-
Hi juano1985,
Great question! The key distinction is what each scaler actually does:
• HPA scales pods within existing nodes — it cannot add new nodes when capacity runs out.
• Cluster Autoscaler provisions new nodes when pods can’t be scheduled due to insufficient compute resources.
The question specifically asks about “provisioning new compute resources,” which is node-level scaling; only the Cluster Autoscaler does this. HPA alone would leave new pods stuck in a Pending state once existing nodes are full.
The Cluster Autoscaler is also the correct fit for Windows Server 2022 node pools in AKS.
If you have other questions, feel free to reach out!
Best regards,
Irene @ Tutorials Dojo
