Re: CMAN related questions

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Moving this thread to corosync mailing list.

On 04/12/2014 03:03 PM, Konstantin Ponomarenko wrote:
> 
> On Thu, Apr 10, 2014 at 12:08 AM, Fabio M. Di Nitto <fdinitto@xxxxxxxxxx
> <mailto:fdinitto@xxxxxxxxxx>> wrote:
> 
>     Just use existing set of applications. look at pacemaker + corosync. You
>     might not need quorum disk at all.
> 
>  
> But how I can dill with split-brain situation in case of 2 nodes cluster?
> 
> It seems to be the solution for split-brain: 
> 
> .B auto_tie_breaker: 1
> .PP
> Enables Auto Tie Breaker (ATB) feature (default: 0).
> .PP
> The general behaviour of votequorum allows a simultaneous node failure up
> to 50% - 1 node, assuming each node has 1 vote.
> .PP
> When ATB is enabled, the cluster can suffer up to 50% of the nodes failing
> at the same time, in a deterministic fashion. The cluster partition, or the
> set of nodes that are still in contact with the node that has the lowest
> nodeid will remain quorate. The other nodes will be inquorate. 
> 
> But, say, the 1st node (with with lowest nodeid) went down. The 2nd node
> will not be active because the lover nodeid in cluster is 1.
> Am I correct?
> 
> The main problem for me is the split brain situation.

As it is for many other users :)

> Maybe you could also point me to some feature in "votequorum" which will
> use quorum disk or at least will solve the split brain problem.
> I've read the man pages about "votequorum", but the only thing I found
> is "auto_tie_breaker". But as I said, it seems not ti be the solution in
> split brain situation.

there is no qdiskd for votequorum yet.

You also need to differentiate between:

1) protect data integrity
2) continue providing services from the cluster.

auto_tie_breaker will surely garantee that in case of 50/50 split, only
node can operate. If that node goes down, the other one cannot make
damage to the data set.

Even with qdiskd, in a 50/50 split, you achieve #1 all the time and if
you have a very good configuration also #2.

Fabio

> 
> My cluster has SAS connection to the shared drive.
> 
> I will be very appreciated for the help.
> 
> Many thanks,
> Kostya
> 
> -- 
> Konstantin Ponomarenko

_______________________________________________
discuss mailing list
discuss@xxxxxxxxxxxx
http://lists.corosync.org/mailman/listinfo/discuss




[Index of Archives]     [Linux Clusters]     [Corosync Project]     [Linux USB Devel]     [Linux Audio Users]     [Photo]     [Yosemite News]    [Yosemite Photos]    [Linux Kernel]     [Linux SCSI]     [X.Org]

  Powered by Linux