This message was deleted.

It's not an option yet as I understand it. There are some underlying changes from 1.24 to 1.25 (and actually N-0 is 1.26 on EKS.) One change is PSP policies. I can't speak to where engineering is at with this but hope someone will chime in

yes I’m aware of PSP policies changes, I’ve updated all namespace with annotations etc. It’s really strange, they say

2.7.3

is compatible with

1.25

Did you try importing a 1.25 cluster? Just curious

no

EKS 1.25 is not verified in Rancher 2.7.3, therefore it is not in the version list for Rancher-provisioned EKS cluster. I think Rancher 2.7.2 can manage the imported EKS 1.25 cluster with no problem.

sorry was too tired when reading support matrix: AKS / EKS…

Supporting up to 1.24 (N-2), when EKS is on 1.26 presents an interesting customer experience if we want folks managing EKS clusters with Rancher. Their default deployment version is 1.25 and will very shortly be 1.26. Our Marketplace offering deploys onto 1.23. I'm hoping the new MP offering will at least leverage 1.25

afaik, the team is working on adding the support for 1.25 and 1.26 for Rancher-provisioned EKS cluster, which we can expect to be available in the coming Rancher release. 🙂

That's great Jack! Interestingly, when I update a downstream cluster from 1.22 to 1.23 in the EKS console, I get this EBS CSI driver notification. https://docs.aws.amazon.com/eks/latest/userguide/ebs-csi-migration-faq.html Rancher 2.6.10 (Rancher-Setup) appears to still manage it fine as a Rancher deployed Downstream EKS cluster. The only driver I'm using is efs.csi.aws.com After updating the control plane: Notification "New AMI release versions are available for 1 node group." so it does not update the node kubelet and I can do a 'Rolling Update' from AMI v1.22.17-20230411 (kubelet v1.22.17-eks-a59e1f0) to AMI version. 1.23-v20230501 (kubelet v1.23.17-eks-a59e1f0) This added 3 nodes to my 2-node group, and drained pods into the new 3, my node group was set up with 2 nodes +-1. Now I can update to 1.24 and kubelet v1.24.11-eks-a59e1f0. --BROKEN-- In Rancher UI, the cluster is in error state now: "Controller.FailureMessage{ClusterName:"test-5", Message_:"Unsupported Kubernetes minor version update from 1.24 to 1.23", NodegroupName:""}" What I have not done yet is upgrade from 1.23 to 1.24 FROM Rancher UI to see if it can update the cluster and node-groups w/o falling into an error state. I also could have upgraded Rancher beforehand. I don't expect anything higher than 1.24 to work at all.