湾省台湾大宽频' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.21s 

SELECT * FROM user WHERE area='美国弗吉尼亚州' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.01s 

SELECT * FROM user WHERE area='德国奔驰汽车' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.28s 

SELECT * FROM user WHERE area='台湾省中华电信' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.02s 

SELECT * FROM user WHERE area='韩国' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.02s 

SELECT * FROM user WHERE area='拉美地区' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.02s 

SELECT * FROM user WHERE area='美国纽约(Prudential)' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.03s 

SELECT * FROM user WHERE area='印度尼西亚' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.67s

发现建立两个单独的索引还不如仅仅建立一个索引 

删除索引： 

发现时间又变成了0.57s左右。

建立一个的联合索引： 

ALTER TABLE user 

ADD INDEX index_last_login_area (last_login,area)
, 

SELECT * FROM user WHERE area='美国ATT用户' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.00s 

SELECT * FROM user WHERE area='泰国' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.00s 

SELECT * FROM user WHERE area='台湾省台湾大宽频' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.00s 

SELECT * FROM user WHERE area='美国弗吉尼亚州' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.00s 

SELECT * FROM user WHERE area='德国奔驰汽车' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.00s 

SELECT * FROM user WHERE area='台湾省中华电信' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.00s 

SELECT * FROM user WHERE area='韩国' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.00s 

SELECT * FROM user WHERE area='拉美地区' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.00s 

SELECT * FROM user WHERE area='美国纽约(Prudential)' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.00s 

SELECT * FROM user WHERE area='印度尼西亚' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.00s 

额。第二条数据这是怎么了，我測试了5次都在这附近晃悠哈！

这尼玛。找对索引啦。就该这么建立，查询不出来须要的时间啦！预计就是我们须要的索引啦！

！

！！

删除索引： 

发现时间又变成了0.57s左右，

建立一个的联合索引： 

ALTER TABLE user 

ADD INDEX index_sex_last_login_area (sex,last_login,area) 

SELECT * FROM user WHERE area='美国ATT用户' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.18s 

SELECT * FROM user WHERE area='泰国' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.17s 

SELECT * FROM user WHERE area='台湾省台湾大宽频' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.81s 

SELECT * FROM user WHERE area='美国弗吉尼亚州' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.01s 

SELECT * FROM user WHERE area='德国奔驰汽车' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.02s 

SELECT * FROM user WHERE area='台湾省中华电信' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.04s 

SELECT * FROM user WHERE area='韩国' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.01s 

SELECT * FROM user WHERE area='拉美地区' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.01s 

SELECT * FROM user WHERE area='美国纽约(Prudential)' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.03s 

SELECT * FROM user WHERE area='印度尼西亚' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.04s 

sex怎么总是你在拖后腿啊！

把你调整到索引的最后一个吧。 

删除索引： 

发现时间又变成了0.57s左右，

建立一个的联合索引： 

ALTER TABLE user 

ADD INDEX index_last_login_area_sex (area,last_login,sex) 

SELECT * FROM user WHERE area='美国ATT用户' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.03s 

SELECT * FROM user WHERE area='泰国' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.07s 

SELECT * FROM user WHERE area='台湾省台湾大宽频' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.50s 

SELECT * FROM user WHERE area='美国弗吉尼亚州' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.02s 

SELECT * FROM user WHERE area='德国奔驰汽车' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.05s 

SELECT * FROM user WHERE area='台湾省中华电信' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.06s 

SELECT * FROM user WHERE area='韩国' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.02s 

SELECT * FROM user WHERE area='拉美地区' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.02s 

SELECT * FROM user WHERE area='美国纽约(Prudential)' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.04s 

SELECT * FROM user WHERE area='印度尼西亚' AND sex=0 ORDER BY lastlogin DESC limit 30; 0.06s

综上所述：1.建立索引不一定可以加快查询效率如sex这样的给反复次数特别多的列添加索引如sex这样的会减少查询效率，详细的原因有待查找 

2.给反复次数比較少的列添加索引还是可以大幅度提高效率 

3.给where和orderby之后的字段加入索引才会加快查询效率 

4.为每个列单独建立索引，不能将索引的效率最大化，应该使用索引合并策略，即依据查询条件。建立联合索引 

5.联合索引的顺序问题：将选择性高的索引放到前面 

6.依据资料建立索引意味着索引依照最左列进行排序，然后事第二列。以此类推。如（lastlogin ,area）就会依照lastlogin进行排序，然后才是area 

7.依据这次的这个查询条件来说最好的索引是：ALTER TABLE userADD INDEX index_last_login_area (last_login,area)。

在公司能有个机会。查看资料和实践索引真的非常不错哈！

推荐书籍：高性能mysql（第三版）

PDF版本号的：http://pan.baidu.com/s/1sjJIyRV