数据如下，表名video，发现vid和section_id重复的数据很多，要删除重复的数据只保留一条
+----------+-----------+------------+
| video_id | vid | section_id |
+----------+-----------+------------+
| 10908 | 284478110 | 528238111 |
| 10909 | 284478110 | 528248111 |
| 10884 | 278748110 | 528048111 |
| 10885 | 278758110 | 528058111 |<-重复
| 54155 | 278758110 | 528058111 |<-重复
| 10886 | 278768110 | 528068111 |
| 54156 | 278768110 | 528068111 |
| 10887 | 278778110 | 528078111 |
| 10888 | 282228110 | 528078111 |
| 54157 | 278778110 | 528078111 |
| 54158 | 282228110 | 528078111 |
| 10889 | 282148110 | 528088111 |
| 54159 | 282148110 | 528088111 |
| 10890 | 282158110 | 528098111 |
| 54160 | 282158110 | 528098111 |
| 10891 | 282168110 | 528108111 |
| 54161 | 282168110 | 528108111 |
| 10892 | 282178110 | 528118111 |
| 10893 | 282238110 | 528118111 |
| 54162 | 282178110 | 528118111 |
+----------+-----------+------------+
如果要保留小id的数据：
delete a from video a,video b where a.video_id>b.video_id and a.vid=b.vid and a.section_id=b.section_id
eg.
update ms_video a,ms_video b set a.status=0 where a.section_id=b.section_id and a.vid=b.vid and a.`status`=1 and b.`status`=1 and a.video_id<b.video_id; //video_id 为自增id
