chris@bcl.co.nz (Chris Mackerell) (04/05/91)
Has anyone experienced problems running multiple external SCSI peripherals off RC3240 or M120 machines? We have three systems here (2 x RC3240, 1 x M120) that seem to be incapable of handling more than one external SCSI peripheral. An RS2030 system handles the same peripherals with no problems. There is little in common between the systems. (Different customers, sites and peripherals). It doesn't seem to matter what peripherals are used (we've tried combinations of Wren drives, Maxtor drives, 8mm tapes and cartridge tapes), or what order they are connected. It doesn't seem to make any difference if the bus is externally terminated, or the last unit is internally terminated. I've checked the terminator voltage at the end of the bus and it seems ok (5v). We've swapped the motherboard on one RC3240, which improved things but hasn't cured the problem entirely. The symptoms vary with the number of devices attached, varying from failing the SCSI power on diags (with serveral devices attached), to the occasional recovered SCSI error (with one attached). All peripherals work fine in any combination when connected to a CMD SCSI controller on a MicroVAX II. Has anyone else experienced problems similar to this? Is there a cure? Thanks, Chris -- _ _ / /_) / / / / / Chris Mackerell <chris@bcl.co.nz> [a/h chris@owdjim.gen.nz] / /_) / /_ / /_ / Ph: +64 (4) 801-6716 Fax: +64 (4) 85-2752 Business Computers Ltd, Box 3236, Wellington 6000, NEW ZEALAND.
rex@cs.su.oz (Rex Di Bona) (04/05/91)
In article <1991Apr4.220929.3759@bcl.co.nz> chris@bcl.co.nz (Chris Mackerell) writes: > Has anyone experienced problems running multiple external SCSI > peripherals off RC3240 or M120 machines? > > We have three systems here (2 x RC3240, 1 x M120) that seem to be > incapable of handling more than one external SCSI peripheral. An RS2030 > system handles the same peripherals with no problems. We have four M120's and a RC3240, each having external peripherals. The longest string of peripherals we have had would be the internal drive and tape, two external drives (wrens) and an exabyte. - Worked fine! > There is little in common between the systems. (Different customers, sites > and peripherals). I'm sorry, what is your involvment with the machines? Are you a support group or what? Were all the machine purchased from the same supplier? When were the machine purchased? > It doesn't seem to matter what peripherals are used (we've tried > combinations of Wren drives, Maxtor drives, 8mm tapes and cartridge > tapes), or what order they are connected. Ok, silly question time: 1) Did all the devices have different SCSI addresses? 2) Did you do an 'sprobe' command (boot sash and type 'sprobe') This will list what devices the machine can see on the SCSI buss. 3) Do you have any problems with the INTERNAL devices? Is there an internal drive and tape? > We've swapped the motherboard on one RC3240, which improved things but > hasn't cured the problem entirely. In what way did things improve? > The symptoms vary with the number of devices attached, varying from > failing the SCSI power on diags (with serveral devices attached), to > the occasional recovered SCSI error (with one attached). > > Has anyone else experienced problems similar to this? Is there a cure? My vote would go for either bad cables or connectors. Have you tried new SCSI cables? The SCSI connectors on the back of our M120 expansion cabinet were thought to be dodgy, and we did get the symptoms you described. The connector wasn't making a good connection. We would get soft errors, and the occasional hard error. If all the machines were purchased from the same supplier (and about the same time) then that would be a similarity between them. Once we got that fixed we had no problem with the bus (although any non-standard peripherals still wouldn't work :-) -------- Rex di Bona (rex@cs.su.oz.au) Penguin Lust is NOT immoral
pnessutt@dmshq.mn.org (Robert A. Monio) (04/05/91)
In article <1991Apr4.220929.3759@bcl.co.nz> chris@bcl.co.nz (Chris Mackerell) writes: >Has anyone experienced problems running multiple external SCSI >peripherals off RC3240 or M120 machines? > >The symptoms vary with the number of devices attached, varying from >failing the SCSI power on diags (with serveral devices attached), to >the occasional recovered SCSI error (with one attached). > >Has anyone else experienced problems similar to this? Is there a cure? Several local customers that I work with have a mixture of M-120s and RC3240s. All have expansion cabinets with several drives and one even has an exabyte in it. For a while we were experiencing some problems with one of the RC3240s and we finally narrowed it down to a faulty power supply in the expansion cabinet. After replacing the power supply, our SCSI problems went away. The only other problems we've seen with the M-120s has been with the 348MB drives. The MIPS-supplied units seem to have a mean failure rate of 6 months per drive and the one customer who has had 2 of them go bad has now gone directly to Seagate for the units. I don't think any of this is MIPS's problem though since even Seagate has mentioned that there were manufacturing problems with those particular drives. Good luck! -Bob -- Robert A. Monio "Politicians are the lowest form of International Quality Institute, Inc. life on the earth. Liberal Democrats pnessutt@dmshq.mn.org are the lowest form of politician." ..uunet!rosevax!sialis!dmshq!pnessutt -- George S. Patton, Jr.
