Hosts

The hosts are the nodes which make the cluster. Currently they have cos installed on them. Navigating to the hosts tab in the Harvester UI shows the details of the nodes.

Negative Test Case Behavior

• If a host goes down
  • The VMs on the host stay in active, or whatever status they were in
  • The host goes to unavailable, but the VM status will not update
  • You can not migrate the VM to a new host, but sometimes a VM will migrate when the host turns back on
• If a host reboots
  • The VM shows as active
  • The host goes to unavailable then to available again as it boots up
• If a host with a replica goes down and the number of replica goes below 3?
• If a host goes into maintenance mode
  • It enters into an entering maintenance mode status
  • All the VMs migrate off to other nodes
  • The node goes into maintenance mode
  • If you take it out of maintenance mode you can migrate the VMs back
• I’ve noticed that the VMs sometimes migrate when you’re turning a powered off node on. That might be some queued migration tasks. I’ll keep an eye on it
• If you delete a host while it is unavailable and there are VMs on it
  • The host is removed
  • The VMs stay in running with no node assigned for a few minutes, then they are started on a new node
  • If you start a migration with the node down but before you delete it then it goes into a fail state and it doesn’t seem to be able to be started on a new node, even if you stop and then start it.

---