Uploaded image for project: 'OpenShift Bugs'
  1. OpenShift Bugs
  2. OCPBUGS-33181

Fix audit-logs container to respect SIGTERM


    • Icon: Bug Bug
    • Resolution: Unresolved
    • Icon: Undefined Undefined
    • None
    • 4.15, 4.16
    • HyperShift
    • None
    • No
    • False
    • Hide


    • Release Note Not Required
    • In Progress

      Description of problem:

          The audit-logs container for kas, oapi and oauth apiservers does not terminate within the `TerminationGracePeriodSeconds` timer. This is due to the container not terminating when a `SIGTERM` command is issued.
      When testing without the audit logs container, oapi and oath-apiserver terminates within a 90-110 second range gracefully. The kas does not terminate with the container gone and I have a hunch that it's the konnectivity container that also does not follow `SIGTERM` (I've attempted 10 minutes and still does not timeout).
      So this issue is to change the logic for audit-logs to terminate gracefully and increase the TerminationGracePeriodSeconds from the default of 30s to 120s.

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


      How reproducible:


      Steps to Reproduce:

          1. Create a hypershift cluster with auditing enabled
          2. Try deleting apiserver pods and watch the pods being force deleted after 30 seconds (95 for kas) instead of gracefully terminated.

      Actual results:


      Expected results:


      Additional info:


            agarcial@redhat.com Alberto Garcia Lamela
            joseph-goergen-2 Joseph Goergen
            0 Vote for this issue
            3 Start watching this issue