Performing a Reconfigure for vSphere HA operation on a primary node may cause an unexpected virtual machine failover
Yep, the Warning Message is that long! 😀
This is what I got from a customer, their vSphere Skyline Health was showing the above mentioned Warning:
*Sorry for the poor quality picture. 🙂
If you manually reconfigure the primary HA host for HA, this forces the secondary host to find a new primary host. The newly elected primary host places the VMs that were running on the old primary to an unknown power state. It waits for up to 10 seconds for a notification that the VMs on the old primary host are powered on and running.
If the old primary does not become secondary within that 10 seconds, the new primary host assumes that the VMs are down and attempts to restart them. To resolve this issue, increase the monitor period by switching over to the vSphere HA advanced settings of the affected cluster.
Add one of the two parameters under Option, depending on which vSphere version you are running and set the Value on 30.
- das.config.fdm.unknownStateMonitorPeriod -> for vSphere 7.0 U1 and later
- das.config.fdm.policy.unknownStateMonitorPeriod -> for pre vSphere 7.0 U1
After clicking OK, switch over to the Skyline Health view of your vCenter (not the vSAN Skyline Health, if you have a vSAN Cluster) and retest. The Warning should be gone.
This blog post was written in conjunction with the following KB article from VMware, which was also the solution for the customers case. https://kb.vmware.com/s/article/2017778
If you have any questions please leave it in the comments. 🙂