SEZ9 commented on issue #12058: URL: https://github.com/apache/seatunnel/issues/12058#issuecomment-5650147685
@Rangsh thanks — the boundaries for this experiment are already set in the previous reply, so nothing new to add there: only the `HdfsWriter.flush()` control-flow change on the unchanged dev baseline, write-through and the caller's completion wait preserved with one hsync-family call per successful append, the local `file:///` result labelled as filesystem-specific, and the same-environment wall, CPU, GC, raw per-fork and Score/Error/CV artifacts with a comparison of the original `InvocationFuture.get`/park hotspot and the storage sync path rather than aggregate scores alone. Agreed that the related PR stays Related only and that a before/after of the full PR is not evidence for this issue. Please go ahead with the A/B and report either outcome here before proposing any production conclusion. <!-- streview-comment:1000 --> -- 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]
