From: | Jeni Fifrick <jfifrick(at)incomm(dot)com> |
---|---|
To: | Gabriele Bartolini <Gabriele(dot)Bartolini(at)2ndQuadrant(dot)it> |
Cc: | "pgsql-admin(at)postgresql(dot)org" <pgsql-admin(at)postgresql(dot)org> |
Subject: | Re: Suggestion on Backup |
Date: | 2013-10-16 16:05:05 |
Message-ID: | E59C57F2AE0F8641B5B428292D88479F26D1C3CA@ATLPXCH01MB.uss.net |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-admin |
Thank you Gabriele & all for your suggestions.
I'm going to review them.
-----Original Message-----
From: Gabriele Bartolini [mailto:Gabriele(dot)Bartolini(at)2ndQuadrant(dot)it]
Sent: Wednesday, October 16, 2013 2:51 AM
To: Jeni Fifrick
Cc: pgsql-admin(at)postgresql(dot)org
Subject: Re: [ADMIN] Suggestion on Backup
Hi Jen,
On Tue, 15 Oct 2013 21:09:50 +0000, Jeni Fifrick <jfifrick(at)incomm(dot)com>
wrote:
> We're having Master & Slave Postgres 9.1.3 on CentOS - using Streaming
> Replication, which serve as a read-only database, and as 'backup' for
> the production database.
Sure.
> I'm trying to setup a process that can be used to have the continuous
> backup of the production, to make sure if some problems happen on the
> production server, we won't loss the data.
It is important that you understand though that such an architecture (with just replication) won't completely protect you from disasters happening on the master. I am referring to data loss caused by undeliberate DELETE or DROP operations, for example, or SQL injections.
Especially if you have a streaming replicated standby, data that disappears on the master will disappear from the standby immediately after.
Physical backup and continuous archiving will add this kind of protection to your system. Using 9.1 unfortunately you can't benefit from cascading replication using the streaming protocol (introduced in 9.2), however you can look into the pg_basebackup utility for this purpose.
Otherwise, there are tools out there that help you achieving these goals, such as OmniPITR (https://github.com/omniti-labs/omnipitr)
pg-rman (http://code.google.com/p/pg-rman/) WAL-E (https://github.com/heroku/WAL-E, mainly for Amazon AWS users) and barman (http://www.pgbarman.org/) - to name a few.
Being one of the developers, I can definitely recommend Barman. It does not yet support streaming replication, but usually with WAL file shipping you can still achieve very good results in terms of Recovery Point Objective (which measures data loss). It is very easy to setup, and already has packages for CentOS.
I hope this helps.
Cheers,
Gabriele
--
Gabriele Bartolini - 2ndQuadrant Italia
PostgreSQL Training, Services and Support
Gabriele(dot)Bartolini(at)2ndQuadrant(dot)it - www.2ndQuadrant.it
From | Date | Subject | |
---|---|---|---|
Next Message | Oscar Calderon | 2013-10-17 15:48:57 | Strange results using pg_restore |
Previous Message | Gabriele Bartolini | 2013-10-16 06:51:15 | Re: Suggestion on Backup |