0

To avoid a single point of failure on the backend SAN system, can I place one quorum witness disk on each site and have the cluster use both?

  • How can a Quorum help to avoid a single point of failure? Does that even make sense? – Daniel Jan 12 '16 at 14:11
3

No you can't. The whole point of witnesses is to avoid a split brain situation.

2

The solution ended up being a file server witness placed at a third secure location with direct links to each datacenter to maximize redundancy.

0

If we just have 2 x nodes in a failover cluster then if they can't communicate both will assume the role of the cluster - the Split Brain Morten mentions. the quorum disk fixes this as the node that can see the quorum disk wins as it gets 2 votes compared with the other node that just gets one. If your quorum disk fails you'll still have your two node cluster so don't worry about that unless a node fails at the same time. You do mention this disk is on a SAN so that does seem unlikely. More on this here http://blogs.msdn.com/b/clustering/archive/2011/05/27/10169261.aspx

  • My concern is that due to severe licensing costs, we do not presently have automatic failover of disks between our two san hosts, meaning that if the site with the quorum disk goes down, one node and the quorum witness both go down at the same time. I would like to avoid any single point of failure like that. – Morten Nilsen Jan 22 '16 at 13:29
  • I am currently thinking of adding a file server witness in a third location as a witness. combined with a quorum disk witness, that should prevent any spof scenario, I hope. – Morten Nilsen Jan 22 '16 at 13:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.