本文目录
显示
1.
1. LIMIT语句
2.
2. 隐式转换
4.
4. 混合排序
5.
5. EXISTS语句
6.
6. 条件下推
7.
7. 提前缩小范围
8.
8. 中间结果集下推
9.
总结
现如今,越来越多的客户将自己的应用建立在MySQL数据库之上,甚至是从Oracle迁移到MySQL上来。但也存在部分客户在使用MySQL数据库的过程中遇到一些比如响应时间慢,CPU打满等情况。阿里云RDS专家服务团队帮助云上客户解决过很多紧急问题。现将《ApsaraDB专家诊断报告》中出现的部分常见SQL问题总结如下,供大家参考。
MySQL数据库的8种常见错误用法:
1. LIMIT语句
分页查询是最常用的场景之一,但也通常也是最容易出问题的地方。比如对于下面简单的语句,一般DBA想到的办法是在type, name, create_time字段上加组合索引。这样条件排序都能有效的利用到索引,性能迅速提升。
SELECT* FROMoperation WHEREtype=\'SQLStats\' ANDname=\'SlowLog\' ORDERBYcreate_time LIMIT1000,10;
好吧,可能90%以上的DBA解决该问题就到此为止。但当 LIMIT 子句变成 “LIMIT 1000000,10” 时,程序员仍然会抱怨:我只取10条记录为什么还是慢?
要知道数据库也并不知道第1000000条记录从什么地方开始,即使有索引也需要从头计算一次。出现这种性能问题,多数情形下是程序员偷懒了。在前端数据浏览翻页,或者大数据分批导出等场景下,是可以将上一页的最大值当成参数作为查询条件的。SQL重新设计如下:
SELECT* FROMoperation WHEREtype=\'SQLStats\' ANDname=\'SlowLog\' ANDcreate_time >\'2017-03-16 14:00:00\' ORDERBYcreate_timelimit10;
在新设计下查询时间基本固定,不会随着数据量的增长而发生变化。
2. 隐式转换
SQL语句中查询变量和字段定义类型不匹配是另一个常见的错误。比如下面的语句:
mysql> explain extended SELECT * > FROM my_balance b > WHERE b.bpn =14000000123 > AND b.isverified IS NULL ; mysql> show warnings; | Warning |1739| Cannot use ref accessonindex\'bpn\'due to typeorcollation conversiononfield\'bpn\'
其中字段bpn的定义为varchar(20),MySQL的策略是将字符串转换为数字之后再比较。函数作用于表字段,索引失效。
上述情况可能是应用程序框架自动填入的参数,而不是程序员的原意。现在应用框架很多很繁杂,使用方便的同时也小心它可能给自己挖坑。
3. 关联更新、删除
虽然MySQL5.6引入了物化特性,但需要特别注意它目前仅仅针对查询语句的优化。对于更新或删除需要手工重写成JOIN。
比如下面UPDATE语句,MySQL实际执行的是循环/嵌套子查询(DEPENDENT SUBQUERY),其执行时间可想而知。
UPDATEoperation o SETstatus=\'applying\' WHEREo.idIN(SELECTid FROM(SELECTo.id, o.status FROMoperation o WHEREo.group =123 ANDo.statusNOTIN(\'done\') ORDERBYo.parent, o.id LIMIT1) t);
执行计划:
+----+--------------------+-------+-------+---------------+---------+---------+-------+------+-----------------------------------------------------+ | id |select_type| table |type| possible_keys |key| key_len |ref| rows |Extra| +----+--------------------+-------+-------+---------------+---------+---------+-------+------+-----------------------------------------------------+ |1| PRIMARY |o| index || PRIMARY |8| |24| Using where; Using temporary | | 2 |DEPENDENT SUBQUERY| || || || |Impossible WHERE noticed after reading const tables| |3| DERIVED |o| ref |idx_2,idx_5| idx_5 |8| const |1| Using where; Using filesort | +----+--------------------+-------+-------+---------------+---------+---------+-------+------+-----------------------------------------------------+
重写为JOIN之后,子查询的选择模式从DEPENDENT SUBQUERY变成DERIVED,执行速度大大加快,从7秒降低到2毫秒。
UPDATEoperation o
JOIN(SELECTo.id,
o.status
FROMoperation o
WHEREo.group =123
ANDo.statusNOTIN(‘done’)
ORDERBYo.parent,
o.id
LIMIT1) t
ONo.id = t.id
SETstatus=‘applying’
执行计划简化为:
