Host cannot communicate with all other nodes in the Virtual SAN enabled cluster

[Paul_Sheard]

I get this error after applying the latest patch :-    http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=213511...

So all hosts are now at the revision :  VMware ESXi, 6.0.0, 3380124

After the reboot I get the error about Host cannot communicate with all other nodes in the virtual san enabled cluster..

However the VSAN health check is green, "esxcli vsan cluster get" shows all 6 members of cluster ok..  and if I reboot a host it comes up ok without the error.. then if I reboot another host this will come up without the error..  but then the host that I rebooted prior to this one will then show the error...?  So I can never get more than 1 host without the error after a reboot.

I've checked multicast which checks out ok, in fact all checks contradict the fact that why I get this error..

Has anyone got any ideas? Could this be the latest patch..

Paul.. 

[zdickinson · Accepted solution]

Good morning, this has been covered here:  6.0 U1b - Hosts cannot communicate  Thank you, Zach.

[vicenac]

I had the same situation.

I found that one of the hosts had wrong DNS settings. Once DNS settings were correct on all the hosts, the errors went away.

[SugeeB]

Hi Guys,

This issue has been fixed in VSAN 6.2

There is a workaround for you guys who are fan of VSAN 6.0

workaround 1: Take-off just one ESX host from the cluster (Should be in maintenance mode) wait for 2 minutes and then add that host back to the cluster VPXA services will start sync and the exclamation mark will be remove from all the host in that cluster.

workaround 2: Putty to ESXi host in the cluster and restart the vpxa service (Need to run the command for all the Hosts belongs to your cluster)

# /etc/init.d/vpxa restart

VMware KB: Following maintenance, hosts in a Virtual SAN cluster may erroneously report the error: H...

- Sugee -

[rrbnc]

For posterity, I ran into this error message in 6.5 because one host had IPv6 enabled after it was rebuilt, while all of the other hosts had IPv6 disabled. I disabled IPv6 on the one host and rebooted, and that solved the problem.