Use case · load and performance
Load and performance testing for document workflows.
Document systems often behave differently when queues fill, indexes grow, or many formats arrive together. Repeatable packs let teams run controlled load tests without recycling production records.
The testing job
Exercise the system with known inputs.
Use a defined pack and repeatable request pattern to inspect throughput, latency, memory, queue handling, retries, and failure recovery. Match the volume to the actual release contents rather than assuming a package size.
- Ingestion and parsing throughput
- Indexing and retrieval under concurrent requests
- Mixed document sizes and formats
- Retries, back-pressure, and partial failures
Useful checks
Measure the workflow, not just the endpoint.
Record the conditions of each run and separate infrastructure limits from document-specific failures. Synthetic inputs make the comparison repeatable while leaving production data out of the test environment.
Deniable approach
Scale is a property of the selected release.
Deniable does not invent volume claims. Each pack states its verified document count and formats so teams can choose a release appropriate to their test run.