DanielLeens commented on issue #12060: URL: https://github.com/apache/seatunnel/issues/12060#issuecomment-5551704651
Thanks @goutamadwant. The reported measurements correctly separate cheaper untimed fixture cleanup from production write latency, and they do not establish a variance or throughput improvement. Please do not submit a production optimization from this evidence. For the next agreed protocol, keep two separately named measurements: startup-sensitive end-to-end storage behavior and steady-state store behavior after a defined readiness boundary. They must use the same JDK, Linux runner class, heap, storage configuration, warmup, measurement count, and reporting rule for baseline and candidate. Do not tune warmup, sleeps, or the fixture only to lower CV. If fixture cleanup is later changed, review it as a benchmark-fixture correctness change with its own tests and report it separately from any production claim. The current no-production-change conclusion is the right outcome until a matched CI measurement identifies a reproducible bottleneck. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
