Container garbage collection failed: operation timeout: context deadline exceeded

Issue

• At times my OpenShift nodes cannot be SSH'd and have high load

• ReadinessProbes failing with connection reset:

Apr 02 15:33:43 node01 atomic-openshift-node[47375]: I0402 15:33:43.472967   47375 prober.go:108] Readiness probe for "example-2-4g545_myproject(xxxxx):test" failed (failure): dial tcp 10.69.29.83:8080: getsockopt: connection refused

• Garbage collection and PLEG failing, context deadline exceeded:

Apr 02 15:25:15 node01 atomic-openshift-node[47375]: W0402 15:25:15.569563   47375 image_gc_manager.go:139] [imageGCManager] Failed to monitor images: operation timeout: context deadline exceeded
Apr 02 15:25:46 node01 atomic-openshift-node[47375]: I0402 15:25:46.729180   47375 container_manager_linux.go:371] Discovered runtime cgroups name: /system.slice/docker.service
Apr 02 15:26:08 node01 atomic-openshift-node[47375]: E0402 15:26:08.455719   47375 generic.go:193] GenericPLEG: Unable to retrieve pods: operation timeout: context deadline exceeded
Apr 02 15:26:13 node01 atomic-openshift-node[47375]: E0402 15:26:13.086673   47375 kubelet.go:1085] Container garbage collection failed: operation timeout: context deadline exceeded