Received: from localhost (unknown [200.46.204.183]) by postgresql.org (Postfix) with ESMTP id AF20D650EA0 for ; Mon, 30 Jun 2008 09:52:59 -0300 (ADT) Received: from postgresql.org ([200.46.204.86]) by localhost (mx1.hub.org [200.46.204.183]) (amavisd-maia, port 10024) with ESMTP id 13237-10 for ; Mon, 30 Jun 2008 09:52:32 -0300 (ADT) X-Greylist: from auto-whitelisted by SQLgrey-1.7.6 Received: from easy2.in-chemnitz.de (easy2.in-chemnitz.de [88.198.17.232]) by postgresql.org (Postfix) with ESMTP id 904FE650DBC for ; Mon, 30 Jun 2008 09:52:36 -0300 (ADT) Received: from localhost (localhost [127.0.0.1]) by easy2.in-chemnitz.de (Postfix) with ESMTP id E3A5647000B for ; Mon, 30 Jun 2008 14:52:31 +0200 (CEST) Received: from easy2.in-chemnitz.de ([127.0.0.1]) by localhost (easy.in-chemnitz.de [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id yESjWHBI-ehr for ; Mon, 30 Jun 2008 14:52:31 +0200 (CEST) Received: by easy2.in-chemnitz.de (Postfix, from userid 1013) id D1438470004; Mon, 30 Jun 2008 14:52:31 +0200 (CEST) Date: Mon, 30 Jun 2008 14:52:31 +0200 From: Tino Schwarze To: pgsql-admin@postgresql.org Subject: Re: Recommended RAID for Postgres Message-ID: <20080630125231.GB24497@easy2.in-chemnitz.de> References: <46F5921DF52F6E40860287B30CADADC79463A3@EROS.nexus-ag.com> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: <46F5921DF52F6E40860287B30CADADC79463A3@EROS.nexus-ag.com> User-Agent: Mutt/1.5.13 (2006-08-11) X-Virus-Scanned: Maia Mailguard 1.0.1 X-Archive-Number: 200806/243 X-Sequence-Number: 29581 On Mon, Jun 30, 2008 at 02:30:23PM +0200, Thomas Bräutigam wrote: > I have a pretty huge Postgres DB, about 1,3 to 1,5 Terra. Thats plenty of data. > What do you guys recommend on RAID Levels for this Database. Which > does Postgres recommend, and with which do Postgres run very good or > in the best way? Avoid RAID5. Go for RAID10. BTW: I wonder how much the used filesystem affects Postgres' performance... > What Backup Strategy do you think would be the best. Dump the DB once > a week or work with the WAL`s? For your info, the data which is feeded > to the database is available and could be feeded again but it would > maybe take a couple of days. So what would be a solution to bring up > the huge DB in about one day after a crash. This would be the target. How much writing activity is on the DB? WALs are probably the way to go. How often you backup the whole DB space (as a starting point for replaying WALs) depends on how much changes the DB gets. And how long it will take to restore the DB space and the WALs afterwards. HTH, Tino. -- "What we nourish flourishes." - "Was wir nähren erblüht." www.craniosacralzentrum.de www.forteego.de