How does HA cluster Host behave when one Host fails that holds a MSCS CIB (MS Cluster on one Host)

[nikkita]

Hello everyone,

i got a quite simple question hoping to get the answer here in the vmware community, before i start testing to get an concrete answer.

My Situation:

I am running a vSphere 5.1 U2 ESXi 2 Host HA Cluster in my vCenter Datacenter where more HA Clusters reside.

This 2 Host ESXi Cluster i have especially provided for Microsoft failover clustering VMs.

Some of my MSCS run as CAB (with RDM) and one as CIB (with vmdk). Further i have DRS affinity and non affinity rules defined.

Now my questions:

Depending on the MSCS as Cluster in a Box (CIB) Scenario i got two VMs on one ESXi Host.

What happens if this ESXi Host fails, which ist hosting the CIB VMs? (Remember i got a second ESXi Host!).

If the ESXi Host1 fails, the MSCS CIB VMs will also fail, this is clear, but will the CIB VMs start on the second remaining ESXi Host2 automaticly or not?

I hope someone can easily answer this question for me, i can only suppose it is so and would be happy if anyone can confirm.

THX in advance

nicole

[weinstein5]

Welcome to the Community - As long as the virtual disks of the two VMs that comprise the CIB (which I assume means Cluster in a Box) VMs are on shared storage both VMs should restart in the even of an HA event -

[nikkita]

Yes, the MS Cluster-in-a-Box VMs have shared disks, actually FC LUNs connected to both of the ESXi Hosts in the HA Cluster, on those LUNs the shared disks as vmdks reside.

So, i understand in case of a ESXi Host failover these VMs will start again on the redundant ESXi Host. Fine, eventhough i still prefer the Cluster across Boxes (CAB) with RDMs

Thank you weinstein5 for answering.

[ScreamingSilenc]

Definitively HA is going to restart the VM's on Host2 unless until you haven't explicitly changed VM Response setting under vSphere HA settings.

[nikkita]

thx Mustafa, i guess next week i will find a gap and test the scenario