Lineage Viewer¶
The Lineage Viewer shows the relationships between files and the jobs they are attached to, visualised as a Mermaid graph.
Navigate to Metadata & Files → Lineage Viewer.
Lineage Graph¶
The graph renders automatically from file_object, job_file_link, and all
job tables. Each file is a rectangular node; jobs are shown as parallelogram
nodes labelled by type.
Edges run from job → file, indicating that the job produced or uses the file.
Graph Limits¶
To keep the graph renderable, it is capped at 50 nodes. If your registry is larger, a warning banner appears. Use the filter/search on the job pages to narrow down to specific jobs and their files.
File Registry Table¶
Below the graph is a paginated table of all ingested files (50 per page):
| Column | Description |
|---|---|
| File Name | Original filename at time of ingestion |
| Use | File use classification (ORIGINAL, DERIVATIVE, etc.) |
| Format | Format name (BRF, PDF, STL …) |
| Size | Human-readable file size |
| SHA-256 | First 12 characters of the checksum (hover for full hash) |
| Ingested | Date of ingestion |
Finding a File by Checksum¶
Copy the full SHA-256 hex string and paste it into the Search page.
An exact match is performed against file_object.checksum_sha256.
Provenance Events¶
The lineage viewer also loads provenance events from the in-memory
ProvenanceRegistry. In development mode (ACCESSMAN_DEV=1), three seed
events are added automatically so the registry is not empty on first load.
Provenance events are separate from the per-job metadata_event table.
They are registered by API-layer operations (workflow enqueue, etc.) and by
direct calls to ProvenanceRegistry.register_event().