Archived RSMS documentation (Parquet / DuckDB era)
These files record historical design and phased plans for storing or querying RSMS scenario results with Parquet files and/or DuckDB. The current product path is rsms-worker-function/rsms_results_storage/ → Azure Table Storage (PlumeByTime / PlumeByMile, float32) plus edge_timeseries.json and mass_balance.json blobs, read by rsms-api-fastapi/app/results/.
Do not follow these archives for new work. Use:
- rsms-results-storage-design — normative storage layout
- rsms-results-table-storage-rework-plan — deployment/phased context (Parquet migration language removed from the active doc body where possible)
- rsms-master-tasklist — living TODO / in progress index
| Archive file | Former role |
|---|---|
| rsms-api-results-duckdb-design | Chart ↔ HTTP route mapping during the DuckDB-named era (Table + blob semantics described here are mostly still accurate; filename is legacy). |
| rsms-api-results-implementation-plan | API migration checklist away from Parquet/DuckDB reads. |
| rsms-results-storage-implementation-plan | Parquet writer / DuckDB validation phased plan. |
Git history also preserves earlier text if a section was trimmed from the active docs.