OpenShift pod to pod internal network connectivity failing.

Issue

• Connections are being dropped when pods are trying to communicate to each other.
• I am seeing intermittent network connectivity between pods in the same project.
• Error: atomic-openshift-node[ multitenant.go:156] refcounting error on vnid 7666890
• Pods are restarting due to failed http health checks.
• Scaling multiple pods up in a project causes issues with pods being able to communicate.
• After scaling down a DeploymentConfig the communication between pods was lost and any attempt to reach the affected service or the pods failed with a timeout.