Basic Informations:
- Host OS: Windows 2003 Enterprise Edition (32 bit)
- VMWare Server Console: 1.0.1build-29996
- Guest OS: Oracle Enterprise Linux 4，Kernel 2.6.9-42.0.0.0.1.ELsmp
- Oracle: Oracle 10g (10.1.0.2)
- OCFS2: OCFS2 1.2.3 Tue Oct 24 16:09:36 PDT 2006 (build 6b798aaadf626d3b137c3952809b2f38)
- Oracle Clusterware Release 2 (10.2.0.1.0) for Linux x86
Other Informations:
- OCFS2 monitors the public network
Assume we have a cluster with 2 nodes named rac1 and rac2. OCFS2 and hangcheck-timer are configured.
1. Network Adapters' Failure
(1). Disconnect rac1's Private Network Adapter...
rac2 will reboot itself.
Explanation: When rac1's private network is disconnected, because CRS use private network for heartbeat messages, so rac2 is rebooted.
(2). Disconnect rac1's Public Network Adapter...
rac2 will go "kernel panic". The error messages are:
(4.0): o2quo_make_decision: 143 ERROR: fencing this node because it is connected to a half-quorum of 1 out of 2 nodes which doesn't include the lowest active node 0
(4.0): o2hb_stop_all_regions: 1908 ERROR: stopping heartbeat on all activate regions.
kernel panic - not syncing: ocfs2 is very sorry to be fencing this sytem by panicing
If you want the system to reboot itself when it meets "kernel panic", do like this:
Q03 How do I configure my system to auto-reboot after a panic?
A03 To auto-reboot system 60 secs after a panic, do:
# echo 60 > /proc/sys/kernel/panic
To enable the above on every reboot, add the following to
/etc/sysctl.conf:
kernel.panic = 60
But still I don't think it's a properl way to deal with this problem since it's really useless to panic and then reboot rac2 when rac1's public network adapter has been disconnected.
阅读(2009) | 评论(0) | 转发(0) |
