Ok, when I do a 'tar -tvf /dev/rst0' on my Archive Viper 150,
I get the following (on a Sun4/260 with the si SCSI):

Mar 16 16:57:31 repeat /netbsd: si0: SCSI bus busy, resetting...
Mar 16 16:57:32 repeat /netbsd: si0: SCSI bus busy, resetting...
Mar 16 16:57:33 repeat /netbsd: si0: SCSI bus busy, resetting...
Mar 16 16:57:35 repeat last message repeated 2 times
Mar 16 16:57:36 repeat /netbsd: st0: bad request, must be between 512 and 512

	Ok... I think to myself, 'dd if=/dev/rst0 of=/dev/null bs=512'
... and I get the following:

Mar 16 16:58:00 repeat /netbsd: si0: SCSI bus busy, resetting...
Mar 16 16:58:06 repeat /netbsd: si0: SCSI bus busy, resetting...
Mar 16 16:58:06 repeat /netbsd: st0(si0:4:0): unit attention, data = 00 00 00 00 00 00

	At this point, things start working (albeit very slowly).
Obviouly you can read/write bigger blocks (they get split up
somewhere, I understand) ... in fact this worked a month ago.  I don't
know where the SCSI reset is coming from, but I'm going to poke around
after the blocksize problem.

	The config message for st0 is:

Mar 11 17:52:20 repeat /netbsd: si0 targ 4 lun 0: <ARCHIVE, VIPER 150  21247, -005> SCSI1 1/sequential removable
Mar 11 17:52:20 repeat /netbsd: st0 at scsibus0: rogue, drive empty

Dave.

-- 
----------------------------------------------------------------------------
|David Gilbert, PCI, Richmond Hill, Ontario.  | Two things can only be     |
|Mail:      dgilbert@jaywon.pci.on.ca         |  equal if and only if they |
|http://www.pci.on.ca/~dgilbert               |   are precisely opposite.  |
---------------------------------------------------------GLO----------------