MySQL 优化总结 (五)

Join Vs select 结果集作列

  有时候我们需要关联很多表统计:

  select f1, f2, count(*) from a join b join c join d join e ……

  如果结果集很小,只统计出几条数据来,那么可以换成这样写

  Select f1, f2, (select count(*) from a join b ) from c……

  比如下面的例子:  

1 select  t1.assignment_id assignmentId,t5.person_id,
2 
                       t2.subject_name subjectName,
3 
                       t3.dict_name assignType,
4 
                       t1.assign_date assignDate,
5 
                       t4.totalcount totalCount,
6                        (select dict_name from bas_diction where dict_id = t5.assign_marks) as
 assignMarks,
7                        (case t5.if_submit when 1 then '是' else '否' end) as ifSubmit,

8                         t6.errorcount errorCount,
9 
                       t6.nocheckCount,
10                        case t1.online_assignment when 1 then '是' else '否' end as ifOnline,

11                        (select dict_name from bas_diction where dict_id = t5.write_appraise) as  writeAppraise
12 
                from hom_assignmentinfo t1,bas_subject t2,bas_diction t3,
13                      (select t1.assignment_id,count(*) as
 totalcount from hom_assignmentinfo t1,hom_assignment t2
14                       where t1.assignment_id =
 t2.hom_assignment_id
15                       group by t1.assignment_id) t4,hom_assignment_appraise t5 left join

16                      (select  t1.assignment_id,t3.appraise_id,
17                             sum(case t3.check_result when 3003001 then 0 else 1 end) as
 errorcount, 
18                             sum(case when t3.check_result=3003001 or t3.recheck_result=3003001 then 0 else 1 end)as
 nocheckCount
19 
                      from hom_assignmentinfo t1,hom_check_assignment t3
20                       where t1.assignment_id =
 t3.assignment_id
21                           and t1.person_id='13042'

22                       group by t1.assignment_id,t3.appraise_id) t6 on (t5.appraise_id= t6.appraise_id)    
23                 where t1.subject_id =
 t2.subject_id
24                       and t1.assign_type =
 t3.dict_id
25                       and t1.assignment_id =
 t4.assignment_id
26                       and t1.assignment_id =
 t5.assignment_id
27                       and not exists (select 1
 from hom_assignmentinfo t11,hom_assignment_appraise t12 
28                        where t11.assignment_id=t12.assignment_id and
 
29                                t11.online_assignment = 0 and
 
30                             t12.person_id='13042' and 

31                             t11.subject_id = t1.subject_id and t11.assign_date >  t1.assign_date)
32                       and t1.assign_date is not null

33                       and t5.person_id='13042'
34                  order by t1.assign_date desc;    
35

  改写成下面的,速度由0.625秒优化到: 0.032  

1 select  t1.assignment_id assignmentId,t5.person_id,
2 
                       t2.subject_name subjectName,
3 
                       t3.dict_name assignType,
4 
                       t1.assign_date assignDate,
5                       (select  count(*) as totalcount from hom_assignmentinfo where assignment_id=
t1.assignment_id) totalcnt,
6 
                            (
7                                select  sum(case t3.check_result when 3003001 then 0 else 1 end
)  
8 
                                from hom_check_assignment t3
9                                 where t1.assignment_id =
 t3.assignment_id 
10 
                            )  errorcount,
11 
                            (
12                                select sum(case when t3.check_result=3003001 or t3.recheck_result=3003001 then 0 else 1 end)as
 nocheckCount
13 
                                from hom_check_assignment t3
14                                 where t1.assignment_id =
 t3.assignment_id 
15 
                            )  nocheckCount,
16                        (select dict_name from bas_diction where dict_id = t5.assign_marks) as
 assignMarks,
17                        (case t5.if_submit when 1 then '是' else '否' end) as ifSubmit,

18                case t1.online_assignment when 1 then '是' else '否' end as ifOnline,
19                        (select dict_name from bas_diction where dict_id = t5.write_appraise) as  writeAppraise
20 
                from hom_assignmentinfo t1,bas_subject t2,bas_diction t3, hom_assignment_appraise t5 
21                 where t1.subject_id =
 t2.subject_id
22                       and t1.assign_type =
 t3.dict_id
23                       and t1.assignment_id =
 t5.assignment_id
24                       and not exists (select 1
 from hom_assignmentinfo t11,hom_assignment_appraise t12 
25                        where t11.assignment_id=t12.assignment_id and
 
