This message was deleted.

The easy answer is to apply the yaml from the release page : https://github.com/harvester/harvester/releases/tag/v1.7.0 aka

kubectl apply -f <https://releases.rancher.com/harvester/v1.7.0/version.yaml>

Thanks, How should one monitor and complete the upgrade procedure once it has already been initiated?

applying that yaml will get you the upgrade button. Then you would watch it in the upgrade dropdown as normal.

The upgrade is stuck after "Images preloaded" . Upgrading System Service @ 50% , "system services" is showing pending.

wait. it can take a while.

OK, is there any pod that shows current stage it's waiting on?

there is.. I think it is a job.. I forge the name. I am waiting on some CPUs before I upgrade today.

I don't see upgrade plan , is this expected ?

k get <http://plans.upgrade.cattle.io|plans.upgrade.cattle.io> -A
NAMESPACE       NAME                            IMAGE                                 CHANNEL   VERSION              COMPLETE   MESSAGE
cattle-system   hvst-upgrade-4jdvn-prepare      rancher/harvester-upgrade:v1.6.1                hvst-upgrade-4jdvn   True
cattle-system   rke2-master-plan                rancher/rke2-upgrade                            v1.33.3+rke2r1       True
cattle-system   rke2-worker-plan                rancher/rke2-upgrade                            v1.33.3+rke2r1       True
cattle-system   sync-additional-ca              <http://registry.suse.com/bci/bci-base:15.6|registry.suse.com/bci/bci-base:15.6>             v1.1.0               True
cattle-system   system-agent-upgrader           rancher/system-agent                            v0.3.14-suc          True
cattle-system   system-agent-upgrader-windows   rancher/wins                                    v0.5.3               True

check for jobs

check apply manifest pod in harvester system. It is probably stuck waiting some chart to complete (cattle fleet most probably)

might be this issue? https://github.com/harvester/harvester/issues/9738

I ended up stopping the first upgrade attempt and starting fresh. I hit a snag though where management IP didn't fail over like it should have when the second node rebooted. I had to restart that node again just to get it to let go of the IP. The setup is 10-host cluster.

did you use dhcp or static? Sorry, did you mean the VIP did not move when the node rebooted?

Yes the VIP didn't move . Is there a way to statically configure vip IP ?