API error (500): Cannot stop error in Red Hat OpenShift Container Platform

Issue

• From time to time we see containers in an error state, where OpenShift/Kubernetes is not able to delete the container. In the events you see messages like the one below:

Feb 28 07:40:48 node.ose.net atomic-openshift-node[9165]: E0228 07:40:48.387666    9165 manager.go:1828] Error killing container xyz-ic"(id={"docker" "9793a976e1402d2ea5338c97cbf2de633a47bd6c9df0ee425f04dafb83097946"}) for pod xyz-ic-14-tedl0_xyz-inte(37b582ff-fce3-11e6-8218-1c98ec3025f0)": API error (500): Cannot stop container 9793a976e1402d2ea5338c97cbf2de633a47bd6c9df0ee425f04dafb83097946: [2] Container does not exist: container destroyed
Feb 28 07:40:48 node.ose.net atomic-openshift-node[9165]: E0228 07:40:48.387809    9165 pod_workers.go:138] Error syncing pod 37b582ff-fce3-11e6-8218-1c98ec3025f0, skipping: failed to "KillContainer" for xyz-ic" with KillContainerError: "API error (500): Cannot stop container 9793a976e1402d2ea5338c97cbf2de633a47bd6c9df0ee425f04dafb83097946: [2] Container does not exist: container destroyed"