Developer documentation
Architecture Overview
The following diagram illustrates the complete Helmholtz-KG infrastructure, showing how metadata flows from external sources through the system to end users:
How to Read This Diagram
| Section | Purpose | Access Point |
|---|---|---|
| Users | Who interacts with the KG | Choose based on your needs: UI for browsing, SPARQL for queries |
| Infrastructure | What runs the system | Not directly accessible, but enables weekly automation |
| Layer 1 (Raw) | Original data as harvested | Data Storage API |
| Layer 2 (Graph) | Harmonized RDF data | QLever (fast) or Virtuoso (compliant) SPARQL |
| Layer 3 (Index) | Optimized for search | Web UI |
Deployment Stack:
- Platform: Jülich Supercomputer Centre (JSC) Cloud
- Orchestration: Kubernetes (K8s)
- CI/CD: GitLab CI/CD + FluxCD (GitOps)
- Pipeline: Apache Airflow (weekly execution)
