Re: WAL recovery failure

From: Fujii Masao <masao(dot)fujii(at)gmail(dot)com>
To: Rick Dicaire <kritek(at)gmail(dot)com>
Cc: pgsql-admin(at)postgresql(dot)org
Subject: Re: WAL recovery failure
Date: 2012-09-26 00:01:52
Message-ID: CAHGQGwEO-Rk1vnVoHZ_Mv+XdHeJYPUL8fQSXwFXt6q2LCnjmfA@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-admin

On Wed, Sep 26, 2012 at 3:22 AM, Rick Dicaire <kritek(at)gmail(dot)com> wrote:
> Hi folks, we're snapshotting at san level, two volumes containing a
> postgres 9.1.5 database, and pg_xlog respectively.
> This database cluster is a streaming replicant.
> We're stopping the database completely, taking the snapshots,
> restarting the database.
> I export these snapshots as luns to a VM, mount the disks in the VM,
> cp -a the database and pg_xlog contents to
> a staging area. Start postgres using -D /staging/9.1/main. Postgres
> fails to start.
>
> From log:
>
> 2012-09-25 12:41:30.147 CDT,,,9597,,5061ecca.257d,1,,2012-09-25
> 12:41:30 CDT,,0,LOG,00000,"database system was interrupted while in
> recovery at 2012-09-25 12:18:06 CDT",,"This probably means that some
> data is corrupted and you will have to use the last backup for
> recovery.",,,,,,,""
> 2012-09-25 12:41:30.155 CDT,,,9597,,5061ecca.257d,2,,2012-09-25
> 12:41:30 CDT,,0,LOG,00000,"starting archive recovery",,,,,,,,,""
> 2012-09-25 12:41:30.354 CDT,,,9597,,5061ecca.257d,3,,2012-09-25
> 12:41:30 CDT,,0,LOG,00000,"restored log file
> ""000000010000074D000000EE"" from archive",,,,,,,,,""
> 2012-09-25 12:41:30.529 CDT,,,9597,,5061ecca.257d,4,,2012-09-25
> 12:41:30 CDT,,0,LOG,00000,"restored log file
> ""000000010000074D000000E2"" from archive",,,,,,,,,""
> 2012-09-25 12:41:30.529 CDT,,,9597,,5061ecca.257d,5,,2012-09-25
> 12:41:30 CDT,,0,PANIC,XX000,"unexpected pageaddr 74D/7C000000 in log
> file 1869, segment 226, offset 0",,,,,,,,,""
> 2012-09-25 12:41:30.530 CDT,,,9595,,5061ecc8.257b,1,,2012-09-25
> 12:41:28 CDT,,0,LOG,00000,"startup process (PID 9597) was terminated
> by signal 6: Aborted",,,,,,,,,""
> 2012-09-25 12:41:30.530 CDT,,,9595,,5061ecc8.257b,2,,2012-09-25
> 12:41:28 CDT,,0,LOG,00000,"terminating any other active server
> processes",,,,,,,,,""
>
> /staging/9.1/main/recovery.conf:
>
> restore_command = 'cp /staging/9.1/main/pg_xlog/%f "%p"'

/staging/9.1/main/pg_xlog is really archive directory? restore_command
should be set to the command which copies the archived WAL files from
the archive area.

Regards,

--
Fujii Masao

In response to

Responses

Browse pgsql-admin by date

  From Date Subject
Next Message Rick Dicaire 2012-09-26 01:56:21 Re: WAL recovery failure
Previous Message Devrim GÜNDÜZ 2012-09-25 21:44:46 Re: Upgrading and streaming replication