Re: RAID for the DB filesystem - Mailing list pgsql-admin

From Scott Marlowe
Subject Re: RAID for the DB filesystem
Date
Msg-id dcc563d10908031459s10f59e9mf015fa385f95b55f@mail.gmail.com
Whole thread Raw
In response to RAID for the DB filesystem  (Brian Modra <brian@zwartberg.com>)
Responses Re: RAID for the DB filesystem  (Brian Modra <brian@zwartberg.com>)
List pgsql-admin
On Mon, Aug 3, 2009 at 10:15 AM, Brian Modra<brian@zwartberg.com> wrote:
> Hi,
> my database is hit with constant inserts to 6 main tables (200 inserts
> per minute to one of the tables, less to the others), some updates,
> but then the selects:
> - large retrievals of randomly different sections of the database
> (indexed maps by postgis). This data is static.
> - medium sized retrievals of the same tables that are receiving the
> inserts. By mediou sized, I mean typically 200 rows at once. These
> retrievals are also randomly different to each other, and typically
> retrieving the newly inserted data rather than the more historical.
> The database size is about 300GB and growing.
>
> What sort of hardware config would you advise?
> I'm thinking of 2x300GB SATA RAID 0 for the OS and application files,

Is there a valid reason you're NOT considering RAID-1 here?  I hope
RAID-0 is a typo.

> and 6x300GB SAS RAID 10 for the database... but some experts have said
> RAID 5 is fine. I'm inlined to think RAID 10, but I'm not an expert.
> Your advice will be much appreciated.

Then I question the expertise of your experts.  RAID5 is not fine.
It's slow, more prone to loss due to drive loss, and generally not a
good choice for databases.

I would gladly have more SATA drives in a RAID-10 than fewer SAS
drives in a RAID-5.

if someone is worried about "wasting" disk space tell them to worry
about something else, like losing data.

pgsql-admin by date:

Previous
From: Brian Modra
Date:
Subject: RAID for the DB filesystem
Next
From: Rodrigo E. De León Plicet
Date:
Subject: Re: RAID for the DB filesystem