Re: Why I lost the last pg_xlog file?

From: Shaun Thomas <sthomas(at)peak6(dot)com>
To: Waldomiro <waldomiro(at)shx(dot)com(dot)br>
Cc: "pgsql-performance(at)postgresql(dot)org" <pgsql-performance(at)postgresql(dot)org>
Subject: Re: Why I lost the last pg_xlog file?
Date: 2011-01-27 14:17:23
Message-ID: 4D417E73.7040901@peak6.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

On 01/27/2011 06:10 AM, Waldomiro wrote:

> 3) I check the pg_xlog, there was created 6 logs, but in the archive
> there was only 5.

xlogs are only "archived" when they'd normally be deleted. If you have
really high data turnover or very frequent checkpoints, that effectively
happens constantly. I'm not sure where the cutoff is, but there's a
certain amount of "reserve" xlog space based on your checkpoint_segments
setting.

It is an archive, after all. pg_start_backup/pg_stop_backup ensure your
backup is consistent, nothing else. Depending on your archives to
capture everything isn't going to work. If you really want everything,
you can either copy the xlogs manually (not safe) or initiate another
backup. You can kinda fudge it doing this:

1. Call pg_current_xlog_location to get the current xlog.
2. Call pg_switch_xlog to force a new xlog.
3. Copy the file from step 1 and anything older than it to your
archive/slave. Doing this *may* confuse the built-in archive system if
your archive_command is too strict.
4. Profit.

--
Shaun Thomas
OptionsHouse | 141 W. Jackson Blvd. | Suite 800 | Chicago IL, 60604
312-676-8870
sthomas(at)peak6(dot)com

______________________________________________

See http://www.peak6.com/email_disclaimer.php
for terms and conditions related to this email

In response to

Browse pgsql-performance by date

  From Date Subject
Next Message Tom Lane 2011-01-27 14:18:11 Re: Real vs Int performance
Previous Message David Greco 2011-01-27 13:48:00 Re: Real vs Int performance