vmware 5.5 / server 2012 / failover clustering

cmatchett used Ask the Experts™
Hi Experts,

We have been asked to do some work on a file and print failover cluster that is using raw device mappings for its storage.

Part of this work will involve testing that the failover of the file server resource actually works as expected.

Is there any reason as to why it might fail or any caveats that somebody should be aware of?
Watch Question

Do more with

Expert Office
EXPERT OFFICE® is a registered trademark of EXPERTS EXCHANGE®
Andrew Hancock (VMware vExpert / EE Fellow)VMware and Virtualization Consultant
Fellow 2018
Expert of the Year 2017
The biggest issue you may come across is whether your SAN supports SCSI-3 Reservations, for the RDM, but most SANs supported this function.

The new Failover Cluster Wizard in WIndows 2012, verifies the requirements and compatibility of RDMs for nodes, and tests failover as part of the wizard.

So, it's very straight foward.

Make sure both nodes are on different ESXi hosts, and if using DRS, create an anti-affinity/affinity rule to ensure that both nodes are on different hosts....

BUT.....why A Cluster...

Why not use a single Print Server and File Server on VMware, and use VMware HA ?

Same function.....as far as availability......if a host fails, VM will be restarted automatically.....

Setup for Failover Clustering and Microsoft Cluster Service - ESXi 5.0/vCenter Server 5.0


or is this requirement for Patching?


Dunno why i said file and print, it's only a file server cluster.  This cluster is already configured and operational, we have been asked to test it.

The only reason i can think of that somebody would want a cluster is for patching reasons.  Like you say, it's already highly available because it is virtual machine that is protected by HA.

A highly available print server in MS terms now is a single virtual machine that is highly available.  No longer any support for clustering of the printer spooler.
VMware and Virtualization Consultant
Fellow 2018
Expert of the Year 2017
Two methods for testing, as testing is already now completed in the Creation through the wizard.

1. Power OFF the VM (not shutdown) Power OFF! This is a hard failure, and Node 2, should take over when failover occurs.

2. Other Test, is to POWER OFF the HOST, the VM is on!

Do more with

Expert Office
Submit tech questions to Ask the Experts™ at any time to receive solutions, advice, and new ideas from leading industry professionals.

Start 7-Day Free Trial