mrwallen@ucsd.edu (Mark R. Wallen) (05/26/90)
Here is what I've uncovered empirically so far. The Quantum ProDrive 80S drives listed in the 4.0.X format.dat files (and thanks for the pointers) seem to be different from the Quantum 280S drives in the Macintosh external 80 meg drives that I have. On my Sun 4/280 running 4.0.3 with a 'sc' (this is probably very important) SCSI controller, I could use format to 'read' and 'write' (in analyze) using the ProDrive disk type. Interestingly, (for me) was that there was a [defect table found] when I selected the ProDrive type. However, doing a format itself would always fail on sector 3 (0/0/3), and "label" would fail writing the backup label (near the end of the drive). By chopping down the number of cylinders I was able to get a configuration that would "label" happily. (Note these disks were already formated on/for a Macintosh so if the formats are the same I win/won big (and did)). Using the Quantum Prodrive format.dat entry and paring it down to ncyl=764 and pcyl=766 got me something workable. That means I could get a working defect list written to the drive and mount and use it (I'm doing so even as I type). But I could never actually format the drives under SunOS 4.0.3. And I'm not real confident about doing a 'repair', should that come up. By going into 'analyze' before formatting and then doing 'setup", one can select extended error messages. The error message that you get when you try to format seems to indicate that there are 823 cylinders and 31 sectors per track, not the 766 and 34 sectors per track that I was using. Using a format.dat entry with those parameters did not work either (still wouldn't format). I also tried running a macintosh utility on the drive to find out how many sectors it thought there were: 156272! (Try running that through /usr/games/factor!) I have also discovered that I can only hang 2 drives off the sc controller because the drives have external termination. After adding drive 2, the Exabyte, which is also on the sc controller, started failing with SCSI parity errors. I'm still tinkering away. Thanks to all who responded. Mark Wallen Cognitive Science, UC San Diego mwallen@ucsd.edu