bin@primate.wisc.edu (Brain in Neutral) (04/06/91)
From article <8530@dmshq.mn.org>, by pnessutt@dmshq.mn.org (Robert A. Monio): > The only other problems we've seen with the M-120s has been with the > 348MB drives. The MIPS-supplied units seem to have a mean failure > rate of 6 months per drive and the one customer who has had 2 of them > go bad has now gone directly to Seagate for the units. I don't think > any of this is MIPS's problem though since even Seagate has mentioned > that there were manufacturing problems with those particular drives. You means the 94171's? Yessir, the two that came with our M/120 went south at 4 months and 7 months. One of the replacements that arrived was DOA, too. In three months the first one that died went out again, and again a year later. They're all working now, though! :-) -- Paul DuBois "The 'C' shell usually doesn't have dubois@primate.wisc.edu have job control." -- LAN TIMES
pnessutt@dmshq.mn.org (Robert A. Monio) (04/06/91)
In article <4165@uakari.primate.wisc.edu> bin@primate.wisc.edu writes: >From article <8530@dmshq.mn.org>, by pnessutt@dmshq.mn.org (Robert A. Monio): >> The only other problems we've seen with the M-120s has been with the >> 348MB drives. The MIPS-supplied units seem to have a mean failure >> rate of 6 months per drive.... > >You means the 94171's? Yessir, the two that came with our M/120 went >south at 4 months and 7 months. One of the replacements that arrived >was DOA, too. In three months the first one that died went out again, >and again a year later. "Yes! That is correct, sir! Yes!" Your record with those drives is a bit more pronounced than ours was, but needless to say we've suggested that anyone who replaces their drive replace it with the 94191 variety.. :-). >They're all working now, though! :-) Sure they are... :-). -- Robert A. Monio "Politicians are the lowest form of International Quality Institute, Inc. life on the earth. Liberal Democrats pnessutt@dmshq.mn.org are the lowest form of politician." ..uunet!rosevax!sialis!dmshq!pnessutt -- George S. Patton, Jr.
bin@primate.wisc.edu (Brain in Neutral) (04/06/91)
>>They're all working now, though! :-) > > Sure they are... :-). Yes, they are! We'll see about tomorrow, though. :-) -- Paul DuBois "The 'C' shell usually doesn't dubois@primate.wisc.edu have job control." -- LAN TIMES
bin@primate.wisc.edu (Brain in Neutral) (04/06/91)
>>They're all working now, though! :-) > > Sure they are... :-). But just to be safe, I moved news processing onto the Fujitsu, so that nothing important will be lost if the 94171's crash again. Ahem. -- Paul DuBois "The 'C' shell usually doesn't dubois@primate.wisc.edu have job control." -- LAN TIMES
stacy@sobeco.com (s.millions) (04/12/91)
In <1991Apr4.220929.3759@bcl.co.nz> chris@bcl.co.nz (Chris Mackerell) writes: >Has anyone experienced problems running multiple external SCSI >peripherals off RC3240 or M120 machines? > ... >I've checked the terminator voltage at the end of the bus and it seems >ok (5v). > ... >Has anyone else experienced problems similar to this? Is there a cure? We have had similiar problems, if the terminator power is ok, try using the twisted pair ribbon cable... it worked for us. We have used M120's and RC3240's with scsi bus full and as long as we are careful with the cable and verify terminator power is ok, we have had little trobule. -stacy -- >Doesn't Sun talk about being standards based? stacy@sobeco.com >Novell is THE standard in PC networks. stacy@sobeco.ca Show us the RFC :-) - aronb@gkcl.ists.ca (Aron Burns) uunet!sobeco!stacy