Question

以下是我的查询，我正在navicat我的sql执行包含超过91000条记录的表。

select l.* 
from localcdrsmem l 
LEFT JOIN (
         SELECT LocalCDRID, ForeignCDRsID 
         from localcdrsmem l 
         INNER JOIN foreigncdrsmem f 
         ON l.CalledNumber=f.CalledNumber and 
         ABS(l.Duration-f.Duration)<= 2 and            
         ABS(TIMESTAMPDIFF(SECOND,f.StartTime,l.StartTime))<=100 
         group by l.CalledNumber) 
as temp ON l.LocalCDRID = temp.LocalCDRID 
where temp.LocalCDRID IS NULL;

表是：

localcdrsmem - > 
+---------+------------+--------+----------
localcdrid | callednumber | duration | starttime
+---------+------------+--------+----------+
foreigncdrsmem- > 
+------------+------------+--------+----------
foreigncdrid | callednumber | duration | starttime
+---------+------------+--------+----------+

优化此查询的任何想法??

Answer 1

select l.* from localcdrsmem l LEFT JOIN 
(SELECT LocalCDRID, ForeignCDRsID from localcdrsmem l 
INNER JOIN foreigncdrsmem f
ON l.CalledNumber=f.CalledNumber 
WHERE ABS(l.Duration-f.Duration)<= 2 
and ABS(TIMESTAMPDIFF(SECOND,f.StartTime,l.StartTime))<=100 
group by l.CalledNumber) as temp
ON l.LocalCDRID = temp.LocalCDRID where temp.LocalCDRID IS NULL;

尝试以上查询。此查询执行得更快。只需在此语句and之后使用where子句更改ON l.CalledNumber=f.CalledNumber语句。您可以看到我们的查询之间的区别。

对于表中的大数据，左连接查询执行速度非常慢

1 个答案: