Re: Need Statspack help

From: Hans Forbrich <news.hans_at_telus.net>
Date: Wed, 20 Oct 2004 21:25:51 GMT
Message-ID: <zJAdd.22596$cr4.11980@edtnps84>

Ed Stevens wrote:

>
> Would it be fair to say that, in spite of the theoretical -- and
> sometimes real -- performance limitations of RAID-5 (putting aside for
> the moment considerations of disk failure/recovery issues) if the SA
> can show that the drive system is not being stressed then RAID-5 is
> not your bottleneck?
>
> I'd be happy to be able to prove otherwise, but I've consistently lost
> these battles to SA's, managers and bean-counters who seem to be able
> to only think in terms of total usable storge. So I've had take a
> more pragmatic position of "if it's working with acceptable
> performance. . . "

With due apologies to those who are competent: too often SAs really don't know whether it's working. However ... if it's working don't fix it.

http://www.baarf.com could be considered to contain the "BAARF manifesto". In summary - we [sic] refuse to argue the issue. We let people know that there _could_ be a problem, point to relevant info & try to ignore the itch to help those who don't want it. (As a relative newbie to BAARF, I'm having an especially hard time with that last point.)

/Hans Received on Wed Oct 20 2004 - 16:25:51 CDT