This message was deleted.

A little yes and no. If there's differences in the CPU VMs may fail to migrate. It's a limitation of the underlying virtualization mechanisms. But if you're nodes are are the same type/generation, then yes.

The scheduler will help distribute the workload depending on the available resources on the nodes.

What is the limit of worker nodes that I can join to a cluster?

It depends on the specs of your machines. According to Kubernetes documentation, it should be no more than 5000 nodes [ref](https://kubernetes.io/docs/setup/best-practices/cluster-large/). We also have an embedded Rancher running on the cluster, so [this](https://ranchermanager.docs.rancher.com/getting-started/installation-and-upgrade/installation-requirements#rke2-kubernetes) might also be worth taking a look.