Transaction ID wraparound is a cluster issue, not an individual database
issue. Due to the way PostgreSQL is designed, you need to vacuum ALL
your databases, but you don't need a FULL vacuum on them all, just a
regular vacuum.
I'm guessing that your other databases aren't real big anyway, so it
shouldn't be a big problem. If your other databases are huge, then
yeah, it might take a while.
On Wed, 2006-09-06 at 12:39, Sriram Dandapani wrote:
> I started this about a few hours ago (I guess the message shows as a
> general warning)..I am only interested in the specific database..will
> this command NOT do a full vacuum of the specific database(I would like
> to save the few hours that I invested in this vacuum command if
> possible)
>
> -----Original Message-----
> From: Scott Marlowe [mailto:smarlowe@g2switchworks.com]
> Sent: Wednesday, September 06, 2006 10:35 AM
> To: Sriram Dandapani
> Cc: Tom Lane; pgsql-admin@postgresql.org
> Subject: Re: [ADMIN] transactoin id wraparound problem
>
> Change the -f to -a
>
> On Wed, 2006-09-06 at 11:02, Sriram Dandapani wrote:
> > I created a super user using the createuser command and issued
> >
> > vacuumdb -f -U superuser <database-name>
> >
> > I still keep getting a decreasing transaction count warning. Am I
> doing
> > something wrong here.(The database is about 120G and while I do expect
> > vacuum full to take time, I expect the warning to show an increasing
> > count)