Uploaded image for project: 'OpenShift Bugs'
  1. OpenShift Bugs
  2. OCPBUGS-65664

IngressOperator not exposing some metrics for existing IngressController after Operator restart

XMLWordPrintable

    • None
    • False
    • Hide

      None

      Show
      None
    • None
    • Moderate
    • None
    • None
    • Rejected
    • NI&D Sprint 279, NI&D Sprint 280
    • 2
    • In Progress
    • Bug Fix
    • Hide
      Cause: Cluster Ingress Operator pod restarts with existing IngressController resources in Available or Degraded status caused the ingress_controller_conditions metric to be non-existent from the Operators /metrics endpoint.

      Consequence: End user could not monitor IngressController status after IngressOperator pod restart.

      Fix: IngressControllerConditions metric is now set at every reconcile, even if there no Ingress controller status update.

      Result: IngressController conditions metric is now consistently available even immediately after an Ingress Operator pod restart, improving the monitoring of IngressController status.
      Show
      Cause: Cluster Ingress Operator pod restarts with existing IngressController resources in Available or Degraded status caused the ingress_controller_conditions metric to be non-existent from the Operators /metrics endpoint. Consequence: End user could not monitor IngressController status after IngressOperator pod restart. Fix: IngressControllerConditions metric is now set at every reconcile, even if there no Ingress controller status update. Result: IngressController conditions metric is now consistently available even immediately after an Ingress Operator pod restart, improving the monitoring of IngressController status.
    • None
    • None
    • None
    • None

      This is a clone of issue OCPBUGS-61508. The following is the description of the original issue:
      —
      Description of problem:

      When an IngressOperator pod is restarted and there are already IngressController resources in Available or Degraded status, the ingress_controller_conditions metric is non-existent from the Operators /metrics endpoint.

      Version-Release number of selected component (if applicable):

      How reproducible:

      Create a cluster with at least the default IngressController
      Restart the Ingress Controller (it could happen if you need to force the cloud provider credentials to be updated after changing something in the IAM role)

      Steps to Reproduce:

          1.Make a cluster with the default ingress controller
    2.Perform an operation that will cause the cloud provider credential to be updated
    3.Restart the Operator pod to force the operator to use immediately the new credentials

      Actual results:

      The ingress_controller_conditions metric is not available.

      Expected results:

      The ingress_controller_conditions metric is available.

      Additional info:

      This is a hypershift cluster, but I think it shouldn't matter.

              dsalerno@redhat.com Davide Salerno
              jbranham.openshift Josh Branham
              None
              None
              Shudi Li Shudi Li
              None
              Votes:
              0 Vote for this issue
              Watchers:
              5 Start watching this issue

                Created:
                Updated: