mysql自带的slow_log分析工具mysqldumpslow
# mysqldumpslow --helpUsage: mysqldumpslow [ OPTS... ] [ LOGS... ]
-s ORDER 按照(al, at, ar, c, l, r, t)进行排序，默认是at
al: average lock time#平均lock时间
ar: average rows sent#平均rows sent
at: average query time#平均query时间
c: count#慢查询个数
l: lock time#lock时间
r: rows sent#rows sent
t: query time#query时间
-r 改变排序方式，原来为倒序改为正序
-t NUM 只展示前边n条记录
-g PATTERN 后边可以写一个正则匹配模式，大小写不敏感的
出现次数(Count) 执行最长时间(Time),累计总耗费时间(Time)等待锁的时间(Lock)发送给客户端的行总数(Rows),扫描的行总数(Rows)
Count: 176Time=6.52s (1146s)Lock=0.00s (0s)Rows=1.0 (176), user@2hosts
select count(*) as aggregate from `order` left join `gift` on `order`.`gift_id` = `gift`.`id` left join `user_1` on `order`.`uid` = `user_1`.`uid` left join `user_wallet_1` on `order`.`uid` = `user_wallet_1`.`uid` where `order`.`status` != 'S' and `order`.`type` = 'S' and `order`.`mall_id` = 'S'
mysqldumpslow会把某些个只是数值不一样的语句认定为同一个语句
页:
[1]