Subject: Re: RAID 0 not as fast as expected
From: Scott Marlowe <smarlowe@g2switchworks.com>
To: "Craig A. James" <cjames@modgraph-usa.com>
Cc: Alan Hodgson <ahodgson@simkin.ca>, pgsql-performance@postgresql.org
In-Reply-To: <4509CB04.4020605@modgraph-usa.com>
References: <450999EC.8070601@modgraph-usa.com>
	<200609141136.03345@hal.medialogik.com>
	<4509CB04.4020605@modgraph-usa.com>
Content-Type: text/plain
Content-Transfer-Encoding: 7bit
Message-Id: <1158271115.24726.24.camel@state.g2switchworks.com>
Mime-Version: 1.0
Date: Thu, 14 Sep 2006 16:58:35 -0500

On Thu, 2006-09-14 at 16:35, Craig A. James wrote:
> Alan Hodgson wrote:
> > On Thursday 14 September 2006 11:05, "Craig A. James" 
> > <cjames@modgraph-usa.com> wrote:
> >> I'm experiment with RAID, looking for an inexpensive way to boost
> >> performance.  I bought 4 Seagate 7200.9 120 GB SATA drives and two SIIG
> >> dual-port SATA cards.  (NB: I don't plan to run RAID 0 in production,
> >> probably RAID 10, so no need to comment on the failure rate of RAID 0.)
> >>
> > 
> > Are those PCI cards?  If yes, it's just a bus bandwidth limit.
> 
> Ok, that makes sense.
> 
>    One SATA disk = 52 MB/sec
>    4-disk RAID0  = 106 MB/sec
>    
>    PCI at 33 MHz x 32 bits (4 bytes) = 132 MB/sec.
> 
> I guess getting to 80% of the theoretical speed is as much as I should expect.

Note that many mid to high end motherboards have multiple PCI busses /
channels, and you could put a card in each one and get > 132MByte/sec on
them.

But for a database, sequential throughput is almost never the real
problem.  It's usually random access that counts, and for that a RAID 10
is a pretty good choice.