No benchmark result is currently published for this repository. The schema and data generator needed for the earlier workflow_events benchmark are not part of the active EF Core model, so the former speedup figures and sample result must not be treated as measured evidence.
- Add a versioned EF Core migration for the tables and indexes under test.
- Generate a deterministic dataset with a documented row count, data distribution, and random seed.
- Record the exact query, PostgreSQL version, hardware/container limits, indexes, and warm/cold cache conditions.
- Capture
EXPLAIN (ANALYZE, BUFFERS)before and after each index change. - Run repeated samples, report median and tail latency, and include the raw output with the change.
- Verify that indexes improve the measured workload without unacceptable write or storage cost.
Until those steps are completed, this document describes a proposed experiment only; it makes no latency or speedup claim.