In this partners MicroShift deployment they have noticed that after the node reboots, two postgres pods fail to start back up properly and are hitting:

Oct 20 09:25:58 spgttmicro-os.schuler.de microshift[3534]: kubelet E1020 09:25:58.438702    3534 log.go:32] "StopPodSandbox from runtime service failed" err=<
Oct 20 09:25:58 spgttmicro-os.schuler.de microshift[3534]:         rpc error: code = Unknown desc = failed to destroy network for pod sandbox k8s_dps-infra-postgres-repo-host-0_dps_8b4cd4e8-32db-48a3-891b-26ae090370f0_0(d21853b5e0c87d279d08f2484bd5d7b8c1a13788d625cc2f2301fcc83532c2e3): error removing pod dps-infra-postgres-repo-host-0 from CNI network "ovn-kubernetes": plugin type="ovn-k8s-cni-overlay" name="ovn-kubernetes" failed (delete): CNI request failed with status 400: '[<namespace>/dps-infra-postgres-repo-host-0 d21853b5e0c87d279d08f2484bd5d7b8c1a13788d625cc2f2301fcc83532c2e3 network default NAD default] [<namespace>/dps-infra-postgres-repo-host-0 d21853b5e0c87d279d08f2484bd5d7b8c1a13788d625cc2f2301fcc83532c2e3 network default NAD default] failed to get container namespace for pod <namespace>/dps-infra-postgres-repo-host-0 NAD default: failed to Statfs "": no such file or directory
Oct 20 09:25:58 spgttmicro-os.schuler.de microshift[3534]:         ': stat netns path "": stat : no such file or directory
Oct 20 09:25:58 spgttmicro-os.schuler.de microshift[3534]:  > podSandboxID="d21853b5e0c87d279d08f2484bd5d7b8c1a13788d625cc2f2301fcc83532c2e3"
Oct 20 09:25:58 spgttmicro-os.schuler.de microshift[3534]: kubelet E1020 09:25:58.438773    3534 kuberuntime_manager.go:1479] "Failed to stop sandbox" podSandboxID={"Type":"cri-o","ID":"d21853b5e0c87d279d08f2484bd5d7b8c1a13788d625cc2f2301fcc83532c2e3"}
Oct 20 09:25:58 spgttmicro-os.schuler.de microshift[3534]: kubelet E1020 09:25:58.438817    3534 kuberuntime_manager.go:1079] "killPodWithSyncResult failed" err="failed to \"KillPodSandbox\" for \"8b4cd4e8-32db-48a3-891b-26ae090370f0\" with KillPodSandboxError: \"rpc error: code = Unknown desc = failed to destroy network for pod sandbox k8s_dps-infra-postgres-repo-host-0-dps_8b4cd4e8-32db-48a3-891b-26ae090370f0_0(d21853b5e0c87d279d08f2484bd5d7b8c1a13788d625cc2f2301fcc83532c2e3): error removing pod dps-infra-postgres-repo-host-0 from CNI network \\\"ovn-kubernetes\\\": plugin type=\\\"ovn-k8s-cni-overlay\\\" name=\\\"ovn-kubernetes\\\" failed (delete): CNI request failed with status 400: '[<namespace>/dps-infra-postgres-repo-host-0 d21853b5e0c87d279d08f2484bd5d7b8c1a13788d625cc2f2301fcc83532c2e3 network default NAD default] [<namespace>/dps-infra-postgres-repo-host-0 d21853b5e0c87d279d08f2484bd5d7b8c1a13788d625cc2f2301fcc83532c2e3 network default NAD default] failed to get container namespace for pod <namespace>/dps-infra-postgres-repo-host-0 NAD default: failed to Statfs \\\"\\\": no such file or directory\\n': stat netns path \\\"\\\": stat : no such file or directory\""
Oct 20 09:25:58 spgttmicro-os.schuler.de microshift[3534]: kubelet E1020 09:25:58.438855    3534 pod_workers.go:1301] "Error syncing pod, skipping" err="failed to \"KillPodSandbox\" for \"8b4cd4e8-32db-48a3-891b-26ae090370f0\" with KillPodSandboxError: \"rpc error: code = Unknown desc = failed to destroy network for pod sandbox k8s_dps-infra-postgres-repo-host-0_8b4cd4e8-32db-48a3-891b-26ae090370f0_0(d21853b5e0c87d279d08f2484bd5d7b8c1a13788d625cc2f2301fcc83532c2e3): error removing pod dps-infra-postgres-repo-host-0 from CNI network \\\"ovn-kubernetes\\\": plugin type=\\\"ovn-k8s-cni-overlay\\\" name=\\\"ovn-kubernetes\\\" failed (delete): CNI request failed with status 400: '[<namespace>/dps-infra-postgres-repo-host-0 d21853b5e0c87d279d08f2484bd5d7b8c1a13788d625cc2f2301fcc83532c2e3 network default NAD default] [<namespace>/dps-infra-postgres-repo-host-0 d21853b5e0c87d279d08f2484bd5d7b8c1a13788d625cc2f2301fcc83532c2e3 network default NAD default] failed to get container namespace for pod <namespace>/dps-infra-postgres-repo-host-0 NAD default: failed to Statfs \\\"\\\": no such file or directory\\n': stat netns path \\\"\\\": stat : no such file or directory\"" pod="<namespace>/dps-infra-postgres-repo-host-0" podUID="8b4cd4e8-32db-48a3-891b-26ae090370f0"

    cri-o: 1.31.11-2.rhaos4.18.git65ec77a.el9
    microshift: 4.18

    Always

    1. With all the pods up & running reboot the node
    2. Wait for the node to come back up and for the pods to be recreated, impacted pods never come back up
    
    

    Impacted pods never manage to come back up on their own and need to be forcefully deleted so a new one can be created

    Pods should be able to come back up without intervention

    Looks to be similar to OCPBUGS-58229 however in that bug the cri-o version was 1.33, in our case is 1.31.11