What are the limits of Kubernetes Horizontal Pod Autoscaler?

What are the limits of Kubernetes Horizontal Pod Autoscaler?

The Kubernetes Horizontal Pod Autoscaler (HPA) can scale the number of pods based on observed CPU utilization or other select metrics. However, it has some limits, such as the minimum and maximum number of replicas you can set, which are defined in the HPA configuration.

What is the default behavior if no metrics are available?

If no metrics are available, the HPA will not scale the pods. It relies on metrics to make scaling decisions, so without them, it maintains the current number of replicas.

Can I set custom metrics for scaling?

Yes, you can set custom metrics for scaling by using the Kubernetes Metrics API or external metrics providers. This allows you to scale based on application-specific metrics.

What happens if I exceed the maximum replicas limit?

If you exceed the maximum replicas limit set in the HPA configuration, the HPA will not scale beyond that limit, even if the metrics indicate that more replicas are needed.