From owner-freebsd-current  Tue Sep  3 09:49:00 1996
Return-Path: owner-current
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id JAA10789
          for current-outgoing; Tue, 3 Sep 1996 09:49:00 -0700 (PDT)
Received: from night.primate.wisc.edu (night.primate.wisc.edu [144.92.43.19])
          by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id JAA10784
          for <current@freebsd.org>; Tue, 3 Sep 1996 09:48:55 -0700 (PDT)
Received: by night.primate.wisc.edu;
          id LAA27866; 8.6.10/41.8; Tue, 3 Sep 1996 11:50:09 -0500
From: Paul DuBois <dubois@primate.wisc.edu>
Message-Id: <199609031650.LAA27866@night.primate.wisc.edu>
Subject: Re: Food for thought
To: imp@village.org (Warner Losh)
Date: Tue, 3 Sep 1996 11:50:09 -0500 (CDT)
Cc: current@freebsd.org
In-Reply-To: <199609031549.JAA26803@rover.village.org> from "Warner Losh" at Sep 3, 96 09:49:53 am
X-Mailer: ELM [version 2.4 PL25]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: owner-current@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

>That said, I recently bought more disk (a Jaz drive), and have been
>very happy with CTM of CVS.  I've not had any make worlds fail since
>I've moved over to that which weren't the result of disk full
>failures.

Had you had problems with FreeBSD locking up if the Jaz has to be
brought out of sleep mode?  (I'm running 2.1 but) I've seen this in
two contexts:

1) During a FreeBSD install.  I got to the point where the install process
began loading my selected subsets.  This was going to take a while, so
I went away.  Then I can back and was ready to finish up.  At some point
during the finish-up, the Jaz got woken up, at which point the install hung.
(I wasn't installing to the Jaz, by the way.)
2) During bootup.  The SCSI probe part of the bootup process woke a sleeping
Jaz, a command timeout and abort occurred, and the boot process was hung.

Does -current handle these situations better?

This was on a Dell OptiPlex with Adaptec 2940UW.  In both cases a power cycle
was necessary to restart the machine.