After upgrading K8s version on EKS to ‘v1.27.12-eks-adc7111’, the cluster agent fails to enter a ready state on the local Rancher cluster.
Checking the Rancher server logs, I see these errors:
2024/05/08 08:23:26 [ERROR] Failed to connect to peer wss://10.42.0.52/v3/connect [local ID=10.42.0.51]: websocket: bad handshake
2024/05/08 08:23:27 [ERROR] Failed to handle tunnel request from remote address 10.42.0.52:47634: response 400: cluster not found
On the EKS cluster, there are no obvious errors on the cattle-cluster-agent, nor the rancher-webhook
Any ideas what might be causing the 400 errors + bad handshake errors? Or advice on where to look? I’ve tried de-registering, removing the rancher-related namespaces and re-registering the EKS cluster, as this has helped in the past, but no luck there.
Rancher version: v2.8.3
Running on k3s version: v1.28.8+k3s1
EKS version: v1.27.12-eks-adc7111