Hello, I have a relatively simple fix for the bug <https://github.com/apache/kafka/pull/12566> reported in KAFKA-13927 <https://issues.apache.org/jira/browse/KAFKA-13927> in which Connect will not commit offsets for a message that a SinkTask retries and eventually succeeds. Would appreciate eyes or feedback on.
Also the CI appears not to succeed for the trunk base (and other branches right now); is that correct? I've added test assertions for this bug and verified that the unit tests for the SinkTask do succeed. Curious what is the course of action if tests do not pass in general for the base branch? Thanks, Jordan