pg.ddx.io  pgsql-hackers@postgresql.org mailing list archive  

help / color / mirror / Atom feed
From: Heikki Linnakangas <heikki.linnakangas@enterprisedb.com>
To: Florian Pflug <fgp@phlo.org>
Cc: Simon Riggs <simon@2ndQuadrant.com>
Cc: Aidan Van Dyk <aidan@highrise.ca>
Cc: Daniel Farina <daniel@heroku.com>
Cc: Chris Redekop <chris@replicon.com>
Cc: PostgreSQL-development <pgsql-hackers@postgresql.org>
Subject: Re: Hot Backup with rsync fails at pg_clog if under load
Date: Thu, 27 Oct 2011 11:18:06 +0300
Message-ID: <4EA913BE.2080102@enterprisedb.com> (raw)
In-Reply-To: <4EA900E5.9070905@enterprisedb.com>
References: <CAC2SuRJTtsXL67EuTdd25GThYZBZTcepfG7Z0Vvv2pZQWzJD=w@mail.gmail.com>
	<CAAZKuFbUSxOb8x80k9-fD9CsXBRmzTNWutqcDVGaTvwFTvURiQ@mail.gmail.com>
	<213AEFBF-479C-4F9E-B0CA-24AAB2C9F4DD@phlo.org>
	<CA+U5nMJksETN=sJEmu2FqhWNv2Uv63e_9T5OrZEGmviNnWEC5Q@mail.gmail.com>
	<CA+U5nMJfnaay7p_rRzrvWLTyO-aKNEh414czr3jyXT8vhHiJVw@mail.gmail.com>
	<1A94CE1A-C3E5-4923-B677-623D41BC273F@phlo.org>
	<CA+U5nMKUUoA8kRG=itfsO5nZuE3x_KDJz78EaUN3_FKmq-uMJA@mail.gmail.com>
	<6C3D7EDA-573E-46DC-9047-5FEB92876DA8@phlo.org>
	<CA+U5nMJLDZ7VP1O3Jdde+p+9-HxuEi8qG+v5+LuoZGSiOvhkxA@mail.gmail.com>
	<CAC_2qU_eCUYMY6O5CgAK0pYugd=moApirdAt8G_G2sYKs0DfOQ@mail.gmail.com>
	<CA+U5nML+fJgz-NdZrGX+R3m6JKnT1JC=v0iXkkSAsojnsqqPoQ@mail.gmail.com>
	<FD57812A-6A9B-44F3-8CDF-3CD2F8D9918E@phlo.org>
	<12E895D5-C6D0-4E71-B9C3-85E1BC14E6B8@phlo.org>
	<CA+U5nMLyW=S1DTBhswOirf_EZX06epvtyUZN364BGPJBDH _tiA@mail.gmail.com>
	<5EDA27FA-963F-413D-9849-874072F317E5@phlo.org>
	<4EA900E5.9070905@enterprisedb.com>

On 27.10.2011 09:57, Heikki Linnakangas wrote:
>> My suggestion is to fix the CLOG problem in that same way that you fixed
>> the SUBTRANS problem, i.e. by moving LogStandbySnapshot() to before
>> CheckPointGuts().
>>
>> Here's what I image CreateCheckPoint() should look like:
>>
>> 1) LogStandbySnapshot() and fill out oldestActiveXid
>> 2) Fill out REDO
>> 3) Wait for concurrent commits
>> 4) Fill out nextXid and the other fields
>> 5) CheckPointGuts()
>> 6) Rest
>>
>> It's then no longer necessary for LogStandbySnapshot() do modify
>> the nextXid, since we fill out nextXid after LogStandbySnapshot() and
>> will thus derive a higher value than LogStandbySnapshot() would have.
>
> Hmm, I don't think that fully fixes the problem. Even if you're certain
> that CheckPointGuts() has fsync'd the clog page to disk, VACUUM might
> decide to truncate it away again while the checkpoint is running.

Oh, scratch that. During recovery, we merrily treat missing slru files 
as they were filled with zeros.

-- 
   Heikki Linnakangas
   EnterpriseDB   http://www.enterprisedb.com



view thread (80+ messages)  latest in thread

Message-ID: <4EA913BE.2080102@enterprisedb.com>
Permalink:  ../4EA913BE.2080102@enterprisedb.com/
Also on:    postgresql.org/message-id/4EA913BE.2080102@enterprisedb.com

 · 

reply
Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Reply to all the recipients using the --to and --cc options:
  reply via email

  To: pgsql-hackers@postgresql.org
  Cc: heikki.linnakangas@enterprisedb.com, fgp@phlo.org, simon@2ndQuadrant.com, aidan@highrise.ca, daniel@heroku.com, chris@replicon.com
  Subject: Re: Hot Backup with rsync fails at pg_clog if under load
  In-Reply-To: <4EA913BE.2080102@enterprisedb.com>

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox