Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!purdue!decwrl!sun!pitstop!sundc!seismo!uunet!van-bc!root From: lphillips@lpami.wimsey.bc.ca (Larry Phillips) Newsgroups: comp.sys.amiga Subject: Re: SCSI question Message-ID: <2270@van-bc.UUCP> Date: 6 Mar 89 14:40:37 GMT Sender: root@van-bc.UUCP Lines: 50 In <12889@gryphon.COM>, jdow@gryphon.COM (J. Dow) writes: >In article rg20+@andrew.cmu.edu (Rick Francis Golembiewski) writes: >>I seem to remember a reading somewhere that 2 machines with SCSI >>controllers could be hooked up to the SAME hard drive (with >>partitions for each if the machines used a different format). Has >>anyone tried something like this? I was kind of wondering if would >>be possible to hook an Amiga with SCSI controller up to a Mac's >>external HD. Kind of a crazy solution, but the other alternative >>requires $$$ for a HD (sigh...). Any comments are appreciated. >Yup, you can. With HardFrame, for example, you could set the two controllers >to separate SCSI addresses and have both connect to the same drive. But what >you cannot do in that situation is have both of them autoboot. (They'd try >to mount the same partitions. While this is not "impossible" and CAN run it >results in corrupted filesystems REAL fast unless you use REAL weird disk >change tricks. (I have a disked script that seems to work. Argh is it a >crungy hack so I won't tell anyone how.)) Fascinating. Does the HardFrame support arbitration automatically or do you have to do anything special with it? About the partitioning and autoboot. I was thinking that two machines (both Amigas of course, I don't have or want a Mac), each with a SCSI host that supported autoboot, could be hooked to the same drive, and could boot from the same partition, _provided_ that the boot partition was stable, or that you were very careful about doing diskchanges whenever the common partition was written to. A number of other 'tricks' come to mind for such operations as updating a common partition. For example, both machines would have read access to a common partition, but only one would have write access. The machine with read-only access would update it by writing to a partition that had the read/read-write permissions reversed. The machine that updates the common partition would occasionally access the other machines writable partition and check for the need to update. There's probably all sorts of holes in this that I haven't thought of, primariy because I haven't thought about it much. Wouldn't it be nice to have a hardware 'disk changed' signal that the driver could recognize? -larry -- Frisbeetarianism: The belief that when you die, your soul goes up on the roof and gets stuck. +----------------------------------------------------------------------+ | // Larry Phillips | | \X/ lphillips@lpami.wimsey.bc.ca or uunet!van-bc!lpami!lphillips | | COMPUSERVE: 76703,4322 | +----------------------------------------------------------------------+