HA isolation address alarm after changed Default gateway

[PonF]

We have changed the default gateway address on our vSphere environment (vcenter and ESXI hosts). Now we have got an alarm on all hosts saying "vSphere HA agent on this host could not reach isolation adress: xxx.xxx.xxx.xxx". But the specified address is an old address that no longer exist.

I have read that HA heartbeat uses the management network default gateway to ping. So why does it still say the old address? On all esxi hosts we have changed the default gateway under Host->Manage->Networking->TCP/IP configuration->System stacks->Default->Routing->VMkernel gateway:.

Is there another place where the old address could still be in? Or do we have to update something so the HA will take the new default gateway as heartbeat address?

We have vmware ESXI 6.0.0, 7967664.

[a_p_]

A quick guess: Did you already run "Reconfigure for HA" after changing the gateway addresses?

André

[depping]

use "reconfigure" or simply disable/enable HA.

[PonF]

Thank you both, it looks like "Reconfigure for vSphere HA" removes the alarm, so I guess it is okay after that. I have tested on two of the eight servers.

The thing is I saw this option earlier but when I use google on this option nothing appears that explain what this does!? Most results is about an bug/error when doing this on the primary node (I have not tested on my primary node yet).

Why is there no information about what this option exactly do? I press reconfigure now and the task takes 20 seconds and then it is done.

[depping]

because it simply "reconfigures HA" on the host I guess no one felt it was useful to document it, let me put that request in. So it takes all the cluster wide settings, and reconfigures those on the host. In your case you changed the gateway, the gateway is used for isolation detection. the isolation address is statically configured on the host, clicking "reconfigure" reconfigures that.