- Extend mod-search.yaml with new API /instance/ids/jobs
- Add unit/integration tests
- New API for /instance/ids/jobs is exist
- Unit/Integration tests created
A possible solution for streaming large volumes of IDs. To begin with, it is worth dividing streaming into two operations. Stream ID from the ES to temporary storage, and the next step is to stream unique IDs to the end-user. As temporary storage, it is most convenient to use the Postgres database.
At request, it is possible to create a temporary table in which to add the IDs from the ES. To optimize export, name the table with the CQL query hash and reuse it. A periodic job can delete result tables after it is no longer needed.