[Ocfs2-users] cluster member hangs during reboot

[Sunil Mushran]

ocfs2 tools has two init scripts, o2cb and ocfs2. The former starts the
cluster stack, the latter mounts the ocfs2 volumes listed in /etc/fstab.
These two scripts run back-to-back.

The scripts expect the network and block device to be ready by the
time they run. Sometimes that is not the case. In older rhel4 we had atleast
one network driver that required more time to initialize. In your case, I
suspect something similar issue with the block device.

Maybe experiment with the init script run order. As in, start with
delaying the ocfs2.init script.

Sunil

Joel Freeman wrote:
> Hi All,
> I inherited a 4-node ocfs2 cluster and recently 2 ocfs2 filesystems 
> were added to be use as temp tablespace. One of the four nodes 
> rebooted during the  creation of the tablespace and hanged at the 
> message below...and it just sits there. If I put the server into 
> rescue mode and comment out all the filesystems it boots up fine and 
> than I can mount the ocfs2 filesystem manuelly but it cannot do it by 
> itself during reboot...
>  
>
> msg
>
> *"EMCPUT MODULE LOADED.*
>
> *OCFS2 USER DLM KERNEL INTERFACE LOADED*..."
>
>  
>
>  
>
> The server is a HP Proliant DL580 running redhat linux AS 40 update 6
>
>  
>
> Any idea why this node cannot fully reboot and join the cluster?? Any 
> help is highly appreciated. Thanks!
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Ocfs2-users mailing list
> Ocfs2-users at oss.oracle.com
> http://oss.oracle.com/mailman/listinfo/ocfs2-users