Rangsh commented on issue #12058: URL: https://github.com/apache/seatunnel/issues/12058#issuecomment-5674694348
@SEZ9 thanks — acknowledged on both points. **1. CPU / GC / after-side hsync** — agreed, record is complete; nothing further from me on that. **2. Flush-collapse experiment** — agreed, concluded for #12058 on the `file:///` / `LocalFileSystem` contract. Storage sync / flush-collapse is ruled out as the driver of the variance reported here. I will not generalize that null result to a real HDFS/DFS client, will not rerun the same single-variable flush A/B against a non-local filesystem to close this line, and will not propose a production change from this evidence. #12081 stays Related-only and cannot claim to address #12058 on this basis. **Next** - Keep #12058 open. - Next experiment on unchanged `dev`: one controlled question about the `InvocationFuture.get` / `LockSupport.park` wait — treated as a wait site, not a proven cause, until the data says otherwise. - When I post those results I will include wake-up/notification and scheduler timing sufficient to distinguish a causal wait condition from a passive symptom, raw per-iteration data, and the exact JDK, machine, JMH, and storage settings. Nothing else outstanding from the flush-path work on my side. Will follow up with the wait-site results. -- 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]
