Ocfs2 users - Nov 2011 - 5 node ocfs2 cluster - what would be the best way to bring all these 5 nodes down without panic any nodes.

Dear ALL,

we have 5 node ocfs2 cluster, we have rebooted this servers recently, when
I rebooted 1st node all the 4 remaining nodes got panicked/rebooted and
these servers has more than 150 ocfs2, because of the unclean shutdown,
servers took more time to do fsck and mount the 150 ocfs2 FS.


and the split-brain algorithm for ocfs2 cluster on odd-numbered nodes are
below.


For an odd-numbered cluster with more than two nodes, the group of nodes
that can still communicate with each other and
whose number is greater than half the number of heartbeating nodes will
form quorum - all other nodes are fenced
(panic/restart).


could any one please suggest what would be the best way to bring all these
5 nodes down without panic any node and for shutdown,  each node its take
more than 20 minutes, because of 150 ocfs2 FS.


and is there any way, we can reduce the mounting/umounting time of these
ocfs2 FS.



Thanks

Veera.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://oss.oracle.com/pipermail/ocfs2-users/attachments/20111107/6bd1146f/attachment.html

I would also love to hear the answer to this.  We have servers panicing quite
frequently.  Makes for an extremely unstable environment, especially if a server
is under high load or if you need to reboot.


On Nov 7, 2011, at 4:43 AM, veeraa bose wrote:
> Dear ALL,
> 
> we have 5 node ocfs2 cluster, we have rebooted this servers recently, when
I rebooted 1st node all the 4 remaining nodes got panicked/rebooted and these
servers has more than 150 ocfs2, because of the unclean shutdown, servers took
more time to do fsck and mount the 150 ocfs2 FS.
> 
> 
> 
> and the split-brain algorithm for ocfs2 cluster on odd-numbered nodes are
below.
> 
> 
> 
> For an odd-numbered cluster with more than two nodes, the group of nodes
that can still communicate with each other and
> whose number is greater than half the number of heartbeating nodes will
form quorum - all other nodes are fenced
> (panic/restart).
> 
> 
> 
> could any one please suggest what would be the best way to bring all these
5 nodes down without panic any node and for shutdown,  each node its take more
than 20 minutes, because of 150 ocfs2 FS.
> 
> 
> 
> and is there any way, we can reduce the mounting/umounting time of these
ocfs2 FS.
> 
> 
> 
> 
> Thanks
> 
> Veera.
> 
> 
> 
> 
> _______________________________________________
> Ocfs2-users mailing list
> Ocfs2-users at oss.oracle.com
> http://oss.oracle.com/mailman/listinfo/ocfs2-users
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://oss.oracle.com/pipermail/ocfs2-users/attachments/20111201/366e58c2/attachment.html

Do you reboot you server, as it? happens if you umount your cluster ??
the procedure are very important. What software of cluster are you
using?? what do you see in you logs?

I hope that my server "not panicing quite frequently" :-)...


On Fri, Dec 2, 2011 at 1:57 AM, Tony Rios <tony at tonyrios.com>
wrote:
> I would also love to hear the answer to this. ?We have servers panicing
> quite frequently. ?Makes for an extremely unstable environment, especially
> if a server is under high load or if you need to reboot.
>
>
> On Nov 7, 2011, at 4:43 AM, veeraa bose wrote:
>
> Dear ALL,
>
> we have 5 node ocfs2 cluster, we have rebooted this servers recently, when
I
> rebooted 1st node all the 4 remaining nodes got panicked/rebooted and these
> servers has more than 150 ocfs2, because of the unclean shutdown, servers
> took more time to do fsck and mount the 150 ocfs2 FS.
>
>
> and the split-brain algorithm for ocfs2 cluster on odd-numbered nodes are
> below.
>
>
> For an odd-numbered cluster with more than two nodes, the group of nodes
> that can still communicate with each other and
> whose number is greater than half the number of heartbeating nodes will
form
> quorum - all other nodes are fenced
> (panic/restart).
>
>
> could any one please suggest what would be the best way to bring all these
5
> nodes down without panic any node and for shutdown,? each node its take
more
> than 20 minutes, because of 150 ocfs2 FS.
>
>
> and is there any way, we can reduce the mounting/umounting time of these
> ocfs2 FS.
>
>
>
> Thanks
>
> Veera.
>
>
>
> _______________________________________________
> Ocfs2-users mailing list
> Ocfs2-users at oss.oracle.com
> http://oss.oracle.com/mailman/listinfo/ocfs2-users
>
>
>
> _______________________________________________
> Ocfs2-users mailing list
> Ocfs2-users at oss.oracle.com
> http://oss.oracle.com/mailman/listinfo/ocfs2-users