Wangtek vs Archive drives with SCO Xenix
Frank Bicknell
frankb at usource.UUCP
Sat Jun 3 13:33:08 AEST 1989
I've noticed in my experience with tape drives under SCO Xenix
2.3.1 that while the Archive drive/drivers seem to behave
themselves, the Wangtek drive/drivers misbehave on occasion.
Specifically, when you use /dev/rct0 with the Archive
drive/drivers, the tape rewinds after the operation. This can
be verified with a 'tape status' command returning
beginning-of-tape. However, try the same operation on the
Wangtek drive/drivers and there is a blank status until you
issue a 'tape rewind' command. What's more, tape operations
mysteriously fail if you forget to do this or haven't yet
realized that it was happening; evidently because the tape picks
up where it left off in the middle of the image. (* more below
for those interested)
Has anyone else noticed this strange behavior? And what has
become of the old 'at-filemark' status? I thought 2.2 returned
such a status after a read from /dev/nrct0 or after a 'tape rfm'
command. The 2.3 drivers seem to not know about this (although
the operation seems to work alright).
* More details in case anyone's interested in the "left in the
middle of the image" problem:
For example, when one uses the 't' or 'T' option of restor, it
is only necessary to read the first part of the dump: the part
with the dump date and inode records in it. The Wangtek
drive/drivers just stop after this read, leaving the tape
between these records and the data records. Trying any other
restor operation after doing this gives a 'Volume is not a
backup volume.' (or some such) On the other hand, the Archive
drive/drivers rewinds automatically after the read. No manual
rewind necessary.
BTW.. notice that I've used drive/drivers throughout. Rumor has
it that there's nothing wrong with the drive itself, but rather
the drivers are a little confused at times. True?
--
Frank Bicknell
UniSource; 1405 Main St, Ste 709; Sarasota, FL 34236
killer!usource!frankb || frankb at usource.UUCP
More information about the Comp.unix.xenix
mailing list