Re: Pg_xlog increase due to postgres crash (disk full)

[Please keep the list copied, and put your reply in-line instead
of at the top.]

Cliff de Carteret wrote:
> On 22 January 2013 16:07, Kevin Grittner <kgrittn@mail.com> wrote:
>
>> Cliff de Carteret wrote:
>>
>>> The current setup has been working successfully for several years
>>> until the recent database crash
>>
>> What file does the server log say it is trying to archive? What
>> error are you getting? Does that filename already exist on the
>> archive (or some intermediate location used by the archive command
>> or script)?

> The sever log is (repeated constantly):
>
> LOG: archive command failed with exit code 1
> DETAIL: The failed archive command was: test ! -f
> /opt/postgres/remote_pgsql/wal_archive/00000001000000A800000078 && cp
> pg_xlog/00000001000000A800000078
> /opt/postgres/remote_pgsql/wal_archive/00000001000000A800000078
> WARNING: transaction log file "00000001000000A800000078" could not be
> archived: too many failures
>
> The file 00000001000000A800000078 exists in the remote archive's
> wal_archive directory. I read a post saying to copy the file over to the
> archive and then delete the .ready file to get postgres to move onto the
> next file but this ended up logging out saying that a log file was missing.
> There are more recent files in this directory but they end at the point
> where I reverted all of the changes I made last night when time was running
> out and the database had to be put back to a known state.

I would have deleted (or renamed) the copy in the archive
directory. Archiving should have then resumed and cleaned up the
pg_xlog directory.

-Kevin