Date: Mon, 21 Mar 2016 16:35:38 +0100
From: hubert depesz lubaczewski <depesz@depesz.com>
To: Nicolas Aubert <aubertnicolas11@gmail.com>
Cc: pgsql-admin@postgresql.org
Subject: Re: vacuum analyse waiting
Message-ID: <20160321153537.GA14547@depesz.com>
Reply-To: depesz@depesz.com
References: 
 <CAEnLzzVU0U1OnYLEq6JVX7LVTywrDRZJseExtdNy9vQC-tbh8g@mail.gmail.com>
 <20160321110250.GE14681@depesz.com>
 <CAEnLzzW16WhwHaV+vpc3QpStQ0YPT3EBsm+jo=MVeb0VzBTqLg@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
In-Reply-To: 
 <CAEnLzzW16WhwHaV+vpc3QpStQ0YPT3EBsm+jo=MVeb0VzBTqLg@mail.gmail.com>
User-Agent: Mutt/1.5.24 (2015-08-30)
Precedence: bulk
Sender: pgsql-admin-owner@postgresql.org

On Mon, Mar 21, 2016 at 04:26:40PM +0100, Nicolas Aubert wrote:
> the vacuum process is the 22510
> here is what I found in pg_locks
> 
> 
> virtualtransaction |  pid  |           mode           | granted | fastpath
> ------------+---------------+---------+-------+----------+--------------------+-------+--------------------------+---------+----------
> 
> 3/376480           | 22510 | ExclusiveLock            | t       |
>             |               |         |       |          |
> 3/376480           | 22510 | ShareUpdateExclusiveLock | f       | f
> 

I'm sure there are more columns in pg_locks.

What you need to find is the rows with pid of vacuum, that have granted
= 'false', and see what the lock is on (the columns you ignored).

Then you check what has conflicting locks on the required things.

Best regards,

depesz


-- 
Sent via pgsql-admin mailing list (pgsql-admin@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-admin