🚀 go-pugleaf

In <Brzs2r.23I@exnet.co.uk> dhd@exnet.co.uk (Damon) writes:

| In article <nodq9mk@zuni.esd.sgi.com> olson@anchor.esd.sgi.com (Dave Olson) writes:
| >In <AJC.92Jul24120000@thendara.pa.dec.com> ajc@pa.dec.com (AJ Casamento) writes:
| >|   My understanding of the SCSI-2 standard is that it calls out a standard
| >| minimum delay of 200MHz (5 Mtransfers/sec). It also allows for the implementa-
| >| tion of Fast Differential (I believe that Fast Single-ended is not supported
| >| and would only allow a cable length of 0.5M) which calls out a minimum delay
| >| of 100MHz (10 Mtransfers/sec).
| >
| >There is an implementors note that fast SCSI might require differential
| >to be reliable.  This hasn't turned out to be the case (actual experience).
| >Several drive vendors are saying that they support single ended fast SCSI,
| >but are also saying that they want the bus to be 4 meters or less, in
| >this case, rather than 6 m.
| 
| Hmm.  I've regularly had problems with single-ended SCSI-1 over 2m on one (Sun)
| system, and other people who've had problems (eg crashes during heavy paging
| activity, which is when we had most problems too) have had the problems go away
| when they've changed to shorter cables.  I'd be cautious about using the full
| 6m, and spent a bit of time getting data on it from the SCSI committee people.
| (I'm working on a project where physical constraints use of the whole 6m.)

I was making the unspoken assumption that you have a clean implementation,
as do the devices.  SGI hasn't been completely squeaky clean on this
issue in the past, but we keep trying, and in general, do fairly well.
--
Let no one tell me that silence gives consent,  |   Dave Olson
because whoever is silent dissents.             |   Silicon Graphics, Inc.
    Maria Isabel Barreno                        |   olson@sgi.com