Re: BUG #13472: VACUUM ANALYZE hangs on certain tables

From: Andres Freund <andres(at)anarazel(dot)de>
To: matthew(dot)seaman(at)adestra(dot)com
Cc: pgsql-bugs(at)postgresql(dot)org
Subject: Re: BUG #13472: VACUUM ANALYZE hangs on certain tables
Date: 2015-06-26 13:53:42
Message-ID: 20150626135342.GG4797@alap3.anarazel.de
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-bugs

Hello,m

On 2015-06-26 11:08:30 +0000, matthew(dot)seaman(at)adestra(dot)com wrote:
> We have two tables which we have not been able to vacuum sucessfully on one
> of our database servers. The vacuum process just hangs almost
> instantaneously after issuing the 'VACUUM ANALYZE' command.

Hm. Is there a chance that your application uses cursors?

Could you send the output of:
SELECT oid::regclass, txid_current(), relfrozenxid, age(relfrozenxid)
FROM pg_class
WHERE oid = 'your relation'::regclass;

My theory is that there's a page that needs to be frozen, but a pin on
the page exists, preventing the cleanup lock from being acquired. Or
rather delaying it long enough that little progress is made.

Greetings,

Andres Freund

In response to

Responses

Browse pgsql-bugs by date

  From Date Subject
Next Message Tom Lane 2015-06-26 13:57:05 Re: operator family changes, sinval bug?
Previous Message Marco Nenciarini 2015-06-26 13:50:41 Re: [BUGS] BUG #13473: VACUUM FREEZE mistakenly cancel standby sessions