Issue

• Heapster pods are not reachable and give the error:

ERROR: [DClu4003 from diagnostic MetricsApiProxy@openshift/origin/pkg/oc/admin/diagnostics/diagnostics/cluster/metrics.go:89]
       Unable to access the metrics API Proxy endpoint /api/v1/proxy/namespaces/openshift-infra/services/https:heapster:/api/v1/model/metrics:
       (*errors.StatusError) an error on the server ("Error: 'dial tcp x.x.x.x:8082: getsockopt: no route to host'\nTrying to reach: 'https://x.x.x.x:8082/api/v1/model/metrics'") has prevented the request from succeeding
       The Horizontal Pod Autoscaler is not able to retrieve metrics to drive scaling.