+----+-------------+-------+------+---------------+-------+---------+-------+------+-----------------------------------------------------+ | id |select_type| table |type| possible_keys |key| key_len |ref| rows |Extra| +----+-------------+-------+------+---------------+-------+---------+-------+------+-----------------------------------------------------+ |1| PRIMARY || || || || Impossible WHERE noticed after reading const tables | | 2 |DERIVED| o |ref| idx_2,idx_5 |idx_5| 8 |const| 1 |Using where; Using filesort| +----+-------------+-------+------+---------------+-------+---------+-------+------+-----------------------------------------------------+
4. 混合排序
MySQL不能利用索引进行混合排序。但在某些场景,还是有机会使用特殊方法提升性能的。
SELECT* FROMmy_order o INNERJOINmy_appraise aONa.orderid = o.id ORDERBYa.is_replyASC, a.appraise_timeDESC LIMIT0,20
执行计划显示为全表扫描:
+----+-------------+-------+--------+-------------+---------+---------+---------------+---------+-+ | id |select_type| table |type| possible_keys |key| key_len |ref| rows |Extra +----+-------------+-------+--------+-------------+---------+---------+---------------+---------+-+ | 1 |SIMPLE| a |ALL| idx_orderid |NULL| NULL |NULL| 1967647 |Using filesort| |1| SIMPLE |o| eq_ref |PRIMARY| PRIMARY |122| a.orderid |1| NULL | +----+-------------+-------+--------+---------+---------+---------+-----------------+---------+-+
由于is_reply只有0和1两种状态,我们按照下面的方法重写后,执行时间从1.58秒降低到2毫秒。
SELECT* FROM((SELECT* FROMmy_order o INNERJOINmy_appraise a ONa.orderid = o.id ANDis_reply =0 ORDERBYappraise_timeDESC LIMIT0,20) UNIONALL (SELECT* FROMmy_order o INNERJOINmy_appraise a ONa.orderid = o.id ANDis_reply =1 ORDERBYappraise_timeDESC LIMIT0,20)) t ORDERBYis_replyASC, appraisetimeDESC LIMIT20;
5. EXISTS语句
MySQL对待EXISTS子句时,仍然采用嵌套子查询的执行方式。如下面的SQL语句:
SELECT* FROMmy_neighbor n LEFTJOINmy_neighbor_apply sra ONn.id = sra.neighbor_id ANDsra.user_id =\'xxx\' WHEREn.topic_status <4 ANDEXISTS(SELECT1 FROMmessage_info m WHEREn.id = m.neighbor_id ANDm.inuser =\'xxx\') ANDn.topic_type <>5
执行计划为:
-----------------------+---------+-------+---------+ -----+ | id |select_type| table |type| possible_keys |key| key_len |ref| rows |Extra| +----+--------------------+-------+------+ -----+------------------------------------------+---------+-------+---------+ -----+ |1| PRIMARY |n| ALL || NULL |NULL| NULL |1086041| Using where | | 1 |PRIMARY| sra |ref| |idx_user_id| 123 |const| 1 |Using where| |2| DEPENDENT SUBQUERY |m| ref || idx_message_info |122| const |1| Using index condition; Using where | +----+--------------------+-------+------+ -----+------------------------------------------+---------+-------+---------+ -----+
去掉exists更改为join,能够避免嵌套子查询,将执行时间从1.93秒降低为1毫秒。
SELECT* FROMmy_neighbor n INNERJOINmessage_info m ONn.id = m.neighbor_id ANDm.inuser =\'xxx\' LEFTJOINmy_neighbor_apply sra ONn.id = sra.neighbor_id ANDsra.user_id =\'xxx\' WHEREn.topic_status <4 ANDn.topic_type <>5
新的执行计划:
