在 2014/7/30 18:03, Rural Hunter 写道:
> I think I understand what happened now. I have another monitor script
> runs periodically and calls pg_cancel_backend and pg_terminate_backend
> for those hanging update sqls. However for some unkown reason the
> cancle and termiante command doesn't work at pgsql side for those
> update sqls.
>
With the log of the monitor&kill scipt, I can confirm that the
CLOSE_WAIT is not caused by it. I logged the netstat before actually
doing the kill and found the CLOSE_WAIT connections were already there.
So it must be something else caused the CLOSE_WAIT connections.