SEZ9 commented on issue #12456: URL: https://github.com/apache/seatunnel/issues/12456#issuecomment-6459147672
@wangbing505 Thanks, that plan works. To be plain about status: the `ReflectionUtils.CONSTRUCTOR_CACHE` cleanup is not in place yet, so there is nothing new to test right now. I will ping you on this thread as soon as it lands on `dev` so you can rerun the 0/5/10-job experiment on a build from that state. When you do rerun it, please keep the same methodology as your earlier snapshots: take each reading after a full GC, and publish the exact `SeaTunnelChildFirstClassLoader` count and Metaspace usage at 0, 5 and 10 jobs. If any loaders are still retained after the final job, a Path to GC Roots for one of them would be very helpful so we can see whether the remaining root is still `CONSTRUCTOR_CACHE` or something else. @modelpath-dev This should also give you the post-job snapshot you were waiting for to see whether the `SeaTunnelChildFirstClassLoader` instances actually drop once the job finishes. If you have your own run in the meantime, feel free to post the loader counts here too. <!-- streview-comment:2082 --> -- 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]
