Re: Perl DBD and an alarming problem

[Please copy the mailing list on replies.]

On Thu, Nov 17, 2005 at 05:38:13PM -0800, Craig A. James wrote:
> >You probably shouldn't set statement_timeout on a global basis
> >anyway
>
> The server is a "one trick pony" so setting a global timeout value is
> actually appropriate.

Beware that statement_timeout also applies to maintenance commands
like VACUUM; it might be more appropriate to set per-user timeouts
with ALTER USER. If you do set a global timeout then you might
want to set a per-user timeout of 0 for database superusers so
maintenance activities don't get timed out.

> >... but did you reload the server after you made the change?
> >Setting statement_timeout in postgresql.conf and then reloading the
> >server works here in 8.0.4.
>
> Yes. By "reload" I assume you mean restarting it from scratch.

Either a restart or a "pg_ctl reload", which sends a SIGHUP to the
server. You can effect some changes by sending a signal to a running
server without having to restart it entirely.

> In this case, I use
>
> /etc/init.d/postgresql restart
>
> It definitely had no effect at all. I tried values clear down to 1
> millisecond, but the server never timed out for any query.

Did you use "SHOW statement_timeout" to see if the value was set
to what you wanted? Are you sure you edited the right file? As a
database superuser execute "SHOW config_file" to see what file the
server is using. What exactly did the line look like after you
changed it?

--
Michael Fuhr