Let start this off with a recent issue I had in my VSAN home lab.
After a recent power outage i was surprised to see that HA had not powered on my VM’s after the hosts were back up. I opened vSphere to each host and found the VM’s were inaccessible. I scanned for new datastores and VMFS volumes but that did not help. I also rebooted all hosts but they all still came back up with inaccessible VM’s.
I did not have access to vCenter because it was on the VSAN datastore so I was not able to do any troubleshooting steps with RVC but I did Putty into each host to get details on the VSAN cluster status. I do keep Veeam on a local datastore separate from the VSAN datastore so i can do i quick restore of vCenter from NAS if needed.
Output of “esxcli vsan cluster get” shows that status for this host is healthy but it is the only host in the Sub-Cluster. What we should see is the Sub-Cluster Membership UUID for the other 2 host listed under Sub-Cluster Member UUID’s. The output on all the hosts showed the node state to be “Master.” This would indicate a network issue. I also looked at “ESXCLI VSAN NETWORK LIST” to make sure it had valid Multicast Addresses listed on each host. I rebooted my switch and with in a minute of it coming back up I saw my VM’s accessible again and checked the cluster status on each host and saw the UUID’s of the other hosts listed under members.
The above shows all the cluster members listed under Sub-Cluster Member UUIDs with VSAN operating correctly and VM’s available. Your problem my not be as simple as a reboot a switch but at least now you can focus on Multicasting as the issue.