Uploaded image for project: 'OpenShift Bugs'
  1. OpenShift Bugs
  2. OCPBUGS-21925

ovnkube-node crashes when nonsensical clusterip-service with itp is in cluster

XMLWordPrintable

    • Quality / Stability / Reliability
    • False
    • Hide

      None

      Show
      None
    • None
    • None
    • No
    • None
    • None
    • Rejected
    • None
    • None
    • None
    • None
    • None
    • None
    • None
    • None

      Description of problem:

      
      

      Version-Release number of selected component (if applicable): 4.13.0-0.okd-2023-05-22-052007

      
      

      How reproducible: always

      
      

      Steps to Reproduce:

      1. have a service of type `ClusterIP` with `internalTrafficPolicy: Local`
      2. update to `okd-4.13`
      3. observe `ovnkube-node`'s crash with
      ```
          State:       Waiting
            Reason:    CrashLoopBackOff
          Last State:  Terminated
            Reason:    Error
            Message:   sbin/iptables -t mangle -C OVN-KUBE-ITP -p TCP -d <nil> --dport 110 -j MARK --set-xmark 0x1745ec --wait]: e
      xit status 2: iptables v1.8.7 (nf_tables): host/network `<nil>' not found
      ```
      
      This does make sense in a way, because there is no external IP that this could apply to.  It arguably should not crash though.
      
      I don't know whether it's possible to create these services with 4.13, but they could be created with 4.12.
      
      

      Additional info:

       Some more info in the downstream bug.
       https://github.com/okd-project/okd/issues/1760
      
      

              bpickard@redhat.com Ben Pickard
              redhat@ibotty.net Tobias Florek (Inactive)
              None
              None
              None
              None
              Votes:
              0 Vote for this issue
              Watchers:
              2 Start watching this issue

                Created:
                Updated:
                Resolved: