| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> | 
|---|---|
| To: | "Casey T(dot) Deccio" <casey(at)deccio(dot)net> | 
| Cc: | pgsql-general(at)postgresql(dot)org | 
| Subject: | Re: Transaction error | 
| Date: | 2005-09-01 17:11:58 | 
| Message-ID: | 3178.1125594718@sss.pgh.pa.us | 
| Views: | Whole Thread | Raw Message | Download mbox | Resend email | 
| Thread: | |
| Lists: | pgsql-general | 
"Casey T. Deccio" <casey(at)deccio(dot)net> writes:
> On Thu, 2005-09-01 at 11:59 -0400, Tom Lane wrote:
>> Define "didn't work" ... what happened, exactly, when you tried it?
> Sorry, I was too tired last night to try it again and post the error.
> Here it is:
> ERROR:  could not access status of transaction 4114057
> DETAIL:  could not read from file "/var/lib/postgres/data/pg_clog/0003"
> at offset 237568: Success
Apparently your backup wasn't very current, and contains a version of
that clog file that hasn't yet reached full size.  Oh well, I was hoping
you'd get valid data from the backup.
> How would I create a dummy clog segment file, if needed?
What you'll have to do is pad out the segment file to full size (256K)
by appending zeroes.  (dd from /dev/zero is useful for this.)
Note that you will probably not get a completely consistent database out
of this --- there are going to be some transactions that appear
uncommitted or partially committed.  I'd recommend a dump and reload to
help you revalidate your data, plus you'll have to apply any manual
consistency checks you can think of.
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | marcelo Cortez | 2005-09-01 17:13:36 | lock problem | 
| Previous Message | marcelo Cortez | 2005-09-01 17:02:48 | Re: lock problem |