Defining the requirements
Before continuing though, letβs run through the specification of the cluster to be created, both in terms of the hardware and what it should support:
- Servers should be the smallest ARM64 instances available (CAX11)
- The region must be one that offers ARM64 instances (Falkenstein)
- There should only be one control plane instance
- Autoscaling should be enabled
- The worker pool should contain a minimum of one instance
- The worker pool should scale up to a maximum of two instances
- Unattended upgrades should be enabled
Once we create the configuration for this, we should expect to see two servers in the Hetzner Cloud console: one will be for the control plane, and another will be a worker node.
graph LR
subgraph workers[Worker Node Pool]
B[Worker 1]
C[Worker 2]
D[Worker <i>...n</i>]
end
A[Control Plane] --> workers