Sure.
I will attempt this on our development server which only occasionally has
this problem. If that won't reproduce it, I will attempt this on our
production server but that must be scheduled at a slow time - maybe Monday.
I'll send this info. once I've collected it.
Thanks for the help,
L.
On Wed, 12 Feb 2003, Tom Lane wrote:
> Laurette Cisneros <laurette@nextbus.com> writes:
> > When I do a shutdown:
> > $pgpath/bin/pg_ctl -D $datadir -m fast stop
> > the server will not completely shutdown.
>
> Hm. -m fast tells the postmaster to send SIGTERM to all the existing
> backends, and then quit when all the backends have quit. So your
> problem is that one or more backend processes isn't exiting in a timely
> fashion after receiving SIGTERM. It would be useful to see what these
> backends are doing. Can you attach to each one with gdb and get a stack
> trace?
>
> regards, tom lane
>
--
Laurette Cisneros, L.D.
The Database Group
(510) 420-3137
NextBus Information Systems, Inc.
www.nextbus.com
----------------------------------
"No man is wise enough by himself"
-- Titus Maccius Plautus
(254 Bc - 184 BC), Miles Gloriosus