[Ocfs2-users] OCFS2 cluster sync problem
Gerrit Bruchhäuser
gbruchhaeuser at nero.com
Thu Jan 11 06:10:39 PST 2007
Dear All,
The two nodes forming my cluster do not see each others file-system changes.
Just as if each node would be mounting a cluster unaware file-system! My
configuration is identical on both the nodes and looks like:
,----[ # cat /etc/ocfs2/cluster.conf ]
| cluster:
| node_count = 2
| name = ocfs2
|
| node:
| ip_port = 7777
| ip_address = 192.168.4.130
| number = 1
| name = gbgen01
| cluster = ocfs2
|
| node:
| ip_port = 7777
| ip_address = 192.168.4.161
| number = 2
| name = gbgen02
| cluster = ocfs2
`----
On the first node (gbgen01) the device /dev/sdd2 is formatted with:
$> mkfs.ocfs2 -L share -N 2 /dev/sdd2
I exported the device via 'aoe' driver and could mount the device on both
nodes with the same command:
$> mount -t ocfs2 -L share /mnt/ocfs2
But fs changes became nor visible on the other node, until I unmount and mount
the share again.
Example:
,----[ gbgen01 ocfs2 # mount | grep ocfs2 ]
| none on /dlm type ocfs2_dlmfs (rw)
| /dev/sdd2 on /mnt/ocfs2 type ocfs2 (rw,_netdev,heartbeat=local)
`----
gbgen01 ocfs2 # cd /mnt/ocfs2
gbgen01 ocfs2 # touch a
gbgen01 ocfs2 # ls -l
total 0
,----[ gbgen02 ocfs2 # mount | grep ocfs2 ]
| none on /dlm type ocfs2_dlmfs (rw)
| /dev/etherd/e0.0 on /mnt/ocfs2 type ocfs2 (rw,_netdev,heartbeat=local)
`----
gbgen02 ocfs2 # cd /mnt/ocfs2
gbgen02 ocfs2 # touch b
gbgen02 ocfs2 # ls -l
total 4
-rw-r--r-- 1 root root 0 Jan 11 15:09 b
drwxr-xr-x 2 root root 4096 Jan 11 14:08 lost+found
On both servers, 'o2cb status' does not report anything indicating the problem
(everything active and loaded ok). I have tried the following kernel
versions, but could see the same result no matter which I used:
vmlinuz-2.6.17-hardened-r1
vmlinuz-2.6.18-gentoo-r6
vmlinuz-2.6.19.1
Is there something very basical I missed during my configuration?
Thanks for your help in advance!
Best Regards,
Gerrit
More information about the Ocfs2-users
mailing list