[Ocfs2-users] strange fencing behavior

[Sean Thon]

I have 10 servers in a cluster running Debian Etch with 2.6.26-bpo.2 
with a backport of ocfs2-tools-1.4.1-1
I'm using AoE to export the drives from a Debian Lenny server in the 
cluster.

My problem is if I mount the ocfs2 partition on the server that is 
exporting it via AoE it fences the entire cluster.  Looking at the logs 
exporting the ocfs2 partition doesn't give much information... Is this a 
known limitation?

Sep 24 14:06:20 storage0 kernel: [650570.916574] ocfs2_hb_ctl[12083]: 
segfault at 0 ip 7f20d9c97a90 sp 7fffe238cf98 error 4 in 
libc-2.7.so[7f20d9c1d000+14a000]
Sep 24 14:06:20 storage0 kernel: [650570.916932] ocfs2: Unmounting 
device (8,17) on (node 0)
Sep 24 14:07:04 storage0 kernel: [650622.608756] ocfs2_dlm: Nodes in 
domain ("E042658E558940E6B05EFE2DBA548DFD"): 0
Sep 24 14:07:04 storage0 kernel: [650622.608818] kjournald starting.  
Commit interval 5 seconds
Sep 24 14:07:04 storage0 kernel: [650622.609924] ocfs2: Mounting device 
(8,17) on (node 0, slot 0) with ordered data mode.
Sep 24 14:07:29 storage0 kernel: [650651.285363] ocfs2_hb_ctl[12160]: 
segfault at 0 ip 7f17160f3a90 sp 7fff1e7e93e8 error 4 in 
libc-2.7.so[7f1716079000+14a000]
Sep 24 14:07:29 storage0 kernel: [650651.285638] ocfs2: Unmounting 
device (8,17) on (node 0)
Sep 24 14:07:57 storage0 kernel: [650683.535472] ocfs2_dlm: Nodes in 
domain ("1C1694511D464F16971020CE53D45401"): 0
Sep 24 14:07:57 storage0 kernel: [650683.566388] JBD: Ignoring recovery 
information on journal
Sep 24 14:07:57 storage0 kernel: [650683.566388] kjournald starting.  
Commit interval 5 seconds
Sep 24 14:07:57 storage0 kernel: [650683.566388] ocfs2: Mounting device 
(8,18) on (node 0, slot 10) with ordered data mode.
Sep 24 14:07:57 storage0 kernel: [650683.566388] 
(12231,1):ocfs2_replay_journal:1149 Recovering node 10 from slot 0 on 
device (8,18)
Sep 24 14:08:00 storage0 kernel: [650687.138110] kjournald starting.  
Commit interval 5 seconds
Sep 24 14:08:00 storage0 kernel: [650687.268898] 
(12231,1):ocfs2_replay_journal:1149 Recovering node 2 from slot 1 on 
device (8,18)
Sep 24 14:08:02 storage0 kernel: [650690.547309] kjournald starting.  
Commit interval 5 seconds
Sep 24 14:08:03 storage0 kernel: [650690.581554] 
(12231,1):ocfs2_replay_journal:1149 Recovering node 3 from slot 2 on 
device (8,18)
Sep 24 14:08:05 storage0 kernel: [650693.934489] kjournald starting.  
Commit interval 5 seconds
Sep 24 14:08:05 storage0 kernel: [650693.983941] 
(12231,1):ocfs2_replay_journal:1149 Recovering node 4 from slot 3 on 
device (8,18)
Sep 24 14:08:08 storage0 kernel: [650697.403252] kjournald starting.  
Commit interval 5 seconds
Sep 24 14:08:08 storage0 kernel: [650697.478297] 
(12231,1):ocfs2_replay_journal:1149 Recovering node 5 from slot 4 on 
device (8,18)
Sep 24 14:08:11 storage0 kernel: [650700.979480] kjournald starting.  
Commit interval 5 seconds
Sep 24 14:08:11 storage0 kernel: [650701.049023] 
(12231,1):ocfs2_replay_journal:1149 Recovering node 6 from slot 5 on 
device (8,18)
Sep 24 14:08:14 storage0 kernel: [650704.525081] kjournald starting.  
Commit interval 5 seconds
Sep 24 14:08:14 storage0 kernel: [650704.569578] 
(12231,1):ocfs2_replay_journal:1149 Recovering node 7 from slot 6 on 
device (8,18)
Sep 24 14:08:17 storage0 kernel: [650708.610043] kjournald starting.  
Commit interval 5 seconds
Sep 24 14:08:17 storage0 kernel: [650708.642044] 
(12231,1):ocfs2_replay_journal:1149 Recovering node 8 from slot 7 on 
device (8,18)
Sep 24 14:08:20 storage0 kernel: [650712.306538] kjournald starting.  
Commit interval 5 seconds
Sep 24 14:08:20 storage0 kernel: [650712.366600] 
(12231,1):ocfs2_replay_journal:1149 Recovering node 9 from slot 8 on 
device (8,18)
Sep 24 14:08:24 storage0 kernel: [650716.491634] kjournald starting.  
Commit interval 5 seconds
Sep 24 14:08:24 storage0 kernel: [650716.587527] 
(12231,1):ocfs2_replay_journal:1149 Recovering node 1 from slot 9 on 
device (8,18)
Sep 24 14:08:27 storage0 kernel: [650720.039214] kjournald starting.  
Commit interval 5 seconds