Rangsh commented on issue #12058: URL: https://github.com/apache/seatunnel/issues/12058#issuecomment-5672694380
Thanks — agreed on the narrow conclusion. This particular flush-collapse A/B is ruled out only for the local file:/// harness; I will not generalize it to a real HDFS/DFS client, and I will not propose a production change from the current evidence. I will keep #12058 open. #12081 remains Related-only and cannot close this issue; its Build check and correctness work will be evaluated independently of this benchmark conclusion. I also treat InvocationFuture.get / LockSupport.park as a wait site, not yet a proven cause. Next I will stay on unchanged `dev` and run one controlled experiment on that wait — capturing wake-up/notification and scheduler timing, plus raw per-iteration data and exact JDK / machine / JMH / storage settings — to distinguish a causal wait condition from a passive symptom, then post the results here. -- 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]
