VACUUM question, OID range

From: bsmith(at)h-e(dot)com (Bob Smith, Hammett & Edison, Inc(dot))
To: pgsql-admin(at)postgresql(dot)org
Subject: VACUUM question, OID range
Date: 2002-01-01 01:22:38
Message-ID: 000001447373092692958@h-e.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-admin


I have a db which has many tables that are mirrors of data from an
outside source. I download text export files and do weekly updates
of the entire tables using TRUNCATE and COPY. Is there any benefit
to a VACUUM ANALYZE immediately afterwards? There are a number of
indices on these tables. Do I ever benefit from VACUUM at all,
considering that the server never inserts/deletes rows from these
tables except during the update, which replaces every row?

Also, what happens if the OID value "rolls over"? Each of these
updates consumes about 3,000,000 OIDs, which is not a problem for
weekly updates, but I've been asked to consider doing daily updates.
Assuming the OID range is 2^31, with daily updates the OID "rolls
over" in less than 2 years (maybe sooner, as there are several other
databases on the same server as well). Nothing to panic about,
true, but I'd still like to know if I might eventually have to re-
build the whole site due to running out of OIDs.

The server is a P3-800, 768MB RAM, 80GB disk, running Red Hat 7.0
and PostgreSQL 7.0.2.

Thanks!

Bob Smith
Hammett & Edison, Inc.
bsmith(at)h-e(dot)com

Browse pgsql-admin by date

  From Date Subject
Next Message Shamik Majumder 2002-01-01 11:51:12 Need help
Previous Message Tony Reina 2001-12-31 19:05:15 Re: Allowing TCP/IP connections