Rundown-Workers is a lightweight, language-agnostic background task and workflow executor. It manages queue states, transaction timeouts, and retries, while allowing your code (workers) to execute anywhere in any language over simple HTTP JSON APIs.
- Installation Guide — Docker Compose and host machine installation.
- Configuration & Usage — HTTP API reference, timeout settings, and lifecycle rules.
- SDKs & Custom Workers — Official SDKs and implementing custom workers.
Reference the official Docker Hub image itsernest/rundown-workers:latest directly in your compose.yml to provision background job scheduling:
services:
# RUNDOWN WORKERS BACKGROUND ENGINE
rundown:
image: itsernest/rundown-workers:latest
container_name: rundown-workers-engine
ports:
- "8181:8181"
environment:
- RUNDOWN_STORE=postgres
- DATABASE_URL=postgres://flare_admin:super_secure_db_password_123@flare-postgres-db:5432/flare_boost_db?sslmode=disable
restart: unless-stoppedAny container in the same service network can poll or dispatch jobs from the engine at http://rundown:8181.
- Language Agnostic: Workers can poll and process tasks using simple JSON HTTP requests.
- Dual Engine Backends: Choose between embedded SQLite for zero-config setups or PostgreSQL for production-ready, scaled compose architectures.
- Concurrency Safe: Uses transactional database locking to ensure jobs are claimed exactly once, even under heavy parallel polling.
- Failures & Automatic Backoff: Scheduled retries calculate automatic exponential backoffs.
- Staleness Loop: Auto-recovers jobs whose workers crashed or hung.
Distributed under the MIT License.