Re: There are many processes created and died frequently during PostgreSQl idle - Mailing list pgsql-general

From Amit Langote
Subject Re: There are many processes created and died frequently during PostgreSQl idle
Date
Msg-id CA+HiwqGNTh4UQ-k+Mvp9SLXgkuBcsumRELsB2STxWFH7PMrTyQ@mail.gmail.com
Whole thread Raw
In response to Re: There are many processes created and died frequently during PostgreSQl idle  (高健 <luckyjackgao@gmail.com>)
List pgsql-general
On Mon, May 27, 2013 at 10:56 AM, 高健 <luckyjackgao@gmail.com> wrote:
> Hi:
> Thanks for Jov's reply.
> I traced it again, and found they are really for autovacuum.
> I found that
> some will call proc_exit() from within  AutoVacLauncherMain function,
> some will call proc_exit() from within AutoVacWorkerMain function.
>
> But I wonder why not using only a few daemon , instead of  making those
> processes be short life.

The short lived processes are autovacuum "worker" processes which need
to exit as soon as their work is done. There does exist a daemon
called autovacuum launcher, which periodically forks autovacuum worker
processes to perform vacuum activity. You might want to read more
about them here:

http://www.postgresql.org/docs/9.2/static/routine-vacuuming.html#AUTOVACUUM

--
Amit Langote


pgsql-general by date:

Previous
From: 高健
Date:
Subject: Re: There are many processes created and died frequently during PostgreSQl idle
Next
From: Ash Damle
Date:
Subject: Human readable or full text of xlog (pg_xlog) for vs 9.2