Hello Managers,
Our environment is an 8400, firmware level Rev: 5.3-11, operating system
Tru64-unix 4.0d, and patch kit 3 (the December version DUV40DAS00003-19981208>
OSF425).
Has anyone seen this: we were performing an upgrade of CAM Medium Changer
Driver from V3.1A to V3.1C (CLCMC311 to CLCMC313). We went to single user,
deleted the 3.1A version, installed the 3.1C version, and let the CLC
installation script back up the kernel and rebuild the new kernel. Before
rebooting with the new kernel, the "robot" CLC command was issued and the
machine panicked (which was to be expected under the old kernel - the reason we
were upgrading the CLC).
The system was reset and began rebooting with the new upgraded kernel, but got
hung during boot. We booted with the old kernel and attempted to rebuild a
new kernel (to pick up the new CLC code again). Could not perform a kernel
rebuild - recieved errors about problems with makelint (noted later that
makelint contained many null characters - \000 ). Renamed and came back up
with orginal/old kernel.
As an aside, we had also installed Tooltalk patch SSRT0566U, which updates
rpc.ttdbserverd and restarts inetd. Though this patch had installed on our
other systems successfully, in this instance it flooded the system, spawning
processes, so we backed it out.
We wonder if the panic could have caused all of the problems that followed: 1)
new kernel hung in boot, 2) inability to perform a kernel rebuild, and 3)
Tooltalk patch spawning processes. If so, we would try the same procedures
again, minus the "panic". If the panic doesn't sound like the culprit, has
anyone else experienced an inability to do kernel rebuilds, or problems with the
Tooltalk patch, in a similar environment?
Thanks,
Lolly Bennett
--------------------------------------------------------------------------
Lolly Bennett
National Institutes of Health
Center for Information Technology
email: lolly_at_nih.gov
Received on Wed Apr 21 1999 - 20:03:25 NZST