For example, "Heartbeat is missing for VM for VirtualMachineName"
Now, if our SQL server went down this could happen cause vCenter would go offline (or better yet, if a datastore was offline and the VM died, etc.). But also if vCenter goes down on its own this happens. I really don't care if vCenter causes this error, but if SQL goes down we want to konw.
To put it another way, picture an island with a bridge leading to it. VeeamONE uses this bridge to check on the island. The island is my infrastructure, and the bridge is vCenter. When the bridge goes down VeeamONE assumes the island is not existent, when it is. So, to me I am having to assume maybe a bomb went off on the island and it is in fact down....
Now if a bomb went off I would get other alarms still. If a datastore goes offline for example. And in that event I want to know all the 100 emails I get about the VMs and their associated issues when a datastore kicks them offline. These two events (vCenter service crashing vs. datastore going offline) are completely different in priority and severity, yet they kick out very similar batch of errors.