Date: Tue, 27 Aug 1996 08:59:25 +0200 (MET DST) From: Lars Koeller <Lars_Koeller@odie.physik2.uni-rostock.de> To: se@ZPR.Uni-Koeln.DE, wjw@IAEhv.nl (Willem Jan Withagen) Cc: hardware@freebsd.org, scsi@freebsd.org, admin@IAEhv.nl Subject: Re: SCSI disks Message-ID: <199608270659.IAA20672@odie.physik2.uni-rostock.de> In-Reply-To: Mail from 'Stefan Esser <se@ZPR.Uni-Koeln.DE>' dated: Mon, 26 Aug 1996 21:09:38 %2B0200 (MET DST)
next in thread | previous in thread | raw e-mail | index | archive | help
Hi!
Stefan Esser writes:
> Willem Jan Withagen writes:
> >
> > Hi,
> >
> > Until recently we as an ISP were using Micropolis disk with pleasure.
> > But since several 4.3Gb started crashing, we've changed our view.
> >
> > So I'm now looking for new disks. I've got 2 offerings:
> > Seagate Barracuda 4.3 GB
> > Quantum XP34300S
> > The need to run on FreeBSD 2.0.5 and up.
> > Long agood I tested the quantum with an NCR 53810 and an Adaptec 1542B
> > and in both case it did not seem to work. Now things might have changed, but
>
> Hmmm, that's surprising, since my current development system uses
> a Quantum Atlas (though its the XP32150S, the 2GB version). This
> drive worked with the NCR driver of Spring 1995 at least, when I
> first tested it.
>
> I'm quite sure that the Adaptec 1542 supported it as well, since
> it is hard to find a device that doesn't work on the Adpatec, and
> the Atlas is really well designed (IMHO).
>
> The Quantum Grand-Prix (XP34301S) is a different beast. It has got
> bad press because of many failed (or DOA) drives, and it is noisy
> and significantly slower than the Atlas, though both are 7200RPM
> drives.
>
> You may also want to have a look at the IBM DHFS series, which is
> also very fast and AFAIK reliable.
>
> Can't comment on recent Seagate drives.
>
> Regards, STefan
Perhaps it's the same phenomena I have with the Quantum and the
Adaptec 1742 EISA SCSI controller. Every time the disc reads out of the
cache, the response is so quick that a lot of controllers aren't ready
to receive the first bytes. As a matter of fact the whole SCSI-bus is
locked! The solution is to edit the modepage of the Quantum Atlas so
that the disc inserts a delay of 15 micro seconds when a block is
transfered out of the cache.
If you need any information how to do this, don't hesitate to contact
me again.
Regards
Lars
--
________________________________________________________________________
Lars Köller Phone: +49 381/498-1665, Fax: -1667
University of Rostock E-Mail:
Department of Physics Lars_Koeller@odie.physik2.Uni-Rostock.DE
Universitätsplatz 3 Anonymous ftp:
D-18051 Rostock (Germany) ftp://odie.physik2.uni-rostock.de/pub
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199608270659.IAA20672>