ceph: handle CephCluster stopCleanupCh better

Investigate a possible memory leak around CephCluster cleanup's
stopCleanupCh channel. Add more close() statements and comment that Go's
garbage collector is able to close the channel when there are no more
referents to the channel.

Signed-off-by: Blaine Gardner <blaine.gardner@redhat.com>

pkg/operator/ceph/cluster/controller.go

@@ -340,9 +340,15 @@ func (r *ReconcileCephCluster) reconcileDelete(cephCluster *cephv1.CephCluster)
 		if doCleanup {
 			cephHosts, err := r.clusterController.getCephHosts(cephCluster.Namespace)
 			if err != nil {
+				close(stopCleanupCh)
 				return reconcile.Result{}, cephCluster, errors.Wrapf(err, "failed to find valid ceph hosts in the cluster %q", cephCluster.Namespace)
 			}
+			// Go will garbage collect the stopCleanupCh if it is left open once the cluster cleanup
+			// goroutine is no longer running (i.e., referencing the channel)
 			go r.clusterController.startClusterCleanUp(stopCleanupCh, cephCluster, cephHosts, monSecret, clusterFSID)
+		} else {
+			// stop channel not needed if the cleanup goroutine isn't started
+			close(stopCleanupCh)
 		}
 	}
 
@@ -352,7 +358,6 @@ func (r *ReconcileCephCluster) reconcileDelete(cephCluster *cephv1.CephCluster)
 		if err != nil {
 			// If the cluster cannot be deleted, requeue the request for deletion to see if the conditions
 			// will eventually be satisfied such as the volumes being removed
-			// TODO: is it a memory leak to not close the channel in other cases?
 			close(stopCleanupCh)
 			return response, cephCluster, errors.Wrapf(err, "failed to clean up CephCluster %q", nsName.String())
 		}