Hi,
In our environment, we have an OpenStack Ocata controller and Clustered Microsoft Hyper-V nodes (Windows Server 2016 Datacenter).
Using the live migration command for OpenStack works fine: the instance is migrated and the database is updated with the information of the new host in which the instance is running. When the instance is migrated directly in the Failover Cluster, the process is completed successully in the hypervisor, but the information isn't updated in the database. The following message is logged when syncing power state in Hyper-V:
2019-04-17 11:25:34.697 23328 WARNING nova.compute.manager [req-553ae0b0-17ac-46bd-8842-4b879d93cc9b - - - - -] While synchronizing instance power states, found 27 instances in the database and 37 instances on the hypervisor.
With that, during the execution of some actions (resize migrate, for example), OpenStack calls the wrong hypervisor, occasioning failure.
There is any configuration/command used to call an information update about instance location?
Thanks in advance.
