Suspicious messages in eks-operator logs when importing EKS cluster

[thehejik]

On attempt to import an eks cluster into rancher:v2.8-head I see following error messages in eks-operator pod logs.

Versions:
rancher:v2.8-74d6c7a4c1f564eb2389508e48466590f12fab5a-head
rancher-eks-operator:103.6.0+up1.3.6-rc.1

Command used for provisioning EKS cluster (using latest eksctl 0.198.0):

eksctl create cluster --name=th-imported --region=eu-west-2 --nodes=1 --node-type=t3.xlarge --node-volume-size=20 --managed --kubeconfig=kubeconfig-th-eks-imported --version 1.28

It is probably harmless and the cluster has been imported successfully, I could scale and add nodepools but the errors seems to be a bit suspicious.

time="2024-12-11T13:36:03Z" level=info msg="Starting eks.cattle.io/v1, Kind=EKSClusterConfig controller"
time="2024-12-11T13:36:03Z" level=info msg="Starting /v1, Kind=Secret controller"
time="2024-12-11T13:36:03Z" level=error msg="Error removing metadata from failure message: message body not formatted as expected"
2024-12-11T13:36:03.955416239Z time="2024-12-11T13:36:03Z" level=error msg="Error recording ekscc [] failure message: resource name may not be empty"
2024-12-11T13:36:04.768479213Z time="2024-12-11T13:36:04Z" level=info msg="cluster [c-5kzvt] finished updating"
2024-12-11T13:36:04.771995869Z time="2024-12-11T13:36:04Z" level=error msg="Error removing metadata from failure message: message body not formatted as expected"
2024-12-11T13:36:04.772034729Z time="2024-12-11T13:36:04Z" level=error msg="Error recording ekscc [] failure message: resource name may not be empty"
2024-12-11T13:36:05.227893452Z time="2024-12-11T13:36:05Z" level=info msg="cluster [c-5kzvt] finished updating"
2024-12-11T13:36:05.755315041Z time="2024-12-11T13:36:05Z" level=info msg="cluster [c-5kzvt] finished updating"
time="2024-12-11T13:38:36Z" level=info msg="waiting for cluster [c-5kzvt] to update nodegroups [th-my-group]"
time="2024-12-11T13:38:37Z" level=info msg="waiting for cluster [c-5kzvt] to update nodegroups [th-my-group]"
time="2024-12-11T13:39:07Z" level=info msg="waiting for cluster [c-5kzvt] to update nodegroups [th-my-group]"
time="2024-12-11T13:39:38Z" level=info msg="waiting for cluster [c-5kzvt] to update nodegroups [th-my-group]"
time="2024-12-11T13:40:08Z" level=info msg="waiting for cluster [c-5kzvt] to update nodegroups [th-my-group]"
time="2024-12-11T13:40:39Z" level=info msg="cluster [c-5kzvt] finished updating"

[kkaempf]

@thehejik - 2.8 is EOM. If the import is successful, I'd vote to close this issue as wontfix (except @yiannistri thinks otherwise 😉)

[thehejik]

Also spotted on AKS rancher/aks-operator#665

[rishabhmsra]

Validated on rancher v2.11-1d345932e5f78b3da07ebe13691c8224c62b6240-head

• Created EKS v1.30.9-eks-8cce635 cluster.
• Imported it in rancher
• Verified that no errors are present in eks-operator logs, hence closing this issue.