Self-Hosting the Hatchet Control Plane

Self-hosting Hatchet means running your own instance of the Hatchet Control Plane - the central orchestration system that manages workflows, schedules tasks, and coordinates worker execution. This is different from running workers, which can connect to any Hatchet instance (self-hosted or Hatchet Cloud).

What You're Self-Hosting

When you self-host Hatchet, you're deploying:

• API Server - REST APIs for workflow management
• Engine - gRPC API for core workflow orchestration and task scheduling
• Database - PostgreSQL for storing workflow state and metadata
• Message Queue (optional) - RabbitMQ for inter-service communication and high-throughput real-time updates
• Dashboard - Web UI for monitoring workflows and debugging

Your workers (the processes that execute your workflow steps) will connect to your self-hosted control plane and execute tasks.

Deployment Options

The fastest way to get a Hatchet instance running locally is with the Hatchet CLI (which wraps Hatchet Lite):

hatchet server start

There are currently three supported ways to self-host the Hatchet Control Plane:

Docker:

1. Hatchet Lite - Single docker image with bundled engine and API (development, testing, or low-throughput production)
2. Docker Compose - Multi-container setup with PostgreSQL and RabbitMQ (production)

Kubernetes:

1. Quickstart with Helm - Production-ready Helm charts (production)