v097pba8@ubvmsd.cc.buffalo.edu (Ken F Morton) (05/04/91)
Somehow this got posted on comp.binaries.apple2 - I have *no* idea how... Anyway, can a mac and an Apple IIgs share the same HD? Thanks... Ken Morton "A vacant engineer rides on a v097pba8@ubvms.Bitnet train of thought that will not v097pba8@ubvmsd.cc.buffalo.edu lead him home" - Edie Brickell
unknown@cats.ucsc.edu (Matt Ackeret) (05/04/91)
In article <74719@eerie.acsu.Buffalo.EDU> v097pba8@ubvmsd.cc.buffalo.edu writes: > > Somehow this got posted on comp.binaries.apple2 - I have *no* >idea how... > Anyway, can a mac and an Apple IIgs share the same HD? Apparently it can be done, but Apple won't tell us how... heh... actually, I've heard it can be done, but it's never really been documented.. (I was told this by someone who used to work at Apple) -- /unknown@ucscb.ucsc.edu Apple IIGS Forever! unknown@cats.ucsc.edu\ |WANT to help get ULTIMA VI //e or GS written?-mail me. CHEAP CD info-mail me.| \ McIntosh Junior: The Power to Crush the Other Kids. /
mattd@Apple.COM (Matt Deatherage) (05/04/91)
In article <15350@darkstar.ucsc.edu> unknown@cats.ucsc.edu (Matt Ackeret) writes: >> Anyway, can a mac and an Apple IIgs share the same HD? > > Apparently it can be done, but Apple won't tell us how... heh... > > actually, I've heard it can be done, but it's never really been >documented.. (I was told this by someone who used to work at Apple) This always turns into a bloody holy war because everyone (except me) thinks they know a whole lot about SCSI. I don't pretend to, but I know my way around our systems, so let me give it a shot. The SCSI specs say that SCSI implementations may or may not have more than one master on the SCSI bus. Apple chooses not to. This makes the cost of some of the SCSI stuff lower since we don't have to include pounds of SCSI bus arbitration software/hardware/voodoo magic. This is perfectly legal within the SCSI spec but really annoys people who want to hook multiple CPUs to one drive, so they often cite the "you can do this" clause and say Apple is not following the SCSI spec because we don't do that. Read it more carefully, prospective flamers. So, if you try this, you'll run into bus arbitration problems whenever more than one of the masters tries to access the drive at the same time. This is similar to the old CMS SCSI card (which did kind of support this) problem, which they solved by telling users to only let one CPU write to any given partition, and they did hard-partitioning things. I think. Remember, I'm not a SCSI expert and don't play one on TV. At any rate, you might be able to pull it off for a short period of time, but unless you always exact extreme care something will eventually go boom. -- ============================================================================ Matt Deatherage, Developer Technical | The opinions expressed herein are Support, Apple Computer, Inc. | not those of Apple Computer, and Personal mail only, please. Thanks. | shame on you for thinking otherwise. ============================================================================
toddpw@nntp-server.caltech.edu (Todd P. Whitesel) (05/04/91)
mattd@Apple.COM (Matt Deatherage) writes: No offense, Matt, but I've been wrong on this more often then you :) ... Here's the Scoop. >The SCSI specs say that SCSI implementations may or may not have more than >one master on the SCSI bus. Apple chooses not to. Actually, Apple chose not to do it in the Mac SCSI manager ROMs. It is possible for Mac SCSI drivers to implement arbitration, although most don't. I had that vehemently pointed out to me by someone who writes those drivers. The DMA SCSI card ROMs, however, are totally ANSI compliant, including support for full arbitration. This I learned from Matt Gulick himself face-to-face. >So, if you try this, you'll run into bus arbitration problems whenever more >than one of the masters tries to access the drive at the same time. This is correct. The only way to get around it is to have ALL the bus masters support arbitration. So if you want to share a SCSI drive, make sure you have a DMA SCSI and a Mac with good SCSI drivers installed. Unfortunately, the Mac HD vendors who write drivers that support arbitration aren't being very vocal about it. The Mac magazines don't seem to include it in their reviews, either. Todd Whitesel toddpw @ tybalt.caltech.edu
gtolar@xcluud.sccsi.com (Glynne Tolar) (05/07/91)
toddpw@nntp-server.caltech.edu (Todd P. Whitesel) writes: > mattd@Apple.COM (Matt Deatherage) writes: > > No offense, Matt, but I've been wrong on this more often then you :) ... > Here's the Scoop. > > >The SCSI specs say that SCSI implementations may or may not have more than > >one master on the SCSI bus. Apple chooses not to. > > Actually, Apple chose not to do it in the Mac SCSI manager ROMs. It is possib > for Mac SCSI drivers to implement arbitration, although most don't. I had tha > vehemently pointed out to me by someone who writes those drivers. > > The DMA SCSI card ROMs, however, are totally ANSI compliant, including suppor > for full arbitration. This I learned from Matt Gulick himself face-to-face. I'll just let my expirence speak for itself and those who want to can judge if Apple's DMA SCSI card follows ALL the rules. I've connected two DMA SCSI cards to one Seagate ST-296N. One card in a IIgs, the other in a //e. Additionally I tested it with block reads (verify disk) in Copy II Plus. They wasted no time in locking up. When I keep the traffic low on the SCSI bus it works most (but not all) the time. The only thing I could posibly find wrong, besides the DMA SCSI card MAYBE being built wrong, is maybe the Seagate should have not been terminated. By the way the same results occoured with a DMA and a Rev.C SCSI card. I plan to try this again in the future and try a few different things but I better move fast. I'm selling my DMA SCSI because now I have a RamFAST. Note: I know the RamFAST card CAN NOT be used in this enviroment. > Todd Whitesel > toddpw @ tybalt.caltech.edu We don't care, we don't have to. We're Apple computer. ----------------------- OR ----------------------------- Apple II forever. I just love hearing the company LInE. Pick your favorite. gtolar@xcluud.sccsi.com I didn't write the organization or path. I just use it.
toddpw@nntp-server.caltech.edu (Todd P. Whitesel) (05/08/91)
gtolar@xcluud.sccsi.com (Glynne Tolar) writes: >I've connected two DMA SCSI cards to one Seagate ST-296N. > They wasted no time in locking up. First, the Seagate had better not have been terminated. Second, the DMA SCSIs must be the ends of the chain (that's probably what you did, hooking both into the two 50-pin connectors on the back of the HD case). > By the way the same results occoured with a DMA and a Rev.C SCSI card. The Rev.C ROMs don't arbitrate so that makes sense. There is one other possibility: the DMA SCSI has terminating resistors built into it, but it does not supply power to them (what, and toast Apple's wimpy power supply?). The drive has to do it and I do not know for sure if Seagates supply terminator power or not. (I know my Quantum does.) Todd Whitesel toddpw @ tybalt.caltech.edu
whitewolf@gnh-starport.cts.com (Tae Song) (05/09/91)
| Somehow this got posted on comp.binaries.apple2 - I have *no* |idea how... | Anyway, can a mac and an Apple IIgs share the same HD? The first thing you'd have to over come is to set one of the computer's SCSI ID to 6. I'm sure if that can be done with the Apple rev C SCSI card or the Mac or the DMA SCSI card, you're half way there. And don't forget get to set the HDs ID to 5 or lower. I haven't tried this nor do have anyway to verify weather it works or not. Not setting the IDs correctly might cause the confusion on the bus. INET: whitewolf@gnh-starport.cts.com UUCP: crash!gnh-starport!whitewolf ARPA: crash!gnh-starport!whitewolf@nosc.mil