SUMMARY: 7 Bay Tower CD questions

From: Dan A. Zambon (dzambon%hawkeye@ra.mcs.anl.gov)
Date: Thu Apr 27 1995 - 12:01:35 CDT


Hi Netters!
Sorry about the very late summary. No real excuse, just normal
work load with a touch of procrastination. Hope the delay didn't cause
any grief....:>)

My original post was:
****************************************************************
Hi Netters!
I need some information from the Masters (not the Augusta, GA ones - you
fine folk!) about some configuration concerns I am having...

I have a need to attach a CD Tower (with anywhere from 4 to 14 CD ROM drives
contained within) to a Sparc2. Each one of the drives has to have it's own
SCSI ID, which would cause definite problems with the Sparc2, especially
if I try and hang an additional 14 SCSI devices on it......:>)

So my questions are:
1) Can I extend the current 7 ID limit on the Sparc2 SCSI bus?
2) Does anyone have experience or suggestions on CD Towers? I
   am currently looking at some hardware from Excel Computer
   from Carrollton, TX.
3) Will hanging 4-14 CD ROM drives on a Sparc2 bring it down to
   its knees?
4) If the kids leave home and you decide to move, are you obligated
   by law to let them know your new address? Does this apply for
   short trips, such as to the grocery store?....:>)

I apologize if this is not a topic that should be posted, but I do rely
quite a bit on the knowledge and helpfulness that is constantly displayed
by this group.

Thanks for your time and information, and I will of course post a summary
if there is enough interest.

******************************************************************

I got many good responses from it...............

On Question 1 (extending the 7 ID limit):
-Most said to extend this PARTICULAR ID limit above 7 cannot be done
 as it is a fixed component of the SCSI specification. However I could ADD
 another SCSI bus via an SBUS card which would give me an additional
 7 devices. This I will probably do.
-Some mentioned the SCSI spec provides for "LUNS" or Logical
 Unit Numbers as a sort of a subdivision. They stated that these were
 not used much, although one person said they had a tower (from JVC) that bypassed
 the SCSI limit by using the unit number (subfield of the SCSI unit field)
 so that the controller uses one SCSI ID and the drives use a second, with
 each drive a unit off of the second.
-I also was told that SunOS 4.x only supports one CD ROM per machine, and
 special drivers are required to support more than one.

On Question 2 (any experience on hardware from Excel Computer):
-Most said they had no experience with this company.
-One respondee mentioned that they used an optical jukebox (16 disk capacity).
-One person mentioned that I look into getting some large capacity disk drives
 instead of the CD ROMs. Drives are now pretty cheap, and they are much faster
 than the ROMs. (But then, what do I do with the $13K CD Mastering system
 I just ordered, Gene?...:>)

On Question 3) (would adding many ROMs to a S2 bring it down to its knees):
-Some interesting comments on this one. A few stated that the throughput of
 the CD ROM is very slow compared to the bandwidth of the SCSI bus. To quote
 John Hall (thanks, John):
        "Traffic is the issue. An SS2 can probably handle two busy SCSI busses
           without too much trouble. A CD ROM drive has limited throughput:

                single speed 150KB/s
                double speed 300KB/s
                quad speed 600KB/s

         This does not even approach the 5MB/s or SCSI-1 or 10MB/s of SCSI-2."
 This makes sense.
-One person did suggest putting all the CD drives on their own bus away from disks
 and tapes, and stating that the impact of adding the ROMs would not mess with the
 performance of the S2 all that much.
-And a few users thought that trying to attach a tower to a SPARC2 with
 upwards of 14 CD ROMs on it would burden the machine, as the S2 is older
 technology. (I don't necessarily agree with this).
  

On Question 4 (moving and not telling the kids):
-We move this weekend....

Many thanks to those who took the time and effort to repond to my post. I have
a better understanding of my systems and I can now make some decisions.

Thanks to:
Srinivasa R. Yalavarthy <srini@concorde.com>
Reto Lichtensteiger <rali@hri.com>
gunn@lardav.com (David Gunn)
brobbins@mako.Newbridge.COM (Bert Robbins)
Will Heres <will@bradman.mit.edu>
poffen@San-Jose.ate.slb.com (Russ Poffenberger)
Paul Caskey <pcaskey@swcp.com>
John Hall <jhall@sqi.com>
Eugene H. Simpson III <Eugene.H.Simpson.III@acenet.auburn.edu>
Mark Ferraretto <mferrare@physics.adelaide.edu.au>
jayl@lattice.com (Jay Lessert)
mad@suned1.Nswses.Navy.Mil (Michael A McDevitt)
johnh@gerbil.umds.ac.uk (John Hearns - System Manager)
ahs!california!david@uunet.uu.net
adamfox@super.org (Adam Fox)
Glenn.Satchell@uniq.com.au (Glenn Satchell - Uniq Professional Services)

A copy of all responses can be emailed to you if you wish. Just let me know
at the address below.

Thanks
Dan
------------------------------------------------------------------
Dan A. Zambon email: dzambon@afit.af.mil
A Packer Backer from way back (remembers when Nitschke had hair)
Air Force Institute of Technology phone: (513) 255-6565, ext
Department of Electrical and Computer Engineering 4280
Wright Patterson AFB, Oh 45433-7765 fax: (513) 476-4055
------------------------------------------------------------------



This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:10:23 CDT