https://rancher.com/ logo
#general
Title
# general
h

happy-elephant-46487

01/03/2023, 6:57 AM
@fast-piano-59234 Hello, - I have accidentally deleted the "Kube-Controller-Manager" docker container of my running RKE cluster. - I thought, it will come up automatically, as it might be running at static pod, but it did not started. Could you please let me know, how can I start "kube-controller-manager" docker container of RKE cluster? Cluster details: - 2 Nodes cluster (Rancher Manager cluster/Upstream) - Both nodes have CP/ETCD/Worker roles - RKE version - 1.22.9
f

fast-piano-59234

01/03/2023, 8:27 AM
If this is RKE1, and in Rancher afaics, you have to trigger cluster provisioning to let RKE check cluster state and fix any missing containers. You can easily do this by editing the cluster as YAML and changing
Copy code
addon_job_timeout
as that does not actually change anything in the cluster but does trigger provisioning. If the cluster is broken beyond repair, you can try to restore a backup. As its just a
docker run
of a container, you can also figure it out yourself and run it manually but it needs a bit more context and preparation to be able to do that (what versions, what config etc)
h

happy-elephant-46487

01/03/2023, 8:30 AM
I tried to find out, how these containers started using below command and output is there. Can I use these details to start "kube-controller-manager" container or not? docker ps --no-trunc | grep kube-controller-manager 864328ddb7fc4c6c08baf2c98e8aaad5f68251ce934c345a6c0cac431dbba05a rancher/hyperkube:v1.22.9-rancher1 "/opt/rke-tools/entrypoint.sh kube-controller-manager --service-account-private-key-file=/etc/kubernetes/ssl/kube-service-account-token-key.pem --allocate-node-cidrs=true --leader-elect=true --authorization-kubeconfig=/etc/kubernetes/ssl/kubecfg-kube-controller-manager.yaml --service-cluster-ip-range=10.43.0.0/16 --node-monitor-grace-period=40s --root-ca-file=/etc/kubernetes/ssl/kube-ca.pem --profiling=false --address=0.0.0.0 --allow-untagged-cloud=true --authentication-kubeconfig=/etc/kubernetes/ssl/kubecfg-kube-controller-manager.yaml --configure-cloud-routes=false --v=2 --cloud-provider= --cluster-cidr=10.42.0.0/16 --kubeconfig=/etc/kubernetes/ssl/kubecfg-kube-controller-manager.yaml --enable-hostpath-provisioner=false --pod-eviction-timeout=5m0s --terminated-pod-gc-threshold=1000 --use-service-account-credentials=true"
I dont want to clean up the Node and Join again as of now, need to know other options to start "kube-controller-manager" container .
13 Views