DB crash SOS - Mailing list pgsql-hackers

From Felde Norbert
Subject DB crash SOS
Date
Msg-id AANLkTiniQ8TGeRnJECR3SDPijuNYtvgNujTAiJnoR61A@mail.gmail.com
Whole thread Raw
Responses Re: DB crash SOS  (Robert Haas <robertmhaas@gmail.com>)
Re: DB crash SOS  (Florian Pflug <fgp@phlo.org>)
List pgsql-hackers
Hi all,

I use 8.2 on a windows server 2008.
Suddenly postgres crashed and I can not do anything.
Strange things happend to postgres in the last few weeks. Once, there
was so many postgres process, that I could not connect to it with
pgAdmin3. It said that too many connections and I had to restart
postgres. Than, after a while the postgres service stopped but some
process ran and it was still accsessable so the application which used
it ran without any problem and I could simply start postgres again.

Now something is wrong, I can not make backup and the vacuum stops too.
The message is
pg_dump: Error message from server: ERROR:  could not access status of
transaction 3974799
DETAIL:  Could not read from file "pg_clog/0003" at offset 204800: No error.
pg_dump: The command was: COPY public.active_sessions_split (ct_sid,
ct_name, ct_pos, ct_val, ct_changed) TO stdout;
pg_dump: *** aborted because of error

The only one thing I found to correct this is to create a file filled
with binary 0 and replace clog/0003. Both files size was the same but
I get that errormessage again.

I tried even with a bigger empty clog/0003 file but than I get that:
pg_dump: Error message from server: ERROR:  xlog flush request
0/A19F5BF8 is not satisfied --- flushed only to 0/A02A1AC8
CONTEXT:  writing block 1149 of relation 1663/4192208/4192508
pg_dump: The command was: COPY public.history (historyid, adatkod,
elemid, userid, ido, actionid, targyid, szuloid, opvalue, longfield,
longtext) TO stdout;
pg_dump: *** aborted because of error

I tried to drop the last few transaction with pg_resetxlog and hoped
to save some data but there is again the original error message.

I checked the permissions of the whole data dir. The owner of it is
postgres and has full permission.


Can anyone suggest something?
Many data would be lost if I can not repaire that so please!

Thanks,
fenor


pgsql-hackers by date:

Previous
From: Magnus Hagander
Date:
Subject: Re: ANNOUNCE list (was Re: New PGXN Extension site)
Next
From: Jean-Baptiste Quenot
Date:
Subject: pg_dump does not honor namespaces when functions are used in index