Re: backup.old

From: Scott Mead <scottm(at)openscg(dot)com>
To: "David G(dot) Johnston" <david(dot)g(dot)johnston(at)gmail(dot)com>
Cc: Steve Pribyl <Steve(dot)Pribyl(at)akunacapital(dot)com>, pgsql-general(at)postgresql(dot)org
Subject: Re: backup.old
Date: 2015-10-07 19:46:28
Message-ID: 1444247189898-d0739f55-ca22542a-4fa76335@openscg.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

On Wed, Oct 7, 2015 at 15:38, David G. Johnston
<david(dot)g(dot)johnston(at)gmail(dot)com>
wrote:
On Wed, Oct 7, 2015 at 3:29 PM, Steve Pribyl <
Steve(dot)Pribyl(at)akunacapital(dot)com [Steve(dot)Pribyl(at)akunacapital(dot)com] > wrote:
Thank you very much. I read someplace if you run pg_start_backup twice the
backup.old will be created, but there was not much beyond that and now I
can't
seem to find the reference.

backup_label gets deleted on pg_stop_backup() on the *master*.
Backup_label will still be in the *backup* itself however (or, more
succinctly,
a slave server). When you start the backup / slave, it will process
backup_label
so that it can start recovery. Once we don't need it anymore, the file is
renamed to backup_label.old. Typically, when you see a backup_label.old on
a
writable master, it was either: * a backup that was restored and put in to
service * a slave server that was promoted A pg_controldata will probably
show a timeline != 1

Scanning the docs and logic tells me that attempting to do pg_start_backup
twice
in a row should result in the second attempt giving an error...but I could
be
misinformed.
The file pg_start_backup creates is named "backup_label" and so I'd also
expect
any attempt to add an old suffix would keep the same base name...
David J.

In response to

Browse pgsql-general by date

  From Date Subject
Next Message droberts 2015-10-07 20:02:56 Re: Best practices for aggregate table design
Previous Message Tom Lane 2015-10-07 19:45:26 Re: How to drop user if objects depend on it