Global: Calico enabled GKE clusters’ pods may get stuck terminating after upgrading to 1.22+

Summary: Global: Calico enabled GKE clusters’ pods may get stuck terminating after upgrading to 1.22+ Description: GKE clusters running versions 1.22 or later and that use Calico Network Policy might experience issues with terminating Pods under some conditions. Our engineering team continues to investigate the issue and are qualifying a potential mitigation for release to the Rapid channel 1.24. After all the qualifications are done, we will expedite the backport of the fix to 1.22 as soon as possible. We will provide an update by Friday, 2022-09-16 15:00 US/Pacific with current details. We apologize to all who are affected by the disruption. Diagnosis: The Calico CNI plugin will show the following error terminating Pods: “Warning FailedKillPod 36m (x389 over 121m) kubelet error killing pod: failed to "KillPodSandbox" for "af9ab8f9-d6d6-4828-9b8c-a58441dd1f86" with KillPodSandboxError: "rpc error: code = Unknown desc = networkPlugin cni failed to teardown pod "myclient-pod-6474c76996" network: error getting ClusterInformation: connection is unauthorized: Unauthorized" Workaround: Affected customers may try the following: 1. Restart the kubelet and calico-node can help getting the pods unstuck. 2. Disable the Calico network policy. (workaround #1 is recommended, as this workaround is only viable if the customer does not have a strong need for Calico).