Open source, API based workflow orchestration tool.
- Homepage: callhouston.io
- Quickstart guide: houston-quickstart-python
- Docs: ./docs
This repo contains the API server, go client, and CLI.
Start a local server with the default config: houston api
Quickly run an end-to-end example workflow: houston demo
Or use the Docker container: docker run -p 8000:8000 datasparq/houston-redis demo
See the quickstart for a guide on how to create microservices and complete Houston missions using them: quickstart
You can download the binary from the releases page.
If you have go installed you can install with:
go install github.com/datasparq-ai/houston@latest
Houston is a simpler, faster, and cheaper alternative to tools like Airflow.
API based orchestration comes with 5 key advantages:
- Code can run on serverless tools: lower cost, less maintenance, infinite scale
- The server isn't under heavy load, so can handle hundreds of concurrent missions
- Pub/Sub message delivery is guaranteed, improving reliability
- Multiple workflows can share the same task runners, aiding collaboration
- Task runners can run anywhere in any language, allowing for rapid development with no vendor lock-in
Please see the contributing guide.
Development of Houston is supported by Datasparq.