On 5/9/07, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> Jim Nasby <decibel(at)decibel(dot)org> writes:
> > Any time this happens it's generally a nasty surprise for users.
> Really? Running out of work memory is expected on large tables.
Sure. Perhaps we should find a better error message but it's an
interesting information. Personnaly, I try to choose a sane value
depending on my database but I'm never sure it's really sufficient or
if I added 100MB it would have made a real difference.
> > It would be nice to throw them an explicit warning that it's occurring.
> I think this is a bad idea. It's furthermore pretty useless in the
> autovacuum world, since no one is likely to see the warning.
IMHO we're far from having everyone using autovacuum. For instance,
for most of our customers, we prefer having a window for vacuuming
(from 3am for example) instead of having autovacuum fired in the
middle of the day during a load peak.
If we can shorten the window by having a sufficient value for
maintenance_work_mem, it's even nicer and Jim's patch could help us
with this point.
In response to
pgsql-patches by date
|Next:||From: Tomas Doran||Date: 2007-05-10 01:59:56|
|Subject: Re: Implemented current_query|
|Previous:||From: Jim Nasby||Date: 2007-05-09 23:18:55|
|Subject: Re: Have vacuum emit a warning when it runs out of maintenance_work_mem |