Skip to main content
Version: v1.0

FAQ

yurtadm

1. yurtadm join error:crictl not found in system path

The node does not have docker installed, and installing docker can solve this problem.

2. yurtadm join error:[ERROR FileExisting-conntrack]: conntrack not found in system path

Execute yum install -y conntrack and then retry the yurtadm join command again.

3. helm installation error during yurtadm init: Release "yurt-app-manager" does not exist. Installing it now. Error: failed to download "openyurt/yurt-app-manager" (hint: running helm repo update may help)

Perform the helm installation process manually:

cd /var/lib/sealer/data/my-cluster/rootfs
helm upgrade --install yurt-app-manager openyurt/yurt-app-manager -n kube-system -f manifests/yurt-app-manager-values.yaml

4. yurtadm init failed, how to clean up the environment?

Since yurtadm init uses sealer to create an OpenYurt cluster from scratch, you can refer to this to clean up the environment.

yurt-tunnel

1. kubectl logs edge node error:error: Error from server (ServiceUnavailable): the server is currently unable to handle the request ( pods/log xxx)

For more details about this problem, you can refer to issue 984.

Please make sure that you are using the latest yurt-tunnel-server/agent image (the latest tag of image is confusing, please check the created time).

2. kubectl logs edge node error:error: You must be logged in to the server (the server has asked for the client to provide credentials ( pods/log xxx))

For more details about this problem, you can refer to issue 984.

This is most likely to be an issue caused by outdated tunnel server certificate. You can clean up the tunnel server certificate and redeploy yurt-tunnel-server.

rm -rf /var/lib/yurttunnel-server/pki
kubectl delete pod yurt-tunnel-server-xxxxxx -n kube-system