>>>>> "Frank" == Frank van der Linden <fvdl@wasabisystems.com> writes:

    >> http://mail-index.netbsd.org/netbsd-bugs/2001/04/10/0000.html

    Frank> Looking at the PR, it looks like mbuf chain corruption
    Frank> somewhere. What is the activity common to all these
    Frank> systems? You have NFS enabled for server and client in the
    Frank> kernel, but are they actually used actively as such? If you
    Frank> avoid any NFS activity, will it still panic at reboot?

    Frank> (I'm trying to work out if this is an NFS problem or not)

2 other facts:

- There other i386 1.5.1_BETA machines around me which reboot fine.

- Once I rebooted a crashing machine into single user mode (changing
  some savecore options) and rebooting from there (== no nfs
  activity). -> The panic occurred, too!

Looking through the different kernel config files of crashing and non
crashing kernels/machines I cannot find one specific
option/device/service only common on crashing machines.

The only common fact of the crashing machines: They have 512 MB ram.
All the fine rebooting machines have less.

Another strange thing I observed the last hours is a kind of chain
reaction in panics of all three machines when loosing some network
services (nis, nfs, dns): E. g. one machine (providing such services
to the others) chrashes -> The other to panic, too. But it seems to be
limited within this machine trio, though other machines are effected
by this missing network service, too, but do not reboot.

-> Memory dependent more general network problems!? Strange...

Markus.