mitzen

 After de-registering the server, we still have some namespaces, rolebindings, role and pods in the cluster that needs to be clean up. To do that without causing your resource wait indefinite due to finalizer, do the following  curl -O https://raw.githubusercontent.com/rancher/rancher/master/cleanup/user-cluster.sh chmod +x user-cluster.sh   ./user-cluster.sh rancher/rancher-agent:v2.13-head                         If everything runs well, you get this 

I guess this can happen to any version of keda metric api metric server getting the error after upgrade/deployment.  grpc: addrConn.createTransport failed to connect to {Addr: "10.0.224.114:9666", ServerName: "keda-operator.keda.svc.cluster.local:9666", }. Err: connection error: desc = "transport: authentication handshake failed: read tcp 10.244.43.200:60114->10.0.224.114:9666: read: connection reset by peer" The solution, delete a secret call kedaorg-cert in the keda namespace and let keda re-generate by doing a deployment rollout for all keda pods.  Check if the certificate exist  kubectl get secrets -n keda Proceed to delete the certificate  kubectl delete secret kedaorg-certs -n keda Redeploy the instances kubectl rollout restart deployment keda-operator -n keda kubectl rollout restart deployment keda-operator-metrics-apiserver -n keda Right at the end you will see connection is established.