Exabyte Woes

[Comer Duncan]

We have an Exabyte and are trying to get it integrated with our Sun4/260
and are having troubles.  I am hoping that some of you have a
configuration similar to ours and can give us a few suggestions as to what
might be the problem.  Here is the situation.  We purchased an Exabyte
from Software Associates/CITA who got it from Artecon.

Let me describe our system and how we have tried to connect the Exabyte.
Our system is a Sun4/260 with 16Mb (8 from Sun, 8 from Helios) and a
cartridge tape and SCSI host adaptor from Sun.  The system has a color
monitor.  For disk we have a CDC Imprimis, which is a SMD with Xylogic 753
controller.  All works just fine until we try to connect up the Exabyte.
The idea was to buy an internal cable to connect to the existing SCSI host
adaptor which drives the cartridge tape.  Artecon claims to have done this
for 'several' systems and told CITA it works just fine.  We can not seem
to verify this good behavior... First, we connected the Artecon internal
cable to the Emulex-202 per instructions with one end connected to the
cartridge tape and the other end going out of the left side of the Emulex
to a face plate containing a 50-pin conncector.  The external cable from
the back of the Sun then goes to the Exabyte.  We configured things so
that the cartridge is st0 and the Exabyte is st1 and of course did a
MAKEDEV st1.  Whenever we try to use the Exabyte nothing happens.  Even
worse, the cartridge tape drive does not function.  In fact the external
red light on the cartridge tape drive is lit EVEN when no tape is in the
cartridge drive...Not good.  After entirely too much time spent talking to
support people about this we were told to try disconnecting the cartridge
tape from the loop and connecting up the internal cable so that only the
Exabyte was at the end of the line.  We did this and set the address to
st0 on the Exabyte but still no success.  I should have mentioned that we
have unterminated the Emulex since the Exabyte was to be the last device.
The Exabyte is terminated.  Actually we have tried unterminating the
Emulex and see no difference in result.

Given that the kernel has the right drivers (we are running 4.0 with the
Artecon-supplied drivers), what gives??  I am hoping that some of you have
the same situation and that you had success with this and can give some
advice.  We are not novices and have given this far more work that we ever
expected.  We are not pleased with the quality of the support and
documentation but do want to point out that CITA and lately Artecon have
worked hard to figure out what is going on.  It is just that they have not
been able to tell us what is wrong....

Please respond to me directly.  If there is enough information of general
interest I will gladly summarize for sun-spots.

Comer Duncan
Department of Physics and Astronomy
Bowling Green State University
Bowling Green, OH 43403
gcd at einstein.bgsu.edu