dragonflyoss/helm-charts

Dragonfly Helm Charts is a collection of Helm charts for deploying Dragonfly, a cloud-native P2P-based data distribution and acceleration system on Kubernetes.

Dragonfly addresses the challenge of efficiently distributing large files, container images, OCI artifacts, and machine learning models across cloud-native environments. It uses peer-to-peer technology to accelerate data delivery while maintaining security and stability. The project includes optional integration with Nydus, a content-addressable filesystem that further accelerates OCI container image launches. The Helm charts provide a standards-based, best-practice approach to deploying these components on Kubernetes clusters.

Teams should adopt these charts if they operate large-scale Kubernetes deployments where efficient distribution of container images, artifacts, or large files is a bottleneck. The charts are suitable for organizations managing multiple clusters or high-volume container image pulls, as well as those working with AI/ML models or other large binary artifacts. The repository provides three main chart options: a complete Dragonfly stack, Dragonfly alone, and Nydus snapshotter integration, allowing teams to choose the deployment model that fits their infrastructure needs.

The project maintains active community engagement through multiple channels including a dedicated Slack community, GitHub discussions, and developer mailing lists. Development activity shows consistent attention to documentation, with separate installation guides provided for each major component and integration scenario.