[Ocfs2-users] problem on one member caused the other member to panic.

[mwoods at fnal.gov]

I have 2 system cluster  running  ocfs2 1.2.5-1.    An OS kernel patch was
errantly applied to one of the machines without the ocfs module being
applied to the new kernel. When the errant machine rebooted, it of course
couldn't mount and ocfs2 file systems. Totally expected and easily
rectified. What was unexpected was, this caused the following message on
the other cluster member and caused it to reboot.

I am assuming this happened because I have something misconfigured.

What caused a problem on one system to panic the other systems and how can
I protect this from happening in the future.

Thanks.


Jul 12 06:04:49 finps03 kernel: (0,0):o2net_idle_timer:1418 here are some
times that might help debug the situation: (tmr 1184238259.664319 now
1184238289.659909 dr 1184238259.6643
03 adv 1184238259.664349:1184238259.664353 func (89487f70:1)
1184238259.664324:1184238259.664339)
Jul 12 06:04:49 finps03 kernel: o2net: no longer connected to node
xxxx.fnal.gov (num 0) at ttt.sss.xxx.yyy:7777
Jul 12 06:04:49 finps03 kernel: (6006,10):ocfs2_process_vote:494 ERROR:
message to node 0 fails with error -112!
Jul 12 06:04:49 finps03 kernel: (6007,5):dlm_send_proxy_ast_msg:459 ERROR:
status = -112
Jul 12 06:04:49 finps03 kernel: (6007,5):dlm_flush_asts:600 ERROR: status =
-112