Best practices for scaling AKS node pools?

Posted by jdoe • Sep 9, 2025 • 34 replies

I'm planning to use AKS for a high‑traffic web application. What are the recommended strategies for scaling node pools dynamically based on CPU and memory metrics? Should I use cluster autoscaler, virtual node, or a combination of both?

Also, any tips on configuring pod disruption budgets for zero‑downtime deployments would be appreciated.

alice

Sep 9, 2025 • 2 likes

You can enable the AKS cluster autoscaler and set min/max node counts per node pool. Combine it with the Virtual Kubelet for burst workloads.

bob

Sep 10, 2025 • 1 like

Don’t forget to set up a pod disruption budget (PDB) for your critical services. A PDB of 1 out of 3 replicas works well for most cases.

Best practices for scaling AKS node pools?

Leave a comment