From: | Simon Riggs <simon(at)2ndquadrant(dot)com> |
---|---|
To: | Fujii Masao <masao(dot)fujii(at)gmail(dot)com> |
Cc: | pgsql-hackers(at)postgresql(dot)org |
Subject: | Re: For what should pg_stop_backup wait? |
Date: | 2008-08-07 13:59:47 |
Message-ID: | 1218117588.4549.497.camel@ebony.2ndQuadrant |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On Thu, 2008-08-07 at 21:11 +0900, Fujii Masao wrote:
> In HEAD, pg_stop_backup waits until the history file has been archived.
> But, in order to ensure that the last wal file was archived, pg_stop_backup
> should wait until not only the history file but also the backup stopping wal
> file has been archived, I think.
>
> Because the alphabetic order of the history file and the backup stopping
> wal file is not constant.
> If the backup starting wal file is the same as the stopping one (that is,
> any wal files were not switched during backup), the history file whose
> name consists of the starting one is behind the stopping one.
> Otherwise, the backup stopping wal file is behind the history file.
pg_stop_backup explicitly requests a log switch prior to writing the
history file, so this occurs in all cases. The assumption is wrong.
So yes, we should wait for stopxlogfilename, not histfilepath.
I'll do a patch. Thanks for your input.
--
Simon Riggs www.2ndQuadrant.com
PostgreSQL Training, Services and Support
From | Date | Subject | |
---|---|---|---|
Next Message | Kevin Grittner | 2008-08-07 14:14:50 | Re: Visibility Groups |
Previous Message | Marko Kreen | 2008-08-07 13:53:49 | [patch] fix dblink security hole |