The Engine_A.log at %VCS_HOME%/logs would report the attempt for manual online as below:
VCS INFO V-16-1-10299 Resource xxx (Owner: Unknown, Group: xyz) is online on Node-A (Not initiated by VCS)
Following an error for Concurrency Violation as below:
VCS ERROR V-16-1-10214 Concurrency Violation:Current Count increased above 1 for failover group
The issue occurs when the Cluster Administrator attempts to manually online a resource on a particular node (Node-A) in the cluster while the resource is already online on another node (Node-B) in the same cluster.
Considering that the group is originally offline on Node-A, VCS monitors each resource every 5 minutes to ensure all resources are offline.
If the VCS monitor finds a resource is online, it reports that the resource is online outside of VCS control and hence throws a concurrency violation error. It then runs a clean (similar to an offline operation) of the resource on the node where it should not be running.
In the above scenario, a clean will be done on Node-A.
VCS will auto-resolve the issue by running a clean against the resource on the node where it is not supposed to be online.