Harvester Bare Metal Mass Deployment & Management Considerations

SUSE Virtualization
1

Reference: PXE Boot Install Options

I’m an old SLE admin here, I know exactly how I would accomplish this “the old way” – so bear with me as I try to clarify the capabilities of Harvester and how we should deploy “the new way”.

Is anyone deploying dozens of Harvester Clusters and nodes on their own hardware?

Our clusters will be managing (running) both VM’s and Containers across over 100 nodes; has anyone automated their deployments and had success? What is the best practice for this?

Also;

  1. I see that SUMA (Uyuni) can now deploy and manage SLE Micro 5.5+. This is what is underneath Harvester by default; can we deploy/Manage Harvester with Uyuni/SUMA yet?

  2. Can Rancher Prime manage/deploy a Harvester Cluster? Is this a better practice than #1 above or can they work together?

2

Yes, several users are deploying large-scale Harvester clusters across many nodes. The recommended best practice is to automate deployments using PXE boot or Terraform with Harvester’s cloud-init configurations. Rancher Prime can fully manage and deploy Harvester clusters, offering centralized control and integration with Kubernetes workloads. Uyuni/SUMA can handle SLE Micro management but not full Harvester lifecycle yet. Using Rancher Prime for orchestration while Uyuni manages the underlying OS is generally the most efficient setup — a powerful combination like a luxury limousine built for performance and control.