You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
What happened: I raised up a new cluster with Terraform, the workers don't get registered into the cluster, when I checked out into logs details, the kubelet was not working, the kubeconfig was misconfigured, the ca /etc/kubernetes/pki/ca.crt was wrong as well.
The file /tmp/describe_cluster_result.txt created by bootstrap.sh was filled with the content below:
None None
What you expected to happen: The bootstrap got the API Endpoint and CA correctly from AWS API.
How to reproduce it (as minimally and precisely as possible):
It's not something easy to reproduce because it depends on the time and on AWS API that has eventual consistency.
@jaksonwkr did this Terraform build cluster provisioning used to work? If so, do you use custom domain names for nodes? If so, this recent change/fix for certificates might be relevant.
If you are using a custom non-EC2 assigned hostname, you'll need to configure the kubelet using the --hostname-override flag to an EC2-identifiable hostname (Ex: the instance's private DNS name) so that our certificate signer can verify that the kubelet certificate is for a legitimate target.
@whereisaaron I didn't build the AMI with Terraform, I used the official AMI and I'm not using a custom domain.
I think if the command aws eks describe-cluster used to get that information returns "None" it should retry instead pass to the next step and could resolve this problem.
What happened: I raised up a new cluster with Terraform, the workers don't get registered into the cluster, when I checked out into logs details, the kubelet was not working, the kubeconfig was misconfigured, the ca
/etc/kubernetes/pki/ca.crt
was wrong as well.The file
/tmp/describe_cluster_result.txt
created by bootstrap.sh was filled with the content below:What you expected to happen: The bootstrap got the API Endpoint and CA correctly from AWS API.
How to reproduce it (as minimally and precisely as possible):
It's not something easy to reproduce because it depends on the time and on AWS API that has eventual consistency.
Anything else we need to know?:
Environment:
aws eks describe-cluster --name <name> --query cluster.platformVersion
): eks.2aws eks describe-cluster --name <name> --query cluster.version
): 1.12uname -a
): 4.14.106-97.85.amzn2.x86_64cat /etc/eks/release
on a node):The text was updated successfully, but these errors were encountered: