Re: safety of vacuum verbose analyze on active tables

> Andrew Snow wrote:
> >
> > > > > what is the relative safety of doing a vacuum verbose analyze
> > > on a 24Gb
> > > > > table while there are selects and updates/inserts happening on it?
> > > >
> > > > As far as I know, the table is locked completely during a Vacuum.
Any
> > > > transactions attempting to do inserts/updates will be paused
> > > safely. So go
> > > > ahead and schedule your vacuums for whenever you need to.
> >
> > > There have been reports of problems (corruption, etc.) when trying to
do
> > > this. See the archive for a discussion along these lines a few
> > > months ago.
> >
> > You've got to be joking.
> >
> > Is the table locking mechanism in Postgresql broken??
>
> I have no idea (but I doubt it). I simply recall this question being
asked
> a few months back and a couple folks said something like "Hey, we tried
> this and had problems." That was with 6.5.* or earlier. Maybe one of
> those folks can pipe up again. I couldn't find them in the archive...
>
> Regards,
> Ed Loehr
>

I've had problems with vacuum on postgres 6.5.3
The didn't occur in 6.3.2 or 7.0beta1.

The problems occured after many revoke and grant statements on the database.
And after and/or during that running a vacuum twice. The second would start
saying something like:
NOTICE: Rel pg_class: TID 294/3: InsertTransactionInProgress 20065 - can't
shrink relation
NOTICE: Rel pg_class: TID 294/4: InsertTransactionInProgress 20065 - can't
shrink relation
NOTICE: Index pg_class_relname_index: NUMBER OF INDEX' TUPLES (1615) IS NOT
THE SAME AS HEAP' (1587)
NOTICE: Index pg_class_oid_index: NUMBER OF INDEX' TUPLES (1615) IS NOT THE
SAME AS HEAP' (1587)
VACUUM

But it works in 7.0

Cheers Wim.