soumadipcodeEnthusiast commented on issue #16403: URL: https://github.com/apache/dubbo/issues/16403#issuecomment-5200830612
@juzi050 @014-code — since there are already two PRs up for this (#16411 and #16412), a few questions before I try anything myself, so I don't duplicate effort: Do either of the two PRs cover all four terminal paths called out in the issue — successful completion, failure, cancellation, and transport closure? The issue specifically notes onDataFinally() is a no-op in GenericHttp2ServerTransportListener, so I want to check cancellation/transport-closure cleanup isn't only handled on the happy path. How do the two approaches handle the multipart safety concern mentioned in the issue — destroying the decoder too early would break FileUpload objects returned by part()/parts() if application code hasn't consumed them yet. Do both PRs defer cleanup to the same terminal lifecycle hook, or do they differ here? Is one of the two approaches currently favored (more complete, better test coverage), or are they both still open for comparison? Since a fix is already in flight from two directions, would it be more useful for me to review/test #16411 and #16412 against the multipart + cancellation edge cases rather than write a third implementation? Happy to help either way. -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
