In cluster view event, we force a release of the lock.

In some cases we don't want to force the release of the lock but make it not usable until be perfom a manual operation/call to make it available again.

This behaviour only makes sense when we allow read writes in split brain. We don't want the cache to go degraded and degrade all the locks, but we want the lock to become unavailable until we decide is safe to use it again