MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8;
 format=flowed
Date: Thu, 21 May 2015 09:16:17 +0200
From: Jan Lentfer <Jan.Lentfer@web.de>
To: <pgsql-admin@postgresql.org>
Subject: Re: File removed while backup is running
In-Reply-To: 
 <CAG59ottwgmaYWBBszozxDQK4SsUS2iEUz2JhUsPacEvq-CQs1Q@mail.gmail.com>
References: 
 <CAG59ottwgmaYWBBszozxDQK4SsUS2iEUz2JhUsPacEvq-CQs1Q@mail.gmail.com>
Message-ID: <96a0ca326b635cb1121b60605297c453@imap.lan.net>
User-Agent: Roundcube Webmail/0.7.2
Content-Transfer-Encoding: quoted-printable
Precedence: bulk
Sender: pgsql-admin-owner@postgresql.org

Hi,

that is generally not a problem. _fsm  is the "Free Space Map" file=20
(http://www.postgresql.org/docs/9.4/static/storage-fsm.html). It's OK=20
and normal, that files change or even disappear within PGDATA directory=20
during backup - because the Postgres cluster is still operational. Those=20
changes end up in the WAL files and that is why you need those from the=20
start of the backup. This is also described in the manual:

"Some file system backup tools emit warnings or errors if the files=20
they are trying to copy change while the copy proceeds. When taking a=20
base backup of an active database, this situation is normal and not an=20
error. However, you need to ensure that you can distinguish complaints=20
of this sort from real errors. For example, some versions of rsync=20
return a separate exit code for "vanished source files", and you can=20
write a driver script to accept this exit code as a non-error case.=20
Also, some versions of GNU tar return an error code indistinguishable=20
from a fatal error if a file was truncated while tar was copying it.=20
Fortunately, GNU tar versions 1.16 and later exit with 1 if a file was=20
changed during the backup, and 2 for other errors. With GNU tar version=20
1.23 and later, you can use the warning options=20
--warning=3Dno-file-changed --warning=3Dno-file-removed to hide the related=
=20
warning messages."

http://www.postgresql.org/docs/9.4/interactive/continuous-archiving.html#BA=
CKUP-LOWLEVEL-BASE-BACKUP


hth

Jan

Am 2015-05-21 09:04, schrieb Mario Aguado:
> Hi,
>
> I have two postgres cluster in 9.4 version and all run backups
> correctly. Recently I add a new cluster and the backup doesnt work on
> it.=C2=A0
>
> All databases have archive_mode =3D on and wal_level=3Darchive
>
> My backup has this steeps:
> =C2=A0
> SELECT pg_start_backup(${clustername});"
> tar PGDATA and TABLESPACES
> Select pg_stop_backup();
> tar WAL files
>
> With this database when I run backup I get errors like:
>
> t3_28961_fsm: File removed before we read it
>
> I dont know what file type is t3_28961_fsm.
>
> If I go to tablespace dir i have several t files and this file types
> are adding and removing continously.
>
> -rw------- 1 postgres postgres 8.0K May 21 09:01 t3_29094
> -rw------- 1 postgres postgres =C2=A0 =C2=A00 May 21 09:01 t3_29087
> -rw------- 1 postgres postgres 8.0K May 21 09:01 t3_29093
>
> Im looking for this error but i didnt found any.
>
> Any idea?
>
> Thanks in advance.
>
> Regards.
>
> Mario


--=20
Sent via pgsql-admin mailing list (pgsql-admin@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-admin