Re: Warm Standby - log shipping

On Fri, 2008-12-19 at 09:14 -0500, Mark Steben wrote:
> What I'm hearing is that I have to perform a base backup on my master in
> Mass. after recovery completes, send that over a secure network
> To Virginia, and lay it down there. Simple enough but the time to travel
> Over the network becomes an issue - 12 - 13 hours at best.
> If we have to do this then we will. I just want to make sure I'm
> understanding your advice.

Yes.

Joshua D. Drake

>
> Thanks
>
> Mark Steben│Database Administrator│
> @utoRevenue-R- "Join the Revenue-tion"
> COME SEE US AT NADA BOOTH #1021, HALL B, IN NEW ORLEANS!
> 95 Ashley Ave. West Springfield, MA., 01089
> 413-243-4800 x1512 (Phone) │ 413-732-1824 (Fax)
> @utoRevenue is a registered trademark and a division of Dominion Enterprises
>
>
>
> -----Original Message-----
> From: Kevin Grittner [mailto:Kevin(dot)Grittner(at)wicourts(dot)gov]
> Sent: Thursday, December 18, 2008 5:20 PM
> To: Mark Steben; pgsql-admin(at)postgresql(dot)org
> Subject: Re: Warm Standby - log shipping
>
> >>> "Mark Steben" <msteben(at)autorevenue(dot)com> wrote:
>
> > We are at postgresql 8.2.5
>
> You really should update to 8.2.11 or consider going to 8.3.5.
>
> http://www.postgresql.org/support/versioning
>
> > We plan on using the Norfolk server not so much as a recovery
> failover but
> > as a replicated database
> > To run reports and establish a data warehousing environment. As such
> the
> > plan is to run the
> > Standby in recovery state for the majority of the day, then
> 'complete'
> > recovery there, bring it
> > Online, perform our reporting and data warehousing functions (in
> read-only
> > mode of course),
> > Then bring it back into recovery mode, letting the updates catch up
> for the
> > next days processing.
> >
> > My questions are:
> >
> > 1. Is this a proper usage of log shipping?
>
> Once you complete recovery you don't have a good way back, short of
> getting a new base backup. If you have space, you could stop the
> replica server without leaving recovery state, do a copy of the
> cluster to another directory, restart your replica, start the copied
> cluster's server, complete recovery, and start running your reports.
>
> > 3. I am currently in a state where a log got partially copied and
> postgres
> > cannot find a valid checkpoint to restart. What is the best way to
> remedy
> > this situation? Pg_resetxlog perhaps?
>
> I'd get a fresh base backup from which to start.
>
> -Kevin
>
>
--
PostgreSQL
Consulting, Development, Support, Training
503-667-4564 - http://www.commandprompt.com/
The PostgreSQL Company, serving since 1997