+----+-------------+-------+--------+ -----+------------------------------------------+---------+ -----+------+ -----+ | id |select_type| table |type| possible_keys |key| key_len |ref| rows |Extra| +----+-------------+-------+--------+ -----+------------------------------------------+---------+ -----+------+ -----+ |1| SIMPLE |m| ref || idx_message_info |122| const |1| Using index condition | | 1 |SIMPLE| n |eq_ref| |PRIMARY| 122 |ighbor_id| 1 |Using where| |1| SIMPLE |sra| ref || idx_user_id |123| const |1| Using where | +----+-------------+-------+--------+ -----+------------------------------------------+---------+ -----+------+ -----+
6. 条件下推
外部查询条件不能够下推到复杂的视图或子查询的情况有:
- 聚合子查询;
- 含有LIMIT的子查询;
- UNION 或UNION ALL子查询;
- 输出字段中的子查询;
如下面的语句,从执行计划可以看出其条件作用于聚合子查询之后:
SELECT* FROM(SELECTtarget, Count(*) FROMoperation GROUPBYtarget) t WHEREtarget =\'rm-xxxx\' +----+-------------+------------+-------+---------------+-------------+---------+-------+------+-------------+ |id| select_type |table|type| possible_keys |key| key_len |ref|rows| Extra | +----+-------------+------------+-------+---------------+-------------+---------+-------+------+-------------+ |1| PRIMARY | <derived2> |ref| <auto_key0> | <auto_key0> |514| const |2|Usingwhere| |2| DERIVED | operation |index| idx_4 | idx_4 |519|NULL|20|Usingindex| +----+-------------+------------+-------+---------------+-------------+---------+-------+------+-------------+
确定从语义上查询条件可以直接下推后,重写如下:
SELECTtarget, Count(*) FROMoperation WHEREtarget =\'rm-xxxx\' GROUPBYtarget
执行计划变为:
+----+-------------+-----------+------+---------------+-------+---------+-------+------+--------------------+ | id |select_type| table |type| possible_keys |key| key_len |ref| rows |Extra| +----+-------------+-----------+------+---------------+-------+---------+-------+------+--------------------+ |1| SIMPLE |operation| ref |idx_4| idx_4 |514| const |1| Using where; Using index | +----+-------------+-----------+------+---------------+-------+---------+-------+------+--------------------+
7. 提前缩小范围
先上初始SQL语句:
SELECT* FROMmy_order o LEFTJOINmy_userinfo u ONo.uid = u.uid LEFTJOINmy_productinfo p ONo.pid = p.pid WHERE( o.display =0) AND( o.ostaus =1) ORDERBYo.selltimeDESC LIMIT0,15
该SQL语句原意是:先做一系列的左连接,然后排序取前15条记录。从执行计划也可以看出,最后一步估算排序记录数为90万,时间消耗为12秒。
+----+-------------+-------+--------+---------------+---------+---------+-----------------+--------+----------------------------------------------------+ | id |select_type| table |type| possible_keys |key| key_len |ref| rows |Extra| +----+-------------+-------+--------+---------------+---------+---------+-----------------+--------+----------------------------------------------------+ |1| SIMPLE |o| ALL |NULL| NULL |NULL| NULL |909119| Using where; Using temporary; Using filesort | | 1 |SIMPLE| u |eq_ref| PRIMARY |PRIMARY| 4 |o.uid| 1 |NULL| |1| SIMPLE |p| ALL |PRIMARY| NULL |NULL| NULL |6| Using where; Using join buffer (Block Nested Loop) | +----+-------------+-------+--------+---------------+---------+---------+-----------------+--------+----------------------------------------------------+
