This message was deleted.
# rke2a
adamant-kite-43734
11/08/2024, 2:04 AMThis message was deleted.
g
glamorous-umbrella-16701
11/08/2024, 2:20 AMLooks like the v1.20 reference may be related to this - https://github.com/helm/helm/issues/13248
Given the Rancher console is performing the upgrade, there's no way to declare the
--kube-versionc
creamy-pencil-82913
11/08/2024, 3:12 AMThis is a Rancher issue, not RKE2. You might search for issues in the rancher/rancher repo on GH, and open an issue if you can’t find one. I’m sure I don’t need to tell you that everything here is all many many months end of life.
g
glamorous-umbrella-16701
11/08/2024, 4:05 AMYep, hence why I'm going through the motions to get everything up to date. Googling seems to have gotten me nowhere in particular, but I'll look to dig deeper and get an issue raised if I can't resolve... Any other possible hints would be appreciated.
c
creamy-pencil-82913
11/08/2024, 6:29 AMYou might see if there are any updates to fleet available? If you're airgapped you might be missing post-release updates to Rancher components that would normally be available to a non-airgapped cluster via driver metadata updates.
👍 1
m
miniature-notebook-6405
11/09/2024, 4:17 PMWhat's the version of the local cluster running Rancher?
g
glamorous-umbrella-16701
11/10/2024, 7:17 PMThe local Rancher cluster is running v1.24.11+rke2r1 - I was wondering if it's worth upgrading this first, but wanted to test the process on a downstream dev environment first.
glamorous-umbrella-16701
11/12/2024, 1:04 AMNegative, upgrading the local cluster to v1.26.11 didn't do the trick, stuck with the same error when trying to upgrade a downstream cluster:
Copy code
level=info msg="While calculating status.ResourceKey, error running helm template for bundle mcc-monitoring-managed-system-upgrade-controller with target options from : chart requires kubeVersion: >= 1.23.0-0 which is incompatible with Kubernetes v1.20.0"glamorous-umbrella-16701
11/25/2024, 10:44 PMFor anyone else stumbling on this, the issue (at least for our clusters) was that the fleet agent couldn't register with the controller. We were using the hosts file on the cluster node to define the IP of the rancher server, which was working for most containers as we set the dnsPolicy to ClusterFirstWithHostNet for cattle cluster agent deployment. Looks like the fleet pod wasn't using that config as it couldn't reach the rancher server. After creating a CNAME to the rancher server & forcing the cluster update via Fleet-Clusters, updating the rke2 version from the console worked.
glamorous-umbrella-16701
11/27/2024, 1:32 AMExactly
m
miniature-notebook-6405
11/30/2024, 6:31 PMYeah you can't alias things, rancher agent and everything related does DNS lookups.
54 Views