WXPNG commented on PR #12755: URL: https://github.com/apache/seatunnel/pull/12755#issuecomment-6162522421
Thanks @DanielLeens for the two rounds of careful review! I've addressed all the points: Round 1: - Issue 3: replaced the fixed 3s sleep with an explicit signal — the test's NodeContext now returns a joiner subclass that counts down a CountDownLatch on blacklist(...), and the master is started only after the latch fires (with an assertion that it fired). Same for the discovery test. - Issue 1: the lite-member branch now continues instead of falling through to tryClaimMastership() / setMasterAddressToJoin(null). - Issue 2: added a one-time INFO log when the lite member first enters the "all candidates temporarily blacklisted" state. - Issue 4: removed the hierarchy-only discoveryJoinerInheritsTcpIpRetryFix test. - Synced with the latest dev (includes #12756), so the Cassandra IT failures should now be resolved. Round 2: - Moved the retry-logged flag from an instance field to a local variable in joinViaPossibleMembers(), so a re-join (e.g. after hot restart) logs again. Local verification: the two test classes pass (4 tests, 0 failures). Could you please take another look and let me know if there is anything else that needs changing? Thanks again! -- 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]
