ARCHITECTURE

One spec. Three runtimes. No rewrites.

Aether's core philosophy is write once, deploy anywhere. A single workload YAML specification describes what to run, and Aether's decision engine automatically selects the optimal runtime.

• 🏛️

  Overview

  Universal runtime portability — deploy once, move workloads across Podman, Kubernetes, and KubeVirt.

  →
• ▣

  Deployment topologies

  How to run Aether in dev, production HA, hybrid cloud, and edge scenarios.

  →
• ⛓

  Fleet architecture

  Multi-cluster inventory today and the federation roadmap.

  →
• ◈

  Networking

  How network settings translate across runtimes and what breaks on migrate.

  →