Hi HN! I'm the author of Dagu, a workflow engine that's been in development for a few years. Just released v1.17.0-beta.1 with some significant improvements.
The core idea: define workflows in YAML, run them as scheduled jobs or on-demand. Single Go binary, no external dependencies, stores everything locally.
New in this release:
- Nested DAGs - workflows can call other workflows
- Refactored execution history storage (10x faster for large histories)
- Better debugging UI showing precondition results and variable outputs
- Proper job queue management
Example workflow:
```yaml
name: data-pipeline
schedule: "0 10 * * *"
steps:
- name: fetch-data
command: curl https://api.example.com/data > raw.json
- name: process
command: jq '.items[] | select(.active)' raw.json > processed.json
depends: fetch-data
- name: load
command: python load_to_db.py processed.json
depends: process
```
Try it:
```
docker run --rm -p 8080:8080 ghcr.io/dagu-org/dagu:1.17.0-beta.1 dagu start-all
```
Then visit http://localhost:8080
We use it in production for ETL pipelines, report generation, and system maintenance tasks. It's not trying to be Airflow - no distributed execution, no Python dependency hell, just a reliable way to run workflows.
GitHub: https://github.com/dagu-org/dagu
Would love feedback on the beta, especially around the nested DAG implementation and any performance issues you encounter.
