[comp.sys.sgi] Sparc tapes

davis@ADENOSINE.PHARM.UTAH.EDU ("Darrell R. Davis") (03/20/90)

My thanks to several people who made suggestions for reading Sparc
tapes, none of them worked, but I did a work-around and will post what
I think the problem was/is.

First the solution was to make a new Sparc tape using tar cvf
/dev/rst8.  The default tape device is /dev/rst0 which must be QIC-150
format, supposedly the same as the PI tape drive, but not
compatible????  After making the /dev/rst8 tape which is supposedly
QIC-24 I used

dd if=/dev/tape conv=swab | tar xvf- 

to read the Sparc tape on my 4D/20, no problem.  It would still be
interesting to find out how to make the QIC-150's talk to each other.


-----Darrell R. Davis 
davis@adenosine.pharm.utah.edu      "Faster, faster, until the 
Assistant Professor                  thrill of speed overcomes
Medicinal Chemistry                  the fear of death."
University of Utah                     --H.S. Thompson  

-----

mitch@rock.sgi.com (Tom Mitchell) (03/27/90)

In article <9003191854.AA20775@adenosine.pharm.utah.edu> davis@ADENOSINE.PHARM.UTAH.EDU ("Darrell R. Davis") writes:
* 
* My thanks to several people who made suggestions for reading Sparc
* tapes, none of them worked, but I did a work-around and will post what
* I think the problem was/is.
* 
* First the solution was to make a new Sparc tape using tar cvf
* /dev/rst8.  The default tape device is /dev/rst0 which must be QIC-150
* format, supposedly the same as the PI tape drive, but not
* compatible????  After making the /dev/rst8 tape which is supposedly
* QIC-24 I used
* 
* dd if=/dev/tape conv=swab | tar xvf- 
* 
* to read the Sparc tape on my 4D/20, no problem.  It would still be
* interesting to find out how to make the QIC-150's talk to each other.
* 
* -----

Darrell got my attention.   Is it possible that the tape drive is
a QIC-11/24 drive and not a QIC-150.  I wonder if Darrell was writing 
a QIC-11 tape.  From a SUN st(4S) man page:

====== snip----------
FILES
     For QIC-150 tape devices (Archive Viper):
          /dev/rst[0-3]        QIC-150 Format
          /dev/rst[8-11]       QIC-150 Format
          /dev/rst[16-20]      QIC-150 Format
          /dev/rst[24-28]      QIC-150 Format
          /dev/nrst[0-3]       non-rewinding QIC-150 Format
          /dev/nrst[8-11]      non-rewinding QIC-150 Format
          /dev/nrst[16-19]     non-rewinding QIC-150 Format
          /dev/nrst[24-27]     non-rewinding QIC-150 Format

     For QIC-24 tape devices (Emulex MT-02 and Sysgen SC4000):
          /dev/rst[0-3]        QIC-11 Format
          /dev/rst[8-11]       QIC-24 Format
          /dev/rst[16-20]      QIC-24 Format
          /dev/rst[24-28]      QIC-24 Format
          /dev/nrst[0-3]       non-rewinding QIC-11 Format
          /dev/nrst[8-11]      non-rewinding QIC-24 Format
          /dev/nrst[16-19]     non-rewinding QIC-24 Format
          /dev/nrst[24-27]     non-rewinding QIC-24 Format

     Note:  The QIC-24 format is preferred over QIC-11 for Sun-3,
     Sun-3x, Sun-4, and Sun386i systems.

====== end snip -------

Note how the name space overlaps.

When READING tapes on an SGI machine (with a SCSI tape
drive) there are only two devices one needs to try.  They
are: /dev/tape and /dev/tapens.  Recently the 'ns' flavor of
device has been added to not swap bytes.  The selection a 'ns'
device can avoid the dd conv=swab stuff.  These two devices
have the complementary no rewind on-open (nr) flavors as well.


The SCSI drives we currently use will switch its own
internal gears between QIC-24 and QIC-150 transparently on
reads, reads, reads, reads.  A QIC-150 drive can write only
QIC-150 format (write, write, write).


Messages about QIC-24 on a QIC-150 drive while attempting to
write to QIC-24 tapes translate to: wrong tape media for
writing For QIC-150 tape drives, this indicates that the
user is trying to write on a DC-300XL (or equivalent) tape.
Only DC-6150 (or equivalent) tapes can be used for writing.
Note: DC-6150 was formerly known as DC-600XTD.  


The unlucky may also fall into the unreliable QIC-120 mode
where a QIC-150 drive was asked to write to a DC-600A tape.
Please avoid this unreliable operation.  The DC-600A media
was not designed to work with the extra track density heads
and is not recommended.



  Thomas P. Mitchell   --  mitch@sgi.com
	"All things in moderation; including Grammarians."