What is the expected behavior of the VMs residing on that host when a controller VM becomes unavailable?
What is the expected behavior of the VMs residing on that host when a controller VM becomes unavailable?
When a controller VM (CVM) becomes unavailable, the expected behavior of the VMs residing on that host is that the host will automatically redirect I/O and the VMs will continue running. This ensures continuity of operations and high availability of services without significant disruptions, leveraging built-in mechanisms designed to handle such failures gracefully.
A is correct, thanks to Happy process (ha.py)
Definitely A. This is how our VMs behave in production when a CVM is off.