Re: Specific questions about wraparound and vacuum

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: "Nick Fankhauser" <nickf(at)doxpop(dot)com>
Cc: pgsql-admin(at)postgresql(dot)org
Subject: Re: Specific questions about wraparound and vacuum
Date: 2007-08-08 16:34:04
Message-ID: 2643.1186590844@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-admin

"Nick Fankhauser" <nickf(at)doxpop(dot)com> writes:
> 1) Do we need to do a *full* vacuum or just a "lazy" vacuum to take care of
> the xid wraparound issues?

Lazy is sufficient --- as long as it's database-wide including the
system catalogs (which means a superuser has to do it).

> Or... is it the case that by one billion, the docs really mean 2^30
> (1073741824) ?

2^30. You appear to have executed a bit under 600000 transactions since
vacuuming, so if that's three hours then you've got about 5000 hours
until wraparound.

> 4) The documentation implies that I need to vacuum *every* database, even if
> I'm not using them all, so I vacuumed template1, but can't connect to
> template0. Do I really need to vacuum databases where no transactions are
> occurring? If so, how to I deal with template0?

You don't need to touch frozen databases. This was discussed yesterday ...

regards, tom lane

In response to

Browse pgsql-admin by date

  From Date Subject
Next Message Andreas 2007-08-08 18:00:34 Re: ssl and odbc standar driver
Previous Message Stefan Zweig 2007-08-08 16:25:58 use replication on a database only on selected schemes?