Site-Network error while VMs are running

[rkabelich]

Hello,

what happens in case of a networkerror in an environment with two sites wile one site is no more reachable for vCenter (HA is enabled)? Do the VMs restart at the reachable site or do they stay running without network at the non reachable site?

Regards

Ronald

[hussainbte]

are you referring to HA across sites cluster or you are referring to a disaster recovery solution SRM(site recovery manager)

[rkabelich]

Thanks hussainbte

I refer to HA.

Chief asked, what would happen, if one site fails completely, network and storage (IBM SVC, mirrored Storage on both sites) too. I think in that case HA would restart the VM at another Host, so if the site becomes online there are two VMs running?

[hussainbte]

In case of a stretched cluster (stretched across multiple sites) when HA restarts VMs on the other hosts. once restarted the host owns the VMs.

in case of the esxi hosts on site coming back online as the VMs are running on the other hosts in the cluster they will not be powered on.. In fact it is not technically possible for the VMs to be powered on and running on 2 ESXi hosts in same cluster even if you want to.

[rkabelich]

Ok, but what happens, when all hosts keep running, while one site has no network- and also no storage-connection to vcenter? In my opinion vCenter would be able to restart the VMs on the hosts residing at the working site, because the lock-file ist no more locked on the mirrored datastore. If the other site becomes connected again there may run 2 of each VM?