GKE Cluster Autoscaler

This repository contains a version of the Kubernetes Cluster Autoscaler used in Google Kubernetes Engine (GKE).

Overview

This project contains the GKE Cluster Autoscaler, a core component of Google Kubernetes Engine (GKE). Its primary function is to automatically resize the number of nodes in a given node pool based on the demands of the cluster workloads.

It dynamically adjusts the size of a cluster in two ways:

• Scaling up: When there are Pods that fail to schedule onto current nodes due to insufficient resources, the autoscaler adds new nodes to the node pool.
• Scaling down: When nodes are underutilized and all of their Pods can be safely scheduled onto other nodes, the autoscaler removes those excess nodes to optimize resource usage and costs.

Documentation

• More details about cluster autoscaler

Requirements

• Go: Version 1.25.0 or higher (as defined in go.mod).

Contributing

See CONTRIBUTING.md for details on how to contribute to this repository.

License

Apache 2.0; see LICENSE for details.