| Failure | Cause | Fix |
|---|---|---|
FileNotFoundError on input data |
Input .pbf/.csv/.gpkg not in input-data/<state>/ |
Download missing file (links in README), place in correct path |
MemoryError / OOM during join |
Large state .pbf processed in memory |
Reduce scope by filtering to a county first |
| Osmium tool not found | Not installed | brew install osmium-tool |
| GDAL import error | GDAL not installed or Python binding missing | brew install gdal, reinstall geopandas |
Dask partial output in result_directory/ |
Pipeline interrupted mid-run | Delete output-data/<state>/csv-files/result_directory/ and re-run |
| JOSM scripting plugin error | Plugin version mismatch | Update JOSM and scripting plugin to latest |
No database mutations. To reset: delete the output directory and re-run.
rm -rf hydrography-approach/output-data/<state>/
cd hydrography-approach && python run-hydrography-pipeline.py- Output is stdout only (
print()statements in pipeline scripts) - No external monitoring or alerting
None.
To reprocess with updated source data:
- Download fresh input files (Geofabrik, FHWA NBI, USGS NHD)
- Replace files in
input-data/<state>/ - Delete
output-data/<state>/ - Re-run the pipeline