On Tuesday 01 March 2005 12:46, Tom Lane wrote:
> Bruce Momjian <pgman@candle.pha.pa.us> writes:
> > Tom Lane wrote:
> >> This has to be coming from exec_simple_query():
> >
> > I wonder if some statistics were turned on at postmaster start and even
> > though it was turned off and 'pg_ctl reload' was done the
> > checkpoint/bgwriter process isn't reloading for some reason.
>
> The checkpoint/bgwriter never executes exec_simple_query(). I think
> the 5-minute cycle is driven by something in Robert's application,
> rather than being tied to checkpoints.
>
Actually looking at it now, it happens every 10 minutes, so you're right in
that it doesn't coincide with the bgwriter. Actually I am pretty sure this
is slony related... the output is always on the same pid, and everything else
output on that pid is slony commands. Is there some way I could see what
that pid thinks the GUC variables are set to? I have a wild theory that
slony is preventing the modification of SUSET level variables, but haven't
really got anything solid to back that up.
--
Robert Treat
Build A Brighter Lamp :: Linux Apache {middleware} PostgreSQL