由于最后WHERE条件以及排序均针对最左主表,因此可以先对my_order排序提前缩小数据量再做左连接。SQL重写后如下,执行时间缩小为1毫秒左右。
SELECT* FROM( SELECT* FROMmy_order o WHERE( o.display =0) AND( o.ostaus =1) ORDERBYo.selltimeDESC LIMIT0,15 ) o LEFTJOINmy_userinfo u ONo.uid = u.uid LEFTJOINmy_productinfo p ONo.pid = p.pid ORDERBYo.selltimeDESC limit0,15
再检查执行计划:子查询物化后(select_type=DERIVED)参与JOIN。虽然估算行扫描仍然为90万,但是利用了索引以及LIMIT 子句后,实际执行时间变得很小。
+----+-------------+------------+--------+---------------+---------+---------+-------+--------+----------------------------------------------------+ | id |select_type| table |type| possible_keys |key| key_len |ref| rows |Extra| +----+-------------+------------+--------+---------------+---------+---------+-------+--------+----------------------------------------------------+ |1| PRIMARY |<derived2>| ALL |NULL| NULL |NULL| NULL |15| Using temporary; Using filesort | | 1 |PRIMARY| u |eq_ref| PRIMARY |PRIMARY| 4 |o.uid| 1 |NULL| |1| PRIMARY |p| ALL |PRIMARY| NULL |NULL| NULL |6| Using where; Using join buffer (Block Nested Loop) | | 2 |DERIVED| o |index| NULL |idx_1| 5 |NULL| 909112 |Using where| +----+-------------+------------+--------+---------------+---------+---------+-------+---
8. 中间结果集下推
再来看下面这个已经初步优化过的例子(左连接中的主表优先作用查询条件):
SELECTa.*, c.allocated FROM( SELECTresourceid FROMmy_distribute d WHEREisdelete =0 ANDcusmanagercode =\'1234567\' ORDERBYsalecodelimit20) a LEFTJOIN ( SELECTresourcesid,sum(ifnull(allocation,0) *12345) allocated FROMmy_resources GROUPBYresourcesid) c ONa.resourceid = c.resourcesid
那么该语句还存在其它问题吗?不难看出子查询 c 是全表聚合查询,在表数量特别大的情况下会导致整个语句的性能下降。
其实对于子查询 c,左连接最后结果集只关心能和主表resourceid能匹配的数据。因此我们可以重写语句如下,执行时间从原来的2秒下降到2毫秒。
SELECTa.*, c.allocated FROM( SELECTresourceid FROMmy_distribute d WHEREisdelete =0 ANDcusmanagercode =\'1234567\' ORDERBYsalecodelimit20) a LEFTJOIN ( SELECTresourcesid,sum(ifnull(allocation,0) *12345) allocated FROMmy_resources r, ( SELECTresourceid FROMmy_distribute d WHEREisdelete =0 ANDcusmanagercode =\'1234567\' ORDERBYsalecodelimit20) a WHEREr.resourcesid = a.resourcesid GROUPBYresourcesid) c ONa.resourceid = c.resourcesid
但是子查询 a 在我们的SQL语句中出现了多次。这种写法不仅存在额外的开销,还使得整个语句显的繁杂。使用WITH语句再次重写:
WITH a AS ( SELECTresourceid FROMmy_distribute d WHEREisdelete =0 ANDcusmanagercode =\'1234567\' ORDERBYsalecodelimit20) SELECTa.*, c.allocated FROMa LEFTJOIN ( SELECTresourcesid,sum(ifnull(allocation,0) *12345) allocated FROMmy_resources r, a WHEREr.resourcesid = a.resourcesid GROUPBYresourcesid) c ONa.resourceid = c.resourcesid
总结
数据库编译器产生执行计划,决定着SQL的实际执行方式。但是编译器只是尽力服务,所有数据库的编译器都不是尽善尽美的。上述提到的多数场景,在其它数据库中也存在性能问题。了解数据库编译器的特性,才能避规其短处,写出高性能的SQL语句。
程序员在设计数据模型以及编写SQL语句时,要把算法的思想或意识带进来。编写复杂SQL语句要养成使用WITH语句的习惯。简洁且思路清晰的SQL语句也能减小数据库的负担 ^^。