Tom Lane wrote:
> Kevin Brown <kevin@sysexperts.com> writes:
> > When a heavy INSERT or UPDATE load on a table is occurring (lots of
> > quick INSERTs or UPDATEs within a single transaction), a VACUUM
> > ANALYZE (or just straight VACUUM) has a really good chance (10% or so)
> > of causing either the INSERT/UPDATE or the VACUUM to fail with a
> > "deadlock detected" error.
>
> I was unable to replicate this in CVS tip, using "pgbench -c 10 -t 1000"
> as the load generator.
I guess I should mention that I was doing VACUUM on the individual
tables, not the entire database. I have no idea if that makes any
difference.
I'll see about compiling and installing CVS tip for testing. I'll
also see about doing some testing on 7.3.2.
Can anyone here independently verify what I've found on 7.2.4? You'll
probably have to run a process that does continuous INSERTs on a
table, and another that does continuous VACUUMs on the same table at
the same time.
--
Kevin Brown kevin@sysexperts.com