I would never sing a SLA for HA if I don't have system, at least, fully redundant in order to deal with these issues.
Single server, single point of failure.
-----Original Message-----
From: Brad Nicholson <
bnichols@ca.afilias.info>
To: Iñigo Martinez Lasala <
imartinez@vectorsf.com>
Cc: Renato Oliveira <
renato.oliveira@grant.co.uk>, pgsql-admin <
pgsql-admin@postgresql.org>
Subject: Re: [ADMIN] Migrate postgres to newer hardware
Date: Tue, 30 Mar 2010 10:38:02 -0400
On Tue, 2010-03-30 at 16:29 +0200, Iñigo Martinez Lasala wrote:
> Hi Renato.
>
> I would follow the ancient method: perform a pg_dump / pg_restore
This is the easiest approach. The problem is that a lot of people have
contractual SLA's that do not allow them to take their systems down long
enough to do a dump and restore upgrade.