PostgreSQL 9.2 archiving last replayed WAL after recovery

Hello,

Following the PostgreSQL 9.2 official documentation, I have set up
continuous archiving on my postgres.conf file:

wal_level = archive
archive_mode = on
archive_command = 'test ! -f /.../archive/%f && cp %p /.../archive/%f'

After a successful backup and recovery, the archive command fails when
the database is started because PostgreSQL is trying to archive the
last replayed WAL file, and obviously, it'd already been archived. My
recovery.conf file is the following:

restore_command = 'cp /.../archive/%f "%p"'
recovery_target_time = '2013-07-02 15:20:12'

And this is the error I'm getting when I try to start the database:

LOG: database system was interrupted; last known up at 2013-07-02 14:53:37 CEST
LOG: creating missing WAL directory "pg_xlog/archive_status"
LOG: starting point-in-time recovery to 2013-07-02 15:20:12+02
LOG: restored log file "000000010000000000000002" from archive
LOG: redo starts at 0/2000020
LOG: consistent recovery state reached at 0/20023D0
LOG: restored log file "000000010000000000000003" from archive
LOG: recovery stopping before commit of transaction 4063, time
2013-07-02 15:20:12.211559+02
LOG: redo done at 0/306D8B8
LOG: last completed transaction was at log time 2013-07-02 15:20:11.189978+02
cp: cannot stat `/.../archive/00000002.history': No such file or directory
LOG: selected new timeline ID: 2
cp: cannot stat `/.../archive/00000001.history': No such file or directory
LOG: archive recovery complete
LOG: autovacuum launcher started
LOG: database system is ready to accept connections
LOG: archive command failed with exit code 1
DETAIL: The failed archive command was: test ! -f
/.../archive/000000010000000000000003 && cp
pg_xlog/000000010000000000000003 /.../archive/000000010000000000000003
LOG: archive command failed with exit code 1
DETAIL: The failed archive command was: test ! -f
/.../archive/000000010000000000000003 && cp
pg_xlog/000000010000000000000003 /.../archive/000000010000000000000003
LOG: archive command failed with exit code 1
DETAIL: The failed archive command was: test ! -f
/.../archive/000000010000000000000003 && cp
pg_xlog/000000010000000000000003 /.../archive/000000010000000000000003
WARNING: transaction log file "000000010000000000000003" could not be
archived: too many failures

Is this an expected behaviour? Should the last replayed WAL file be
archived after recovery? If so, the archive command should not fail in
case the file is already archived (as mentioned in the documentation).

Cheers,

Daniel

Dear Daniel Blanco,

I think you copied same postgres.conf from master to standby, right?

if so please remove the following setting from standby before starting it

wal_level = archive
archive_mode = on
archive_command = 'test ! -f /.../archive/%f && cp %p /.../archive/%f'

hope it will work :D

Regards,
Amul Sul

--
View this message in context: http://postgresql.1045698.n5.nabble.com/PostgreSQL-9-2-archiving-last-replayed-WAL-after-recovery-tp5763207p5764173.html
Sent from the PostgreSQL - novice mailing list archive at Nabble.com.

Hi Amul,

No, I didn't copy my postgres.conf from master to standby. This is always
working on a single instance, archiving WAL files to be able to achieve
backup and recovery.

The problem arises after recovery. We go back to a certain snapshot
(created between pg_start_backup and pg_stop_backup as advised) and then
start replaying logs. It reaches a consistent state after the
pg_stop_backup point and all the rest of the logs are replayed, and
recovery is successful. But after that, when the database starts up and
starts recycling WAL files, it always archives the last replayed one,
although that file was already archived and has changed because another
timeline is started.

For the moment, I solved the problem modifying the archive_command to 'test
-f /.../archive/%f || cp %p /.../archive/%f' so if the file exists it
doesn't archive it. But that is not the recommended procedure in the
documentation, where it says that a the archive command should fail if a
fail is tried to be archived twice.

Cheers,

Daniel

On 18 July 2013 04:06, amulsul <sul_amul(at)yahoo(dot)co(dot)in> wrote:

> Dear Daniel Blanco,
>
> I think you copied same postgres.conf from master to standby, right?
>
> if so please remove the following setting from standby before starting it
>
> wal_level = archive
> archive_mode = on
> archive_command = 'test ! -f /.../archive/%f && cp %p /.../archive/%f'
>
>
> hope it will work :D
>
> Regards,
> Amul Sul
>
>
>
> --
> View this message in context:
> http://postgresql.1045698.n5.nabble.com/PostgreSQL-9-2-archiving-last-replayed-WAL-after-recovery-tp5763207p5764173.html
> Sent from the PostgreSQL - novice mailing list archive at Nabble.com.
>
>
> --
> Sent via pgsql-novice mailing list (pgsql-novice(at)postgresql(dot)org)
> To make changes to your subscription:
> http://www.postgresql.org/mailpref/pgsql-novice
>

Sorry, in the second paragraph I meant to say "it always archives the last
replayed one, although that file was already archived and has not changed
because another timeline is started."

Daniel

On 18 July 2013 09:58, Daniel Gomez Blanco <nanodgb(at)gmail(dot)com> wrote:

> Hi Amul,
>
> No, I didn't copy my postgres.conf from master to standby. This is always
> working on a single instance, archiving WAL files to be able to achieve
> backup and recovery.
>
> The problem arises after recovery. We go back to a certain snapshot
> (created between pg_start_backup and pg_stop_backup as advised) and then
> start replaying logs. It reaches a consistent state after the
> pg_stop_backup point and all the rest of the logs are replayed, and
> recovery is successful. But after that, when the database starts up and
> starts recycling WAL files, it always archives the last replayed one,
> although that file was already archived and has changed because another
> timeline is started.
>
> For the moment, I solved the problem modifying the archive_command
> to 'test -f /.../archive/%f || cp %p /.../archive/%f' so if the file exists
> it doesn't archive it. But that is not the recommended procedure in the
> documentation, where it says that a the archive command should fail if a
> fail is tried to be archived twice.
>
> Cheers,
>
> Daniel
>
>
> On 18 July 2013 04:06, amulsul <sul_amul(at)yahoo(dot)co(dot)in> wrote:
>
>> Dear Daniel Blanco,
>>
>> I think you copied same postgres.conf from master to standby, right?
>>
>> if so please remove the following setting from standby before starting it
>>
>> wal_level = archive
>> archive_mode = on
>> archive_command = 'test ! -f /.../archive/%f && cp %p /.../archive/%f'
>>
>>
>> hope it will work :D
>>
>> Regards,
>> Amul Sul
>>
>>
>>
>> --
>> View this message in context:
>> http://postgresql.1045698.n5.nabble.com/PostgreSQL-9-2-archiving-last-replayed-WAL-after-recovery-tp5763207p5764173.html
>> Sent from the PostgreSQL - novice mailing list archive at Nabble.com.
>>
>>
>> --
>> Sent via pgsql-novice mailing list (pgsql-novice(at)postgresql(dot)org)
>> To make changes to your subscription:
>> http://www.postgresql.org/mailpref/pgsql-novice
>>
>
>