Is it always the mount using "node slot 1" that fails? If so, the jbd
superblock
may be corrupted for that slot.
Grow the journal by, say, 1MB. It will reinitialize the JBD superblock
for all
the slots. Either that or just reformat the device.
To see the size of the existing journal, do:
# echo "ls -l //" | debugfs.ocfs2 -n /dev/sda1 | grep journal
36 -rw-r--r-- 1 0 0 67108864
21-Jun-2006 11:58 journal:0000
37 -rw-r--r-- 1 0 0 67108864
21-Jun-2006 11:58 journal:0001
38 -rw-r--r-- 1 0 0 67108864
21-Jun-2006 11:58 journal:0002
39 -rw-r--r-- 1 0 0 67108864
21-Jun-2006 11:58 journal:0003
The grow the journal, do:
# tunefs.ocfs2 -Jsize=65M /dev/sdX
Zachary Williams wrote:> I am attempting to setup a 2 node ocfs2 cluster. At this point, I
> have the latest 1.2.1 version of the tools on both nodes. They are
> not running identical kernels (one is 2.6.16.18 <http://2.6.16.18>,
> the other is 2.6.17.1 <http://2.6.17.1>) both are using the kernels
> built in OCFS2 modules, not using from source.
>
> I can mount my iscsi volume on either node individually, but when I
> attempt to mount two nodes, I get the following error. (To confirm, I
> have 2 nodes setup in the config file, and the filesystem set to a
> maximum of 4 nodes)
>
> The error is JDB: no valid journal superblock found
>
> I have searched high and low for this, but wasn't able to come up with
> anything as to why I get this. This error will occur on either node.
>
> (3509,0):o2net_set_nn_state:415 accepted connection from node bsp (num
> 1) at 10.1.1.11:7777 <http://10.1.1.11:7777>
> (3575,0):ocfs2_initialize_super:1326 max_slots for this device: 4
> (3575,0):ocfs2_fill_local_node_info:1019 I am node 0
> (3575,0):__dlm_print_nodes:377 Nodes in my domain
> ("E09A0D90C8454749B81E9754438611B8"):
> (3575,0):__dlm_print_nodes:381 node 0
> (3575,0):__dlm_print_nodes:381 node 1
> (3575,0):ocfs2_find_slot:267 taking node slot 1
> JBD: no valid journal superblock found
> (3575,0):ocfs2_journal_wipe:814 ERROR: status = -22
> (3575,0):ocfs2_check_volume:1581 ERROR: status = -22
> (3575,0):ocfs2_mount_volume:1087 ERROR: status = -22
> ocfs2: Unmounting device (8,16) on (node 0)
> (3577,0):o2net_set_nn_state:400 no longer connected to node bsp (num
> 1) at 10.1.1.11:7777 <http://10.1.1.11:7777>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Ocfs2-users mailing list
> Ocfs2-users at oss.oracle.com
> http://oss.oracle.com/mailman/listinfo/ocfs2-users
>