Open Source

Building in the Open

Odin

Production-Ready Multi-tenant Deployment Platform

Java Vert.x Kubernetes Helm Multi-tenant

A developer-first deployment platform that enables engineers to define software once and deploy it anywhere, any number of times. Odin provisions isolated ephemeral environments, handles service orchestration via DAG-based execution plans, and guarantees environment parity from development through production.

Battle-tested at Dream11 scale, handling 15M+ concurrent users during IPL season with zero-downtime deployments.

  • Define services as JSON blueprints, deploy with zero config knowledge
  • Ephemeral environments on demand: spin up isolated dev/staging/prod instantly
  • Stateless orchestrator with DAG-based execution and automatic rollback
⭐ 151 🔀 7 forks 🛡️ LGPL-3.0
Repository 📖 Documentation

Odin Component Interface

Plugin Architecture for Cloud-Native Deployments

Groovy Shell HCL Jinja

The contract layer that decouples Odin components from the orchestration engine, enabling anyone to build custom deployment components that integrate seamlessly with the platform. Think of it as Docker's image spec, but for infrastructure deployments.

Defines a complete component lifecycle: preDeploy, deploy, postDeploy, discovery, healthCheck, operate, and undeploy. Components can be written in any language and support multiple cloud provider flavours (AWS, GCP, container) with the same interface.

  • Top contributor: designed and implemented the core DSL specification
  • Multi-flavour support: same component runs on AWS RDS, GCP Cloud SQL, or container
  • Jinja-based config interpolation with state persistence across invocations
⭐ 3 🔀 0 forks 🛡️ LGPL-3.0
Repository

Odin Components

Production-Grade Infrastructure Components Registry

Java Shell Groovy Terraform

The official registry of production-ready components for Odin: MySQL, Kafka, Spark, PostgreSQL, Redis, Cassandra, and more. Each component implements the Odin Component Interface and provides multiple cloud flavours (AWS RDS, GCP Cloud SQL, container) out of the box.

24+ components covering databases, message queues, big data platforms, and caching layers. Each ships with operational actions (scale, backup, add-reader) and health checks.

  • Contributor to component design and the schema guiding principles
  • 24+ production-grade components across databases, messaging, and compute
  • Multi-cloud: same component definition, different cloud provider flavours
⭐ 2 📦 6 releases 🛡️ LGPL-3.0
Repository