Application guide

This section introduces the main differences in running a Ray application on your laptop vs on a Ray Cluster. To get started, check out the job submissions page.

To develop and deploy a Ray Serve application, see the Ray Serve guides. Development Workflow covers the local development loop. Deploy on VM covers VM clusters that the cluster launcher started, and Deploy on Kubernetes covers Kubernetes with KubeRay.

• Ray Jobs Overview
  • Ray Jobs API
  • Running Jobs Interactively
  • Contents
• Programmatic Cluster Scaling
  • ray.autoscaler.sdk.request_resources