This is very unclear, but there is some chance that the automount fix
described previously (starting it after nfs) is causing panics on our
OSF/1 v3.0 alphas.
Since modifying the /sbin/init.d/nfs and /sbin/init.d/nfsmount files,
5 out of 8 alphas have had repeated crashes. Error messages in "uerf" are
as follows:
date:time PANIC
panic (cpu 0): m_copym offset
Has anyone seen this message, and do you have any insight into what it is
trying to tell us?
There is an additional factor that muddies the whole situation - we have a
user running tests using "non-blocking TCP", and he has succeeded in
hanging at least 2 of the 5 crashing nodes several times over the last
few days. Therefore, problem may well have nothing to do with the automounter,
but insights will be gratefully accepted nonetheless. We are getting
crash dumps, but I'm not sure how to read them - insights here gratefully
accepted too, especially pointers to how to find info in documentation
(online).
Thanks.
Judith Reed
judith_at_npac.syr.edu
systems_at_npac.syr.edu
Received on Fri Feb 17 1995 - 13:02:43 NZDT