New autoscaler nodes fail to register within 15min - How to debug?

[schlichtanders]

Hi there,
last week my autoscaling group just does not autosale any longer. I inspected it and found that it is because new nodes are not registering. Somehow they run int longUnregistered state and are killed again. I can see the nodes on hetzner and can ssh into them for the 15 minutes they last.

Has someone a hint how I can debug why this node is not joining the cluster successfully?

• Which files to look at or re-execute?
• how to get which logs?

Thank you for your help

EDIT: using journalctl I was able to find the error (which turned out to be solved in most recent kube-hetzner) and reported it as an issue #1231