HA doesn't just use network heartbeats, but also datastore heartbeats to avoid outages due to e.g. network timeouts.
I'm not familiar with StarWind VSAN, but it looks like the host still has access to the datasatore(s).
If you want to test HA, you may need to kill the host (unplug power cables, or do a hard switch off).
I tried your suggestion but it`s same
VM becomes unresponsive with no try to migrate to another host.
I`m having filing that I missed something in configuration steps but dont know what.
How did you configure Starwind VSAN? Is it configured for automatic failover, and do you have a witness to give the surviving node the majority?
It`s configured for automatic replication.
When I install VM on shared storage (provided by StarWind) it is visible in datastores on both ESXi hosts so I presume I`m failing on vsphere level
I think i found something last night that might be related to issue.
As I mentioned before, when I go to datastore list of each host I see shared data store (DS1)
I suppose it`s shared because I created DS1 from ESXi1 and after adapter rescan it automatically appeared on ESXi2, so it`s visible from both host
Also I can see iSCSI targets from both hosts
When I go to vCenter client and click on SWdatacenter > Datastores i can see local datastores from both hosts and shared datastore DS1
But when I go to SWcluster and click on Datastores tab there`s no datastores
Also, when I go to vSphere Availability >Edit Cluster Settings >Heartbeat Datastores tab there is no data stores to select.