On Wed, Apr 14, 2010 at 5:00 PM, raghavendra t
<raagavendra.rao@gmail.com> wrote:
Hi Shoaib,
Tried with pg_resetxlog
[postgres@dbarhel564 bin]$ pg_resetxlog /usr/local/pgsql/mypg/
The database server was not shut down cleanly.
Resetting the transaction log might cause data to be lost.
If you want to proceed anyway, use -f to force reset.
[postgres@dbarhel564 bin]$ pg_resetxlog -f /usr/local/pgsql/mypg/
Transaction log reset
[postgres@dbarhel564 bin]$ pg_ctl start
server starting
[postgres@dbarhel564 bin]$ LOG: database system was shut down at 2010-04-12 12:26:15 IST
FATAL: could not access status of transaction 889
DETAIL: Could not read from file "pg_clog/0000" at offset 0: Success.
LOG: startup process (PID 1335) exited with exit code 1
LOG: aborting startup due to startup process failure
In that case, I would go and see if there is a clog file DB server is complaining for there or not i.e. "pg_clog/0000" and if it is check for permissions with PG user. Do you by any chance deleted stuff from pg_clog folder?
If they are not there then create 256k files for each of them filled with zeros that server complaints for like the we just saw "pg_clog/0000", remember this way you might be able to start the server but you will lose the data from the time after your last checkpoint.
--
Shoaib Mir
http://shoaibmir.wordpress.com/