Re: Start up error

Hi Tom,

Thank you for your time.

Below are the answers for your queries.

1)I am using /etc/rc.d/init.d/postgresql stop/start to stop and start
the postgresql service.

2)There are several postgresql process running after a failed
restart.Below are the proccesess

ps -ef|grep post
root 730 30758 0 12:22 pts/2 00:00:00 grep post
postgres 2700 15927 73 12:00 ? 00:16:15 postgres: cscti csctools
[local] DELETE
postgres 6706 15927 72 12:00 ? 00:15:42 postgres: cscti csctools
[local] DELETE
postgres 14351 15927 69 12:02 ? 00:13:38 postgres: cscti csctools
[local] DELETE
postgres 15927 1 0 11:58 ? 00:00:05 /usr/bin/postmaster -p
5432 -D /var/lib/pgsql/data
postgres 15936 15927 0 11:58 ? 00:00:00 postgres: logger process

postgres 15967 15927 0 11:58 ? 00:00:00 postgres: writer process

postgres 15968 15927 0 11:58 ? 00:00:00 postgres: stats buffer
process
postgres 15969 15968 0 11:58 ? 00:00:00 postgres: stats
collector process
postgres 16118 15927 73 11:58 ? 00:17:23 postgres: cscti csctools
[local] DELETE
postgres 23600 15927 65 12:08 ? 00:09:25 postgres: cscti csctools
[local] DELETE

3)There is no process running with PID 25372.

I have been asked to reboot the server to fix this, what precautions
should be taken before reboot to get the postgresql DB working after
reboot.

Is really reboot will fix the issue.Could you please suggest.

Thanks & Regards,

-Jawad.

-----Original Message-----
From: Tom Lane [mailto:tgl(at)sss(dot)pgh(dot)pa(dot)us]
Sent: Saturday, November 17, 2007 9:48 PM
To: Hussain Jawad-FXRM43
Cc: pgsql-bugs(at)postgresql(dot)org
Subject: Re: [BUGS] Start up error

"Hussain Jawad-FXRM43" <FXRM43(at)motorola(dot)com> writes:
> When I am trying to stop and start the DB, below error is reported in
> postgresql start up log file.

Exactly what are you doing to start and stop the server? If you are
using a startup script, whose is it?

> I have deleted the file postmaster.pid in the directory
> /var/lib/pgsql/data for number of times and restarted the postgresql
> service, but I am still not able to restart the server and the same
> error is repeating again.

The best theory that comes to mind is that your start procedure is
somehow starting multiple copies of the postmaster. The first one
starts OK and then the second (and third?) ones fail with the lockfile
complaint --- as well they should.

I'm a bit afraid that your manual removals of the lockfile (which is A
Bad Idea as a rule) have left you with multiple versions of the
postmaster running sans lockfile. This would be very bad because you
can easily end up with a corrupted database. Have you looked around
with "ps" to verify that there really aren't any postgres-owned
processes left over after a "failed" start?

In particular, you should absolutely not see any lockfile complaints
unless the PID mentioned in the message is a live process. What is PID
25372 and what is it doing?

regards, tom lane