[5379] in linux-scsi channel archive
Re: SCSI write caching (was: Dumb question: Which is "better" SCSI or IDE disks?)
daemon@ATHENA.MIT.EDU (C S Hendrix)
Thu Dec 10 07:26:38 1998
To: linux-kernel@vger.rutgers.edu, linux-scsi@vger.rutgers.edu
Reply-To: shendrix@escape.widomaker.com
In-reply-to: Your message of "08 Dec 1998 16:51:19 EST."
<5ln24ye208.fsf@tequila.cs.yale.edu>
Date: Wed, 09 Dec 1998 12:20:18 -0500
From: C S Hendrix <shendrix@escape.widomaker.com>
In message <5ln24ye208.fsf@tequila.cs.yale.edu>, Stefan Monnier writes:
> >>>>> "Christopher" == Christopher Smith <cbsmith@envise.com> writes:
> > In general, I'd say, unless you know you need write caching disabled, you
> > probably should have it enabled.
>
> Beside the `is it safe', I believe the main question is:
> how much performance improvement can I expect (this will of course depend on
> the load, but then I'd like to know about that too) ?
Last night I ran some tests with bonnie.c on my system with a DPT
2044 controller and two Seagate Hawk fast-wide drives.
The tests were actually faster with write cache turned off. I'm not
sure I really expected it to be faster, but I was surprised that
cache slowed things down.
I'm talking too about a fairly significant different of nearly
2MB/sec on the block output test.
Granted, bonnie and other benchmarks are not the best, but I think
it is safe to at least say write-cache enabled was slower.
--
Shannon - shendrix@widomaker.com - InfiNet?
----------------------------------------------------------------------
"We are all of us in the gutter, some of us looking at the stars." --
Alexander Pope
-
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@vger.rutgers.edu