26                                t11.online_assignment = 0 and
 
27                             t12.person_id='13042' and 

28                             t11.subject_id = t1.subject_id and t11.assign_date >  t1.assign_date)
29                       and t1.assign_date is not null

30                       and t5.person_id='13042'
31                 order by t1.assign_date desc;    

 

 

如果什么都做不了,试试全索引扫描

  如果一个语句实在不能优化了,那么还有一个方法可以试试:索引覆盖。

  如果一个语句可以从索引上获取全部数据,就不需要通过索引再去读表,省了很多I/O。比如这样一个表        

  如果我要统计每个学生每道题的得分情况,我们除了要给每个表的主键外键建立索引,还要对【得分情况】的实际得分字段索引,这样,整个查询就可以从索引得到数据了。

 Join、In、not in、exist、not exist并不是绝对的

  网上很多教程讨论了join、in和exist 的性能差异,其实这不是绝对的,对于效率不理想的语句,还是应该换换写法试试看。

  Like

  Like毕竟效率太低,必要的话可以试试全文检索。对于中文全文检索,可以结合程序分词来实现。

  什么情况下查询用不到索引

  参见手册 7.4.3列索引, 7.4.4所列索引, 7.4.5 mysql如何使用索引

  去掉不必要的排序,如果必要,尽量用主键排序代替

  显而易见却容易被忽视的问题。

数据库参数配置

  最重要的参数就是内存,我们主要用的innodb引擎,所以下面两个参数调的很大

  # Additional memory pool that is used by InnoDB to store metadata

  # information. If InnoDB requires more memory for this purpose it will

  # start to allocate it from the OS. As this is fast enough on most

  # recent operating systems, you normally do not need to change this

  # value. SHOW INNODB STATUS will display the current amount used.

  innodb_additional_mem_pool_size = 64M

  # InnoDB, unlike MyISAM, uses a buffer pool to cache both indexes and

  # row data. The bigger you set this the less disk I/O is needed to

  # access data in tables. On a dedicated database server you may set this

  # parameter up to 80% of the machine physical memory size. Do not set it

  # too large, though, because competition of the physical memory may

  # cause paging in the operating system. Note that on 32bit systems you

  # might be limited to 2-3.5G of user level memory per process, so do not

  # set it too high.

  innodb_buffer_pool_size = 5G

  对于myisam,需要调整key_buffer_size

  当然调整参数还是要看状态,用show status语句可以看到当前状态,以决定改调整哪些参数

  Cretated_tmp_disk_tables 增加tmp_table_size

  Handler_read_key 高表示索引正确 Handler_read_rnd高表示索引不正确

  Key_reads/Key_read_requests 应小于0.01 计算缓存损失率,增加Key_buffer_size

  Opentables/Open_tables 增加table_cache

  select_full_join 没有实用索引的链接的数量。如果不为0,应该检查索引。

  select_range_check 如果不为0,该检查表索引。

  sort_merge_passes 排序算法已经执行的合并的数量。如果该值较大,应增加sort_buffer_size

  table_locks_waited 不能立即获得的表的锁的次数,如果该值较高,应优化查询

  Threads_created 创建用来处理连接的线程数。如果Threads_created较大,要增加 thread_cache_size值。

  缓存访问率的计算方法Threads_created/Connections。

合理的硬件资源和操作系统

  如果你的机器内存超过4G,那么毋庸置疑应当采用64位操作系统和64位mysql

  读写分离

  如果数据库压力很大,一台机器支撑不了,那么可以用mysql复制实现多台机器同步,将数据库的压力分散。  

  Master

  Slave1

  Slave2

  Slave3

  主库master用来写入,slave1—slave3都用来做select,每个数据库分担的压力小了很多。

  要实现这种方式,需要程序特别设计,写都操作master,读都操作slave,给程序开发带来了额外负担。当然目前已经有中间件来实现这个代理,对程序来读写哪些数据库是透明的。官方有个mysql-proxy,但是还是alpha版本的。新浪有个amobe for mysql,也可达到这个目的,结构如下  

  使用方法可以看amobe的手册。

评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值