Path: utzoo!censor!geac!lethe!yunexus!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!cs.utexas.edu!samsung!uunet!mcsun!hp4nl!phigate!ehviea!hulsebos
From: hulsebos@ehviea.ine.philips.nl (rob hulsebos)
Newsgroups: comp.periphs.scsi
Subject: Re: MAXTOR 4380s
Message-ID: <946@ehviea.ine.philips.nl>
Date: 26 Feb 91 09:06:07 GMT
Article-I.D.: ehviea.946
References: <1991Feb25.011352.3968@watserv1.waterloo.edu>
Reply-To: hulsebos@ehviea.UUCP (rob hulsebos)
Organization: Philips I&E Eindhoven
Lines: 30

In article <1991Feb25.011352.3968@watserv1.waterloo.edu> mann@watserv1.waterloo.edu (Shannon Mann) writes:
>Has anyone had any experience with this drive?
Yes, I had one about a year ago.

>I would like to place this drive on a SCSI controller that already has a
>tape drive and a >100 Meg hard drive on it.
>I have heard rumours that this drive does
>not like being on a SCSI bus with any other devices.  Can anyone
>confirm/dispel this rumour? 
The one I had worked fine with other devices on the bus. In my system
I had an Archive streamer, Micropolis 1375 disk and a Manta floppycontroller.

>Are there any other items of note I might be interested in?
I noticed that after an SCSI-bus reset the 4380 is able to accept commands
sent to it, but they are ignored for a certain time (apparently while the
drive is handling its own reset). You don't get an error from the drive, it
just does nothing.

This means that drivers without timeouts may hang. I did not have time to
find out how this is caused, but as a quick-and-dirty fix I just added a
delay of several seconds in my driver to get it to work. The behaviour of 
the 4380 is different from the Micropolis disks, which do not ignore commands
while being reset. Another issue which makes life difficult for vendors
saying that they sell SCSI-drivers (because they have never tested their
drivers with all available SCSI-disks). So much for 'standards' :-(

Hope this helps.

Rob.
hulsebos@ine.philips.nl