From: | Ogden <lists(at)darkstatic(dot)com> |
---|---|
To: | Igor Chudov <ichudov(at)gmail(dot)com> |
Cc: | pgsql-performance(at)postgresql(dot)org |
Subject: | Re: Postgres for a "data warehouse", 5-10 TB |
Date: | 2011-09-11 18:36:56 |
Message-ID: | CD9F56F6-5951-4E60-85E7-519EDF36F43B@darkstatic.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-performance |
On Sep 11, 2011, at 9:21 AM, Igor Chudov wrote:
>
>
> On Sun, Sep 11, 2011 at 9:16 AM, Claudio Freire <klaussfreire(at)gmail(dot)com> wrote:
> On Sun, Sep 11, 2011 at 3:59 PM, Igor Chudov <ichudov(at)gmail(dot)com> wrote:
> > Well, right now, my server has twelve 7,200 RPM 2TB hard drives in a RAID-6
> > configuration.
> > They are managed by a 3WARE 9750 RAID CARD.
> >
> > I would say that I am not very concerned with linear relationship of read
> > speed to disk speed. If that stuff is somewhat slow, it is OK with me.
>
> With Raid 6 you'll have abysmal performance on write operations.
> In data warehousing, there's lots of writes to temporary files, for
> sorting and stuff like that.
>
> You should either migrate to raid 10, or set up a separate array for
> temporary files, perhaps raid 0.
>
> Thanks. I will rebuild the RAID array early next week and I will see if I have a Raid 10 option with that card.
>
> Quantitatively, what would you say is the write speed difference between RAID 10 and RAID 6?
>
As someone who migrated a RAID 5 installation to RAID 10, I am getting far better read and write performance on heavy calculation queries. Writing on the RAID 5 really made things crawl. For lots of writing, I think RAID 10 is the best. It should also be noted that I changed my filesystem from ext3 to XFS - this is something you can look into as well.
Ogden
From | Date | Subject | |
---|---|---|---|
Next Message | J Sisson | 2011-09-11 19:08:36 | Re: Postgres for a "data warehouse", 5-10 TB |
Previous Message | Marti Raudsepp | 2011-09-11 17:02:06 | Re: Postgres for a "data warehouse", 5-10 TB |