OpenShift pod consuming more CPU cores than its defined "limits" in YAML

Solution Verified - Updated -

Issue

  • We have some pods with the following resource limits and requests assigned; however we see that CPU millicores usage (registered in hawkular-metrics) are higher than 1000millicores assigned as "limits" for the pod. We have seen the POD running with a peak of 1892 millicores.
  • What is the behavior of OpenShift when the pod requires more cores than its "limits" defined in the YAML configuration?
  • Can a pod with limits still receive as many cores as it requires if there are cores available in the OpenShift node?
        resources:
          limits:
            cpu: "1000m"
            memory: "4096Mi"
          requests:
            cpu: "700m"
            memory: "3072Mi"
  • We expected to see no more than 1000 millicores used, so we need to know the real behavior in these scenarios.
  • OpenShift pods exceeded their millicore limit which caused a licensing issue with a 3rd party app.

Environment

  • OpenShift Container Platform 3.2+

Subscriber exclusive content

A Red Hat subscription provides unlimited access to our knowledgebase of over 48,000 articles and solutions.

Current Customers and Partners

Log in for full access

Log In