5.0A cluster crashed after disks switched to redundant controller

From: Raymond Pang <wmpang_at_se.cuhk.edu.hk>
Date: Wed, 3 Jan 2001 17:12:02 +0800

Hi Managers,

Our testing single member cluster (Tru64 5.0A) keeps crashing after
I issued the command "set this prefer=(0,1,2,3)" on the HSZ40 during
the booting process of the cluster. Since then, I cannot even boot it up
again to the single user mode. It keeps complaining about the inconsistent
of the hardware components databases and the cluster database.
Any ideas or solutions?

Thanks in advance.
Raymond


clsm: checking for peer configurations
clsm: no peer data available
clsm: initialized
hwc: Your system has to be rebooted because your
     hardware components databases were out of
     synchronization with the cluster database and
     a reboot is required to bring them back into
     synchronization

trap: invalid memory read access from kernel mode

    faulting virtual address: 0x0000000000000028
    pc of faulting instruction: 0xfffffc000079c2c0
    ra contents at time of fault: 0xfffffc00007faae0
    sp contents at time of fault: 0xfffffe04793cf850

panic (cpu 0): kernel memory fault
drd: Clean Shutdown
Received on Wed Jan 03 2001 - 09:14:04 NZDT

This archive was generated by hypermail 2.4.0 : Wed Nov 08 2023 - 11:53:41 NZDT