Problem
As we optimize the batch processor and NATS handling, we have no way to prove that changes actually improve speed or to catch "performance regressions" where a new feature accidentally slows down ingestion by 50%.
Proposed Solution
Add Go benchmarks (go test -bench) for critical paths: batch serialization, NATS message ingestion, and SQL template rendering. These will provide a baseline to ensure WaveHouse remains high-performance as it grows.
Alternatives Considered
Ad hoc testing with high volume data, but benchmarks provide a more stable and repeatable metric.
Additional Context
None
Problem
As we optimize the batch processor and NATS handling, we have no way to prove that changes actually improve speed or to catch "performance regressions" where a new feature accidentally slows down ingestion by 50%.
Proposed Solution
Add Go benchmarks (go test -bench) for critical paths: batch serialization, NATS message ingestion, and SQL template rendering. These will provide a baseline to ensure WaveHouse remains high-performance as it grows.
Alternatives Considered
Ad hoc testing with high volume data, but benchmarks provide a more stable and repeatable metric.
Additional Context
None