On Fri, Apr 8, 2011 at 9:28 PM, Alvaro Herrera
<alvherre@commandprompt.com> wrote:
>
> Why is it important to have the original pg_clog files around? Since
> the transactions in question are below the freeze horizon, surely the
> tuples that involve those transaction have all been visited by vacuum
> and thus removed if they were leftover from aborted transactions or
> deleted, no? So you could just fill those files with the 0x55 pattern
> (signalling "all transactions are committed") and the net result should
> be the same. No?
>
> Forgive me if I'm missing something. I haven't been following this
> thread and I'm more than a little tired (but wanted to shoot this today
> because I'm gonna be able to, until Monday).
>
> --
> Álvaro Herrera <alvherre@commandprompt.com>
> The PostgreSQL Company - Command Prompt, Inc.
> PostgreSQL Replication, Consulting, Custom Development, 24x7 support
Update on the status of the steps we took, which were:
- test on a hot standby by bringing it live, running the script,
determing the missing clog files, copying them into the live (hot
standby) pg_clog dir
Now, on the master, copied the same old clog files into the production
*master*, ran vacuumdb -a -v -F. The step I should have taken on the
master before the vacuumdb -F would have been to run the
http://wiki.postgresql.org/wiki/20110408pg_upgrade_fix script to see
if I was missing any clog files on the master.
That vacuum freeze step pointed out a clog file, I copied that into
the master pg_clog dir, ran the aforementioned script. It didn't fail
on any of the clog files this time, so now I am rerunning the vacuum
freeze command and hoping like hell it works!
If the current run of the vacuum freeze fails, I'll report back.
Thanks again for everyone's help.