Bulk import
Loading many rows at once, from a CSV file, a filtered copy of another table, or an external API.
Table of contents
Bulk import is a data-API capability, not a dedicated screen in the admin or tenant console. You call it directly (see API reference → Bulk endpoints), or it’s surfaced as buttons in whatever frontend you build against the generated dashboard config — the config’s
bulkActionsfeature wires each table to whichever of the three sources below you asked Generate Config to include for it.
Three sources
| Source | Good for |
|---|---|
| CSV file upload | A one-off import from a spreadsheet export — includes a downloadable template with the right headers, foreign-key and enum hints baked in. |
| Filtered copy from another table | Cloning/archiving rows already in your database into a different table, optionally stamping fixed override values on every copied row. |
| External API | Pulling from a third-party API and mapping its fields onto your table’s columns. |
Full endpoint shapes, request/response examples, and execution-mode behavior (realtime vs. background, based on row count): API reference → Bulk endpoints.
Realtime vs. background
A small import runs synchronously and you get the final result immediately; a large one is queued and you get a job id back instead. The threshold is configurable platform-wide (bulk.realtime_row_threshold, see Admin console → Settings → Host Settings), or you can force either mode explicitly per request.
Row policies apply during import too
If the caller’s role carries a row policy, rows that would violate it are rejected rather than written — for the filter and API-source sources, this is checked before anything is inserted; a batch where every row gets rejected comes back as a downloadable CSV of the rejected rows and why, instead of a misleadingly empty success response.
Partial failures
A batch with some bad rows still inserts everything that’s valid — the response (or, for a realtime run with failures, a downloadable CSV) tells you exactly which rows failed and why, so you can fix and re-run just those instead of starting over.