Hi David, We are also working on fixing flaky tests and e2e :)
Best, Chia-Ping > Justine Olshan <jols...@confluent.io.invalid> 於 2024年12月19日 凌晨3:30 寫道： > > Hey David, > It was actually my plan to look at those today :) > > Justine > >> On Wed, Dec 18, 2024 at 10:50 AM David Arthur <mum...@gmail.com> wrote: >> >> Hey folks! >> >> Now that the 4.0 branch has been created, I think we need some focus on >> stabilizing a few flaky tests. While it is a good practice to keep our >> mainline branch as green as possible, I think it is especially important >> for release branches to be green. >> >> Here is a report of the current situation on trunk which is essentially the >> same as 4.0 right now. >> >> https://github.com/apache/kafka/actions/runs/12398964575#summary-34612865771 >> >> We have a number of tests which have been quarantined, but remain unfixed. >> These are listed in the "Most Problematic Tests" (ignore the results with a >> low number of Build Scans, that's a report bug). >> >> In particular some transactions tests (both tiered and non-tiered) remain >> flaky. These are tracked by KAFKA-18092, KAFKA-18035, and KAFKA-18036. >> >> * testBumpTransactionalEpochWithTV2Enabled(String, String, boolean)[1] >> * testBumpTransactionalEpochWithTV2Disabled(String, String, boolean)[2] >> * testBumpTransactionalEpochWithTV2Disabled(String, String, boolean)[1] >> * testBumpTransactionalEpochWithTV2Enabled(String, String, boolean)[2] >> >> As per usual, PlaintextConsumerTest has some flakiness (tracked by >> KAFKA-18031 and KAFKA-15920) >> >> If anyone has knowledge on these areas and could investigate the flakiness, >> it would be super helpful. I'm sure David J (the 4.0 release manager) would >> love to see green builds :) >> >> Cheers, >> David A >>
