Kubernetes upgrade

You can follow the blue-green deployment strategy for your EKS Cluster Upgrade.

You can refer to this guide: https://engineering.chingari.io/upgrading-amazon-eks-from-version-1-21-to-1-22-a-step-by-step-guide-69172993228f

I would suggest updating minimum to 1.24 as 1.23 will be supported by AWS only until October, 2023.

Any idea if the worker nodes currently running on 1.19 will remain running after the management nodes are upgraded to 1.22?

A rolling, zero-downtime, cluster upgrade is feasible but a new Node Group should be created for smoothest transition.

If you say that it's feasible, the answer to a 1.19 worker, will continue to run on a 1.22 cluster would be yes right? Under normal circumstances the worker nodes would need to be upgrade to at least 1.20 to be able to run on 1.22, but in this case I'm not allowed to do this anymore...

What about your resources running in the cluster ? there gonna be multiple API deprecations. If you face any of it on your helm releases try this tool :) https://github.com/helm/helm-mapkubeapis

Ok, well it seems that all the suggestion work fine when you're not this far behind. I'm familiar wit blue green upgrades, but not when you're 3 versions apart And I'm not hearing a definitive answer about wether 1.19 workers can work with 1.22 master nodes, so I will rebuild the cluster and migrate the parts over. The breaking changes from 21 to 22 are also not fun to deal with. No biggy, we don't use much of the k8s ecosystem, so we can easily build from scratch in a couple days :)