Svr 16 and Hyper-V cluster - VM service monitoring

---

You have a two node Hyper-V cluster named Cluster1. A virtual machine named WIN16 runs on Cluster1. You need to configure monitiroing of VM1. The solution must move WIN16 to a different node if the Print Spooler service on VM1 stops unexpectedly.

Which tool should you use?

A. the clussvc.exe command
B. the cluster.exe command
C. the Computer Management snap-in
D. the configurhyperv.exe command
E. the Disk Management snap-in
F. the Failover Cluster Manager snap-in
G. the Hyper-V Manager snap-in
H. the Server Manager app.

Initially this question seemed pretty straight forward, but I hadn't actually done it so went to the lab... I built a two node cluster with servers named LAB-SVR1 and LAB-SVR2. Created a guest server 16 VM named WIN16 just like in the question. Enabled the firewall rule to allow for "Virtual Machine Monitoring" and enabled the Printer Spooler service monitor from the Failover Cluster Manager.

However, getting this VM to failover to the second node upon stopping the spooler service has turned out to be a real challenge.

Here is the firewall exception:

Here is how service monitoring is configured:

Here is the process I follow to kill the spooler service from within the guest VM:

Here is an example of the VM restarting and the related event log entry:

What is driving me nuts is that the VM never fails over to LAB-SVR2.. No matter how many times the VM is rebooted and the spooler service is failed. Going back to the main question, at the top of this post, I believe the answer is "F" for the Failover Cluster Manager, however it's a tad embarrassing that I can't figure this out.

I did discover these nobs in the Failover Cluster Manager, but no amount of tweaking ever gets the VM to live migrate to a different node. Notice the different set of options for "Virtual Machine" vs "Virtual Machine Configuration" WIN16 Properties...

Adam,
You are definitely on the right track! (F) would be my answer as well. Kudos for setting up your own lab! The cluster resilency setting kick in as kind of the last ditch effort. The VM's local Service Control Manager tries to handle the problem first. Please check on the Recovery tab of the Print Spooler Service from within the VM to see if the first and second failure are configured in a conflicting way. If it states "Restart the Service" then test that and see if the spooler service kicks back on. Try changing it to "Take No Action" and see if you get a failover. Don't forget we do have Practice Labs set up for use if you want to try it in another environment. I hope this helps! Happy studies.
Kindly,
Cherokee

Thanks Cherokee! I changed this "First Failure" and "Second Failure" option within the guest VM to "Take no action" and caused the service to fail again. This time it restarted the VM first, but upon failing the service a second time the VM was finally migrated over to the other node!

Last question.. Is there an option within the Cluster Config that would move the VM to a different node sooner? Or is this the only and default behavior? Rebooting the guest first, then live migrating upon a second failure..

Regards,
Adam Tyler

Adam,
I wouldn't expect the VM to restart. Can you repeat this action to verify there is a correlation with these settings rather than an external variable?
Thanks,
Cherokee