Which of the following statements describes fencing resources in a high availability cluster?

A.
Fencing is the automated stop and restart of cluster services that suffer from software errors.

B.
Fencing is the disconnection of a failed cluster node from any other cluster resources.

C.
Fencing is the forwarding of incoming network connections to backend servers.

D.
Fencing is the assignment of services to different cluster nodes to avoid interference of the services.

E.
Fencing is the accounting and limitation of disk usage on shared storage.