vamossagar12 commented on PR #13646: URL: https://github.com/apache/kafka/pull/13646#issuecomment-1548015263
Thanks @sambhav-jain-16 , @kirktrue , just to give more context, there are ITs in connect for EOS which rely upon the position() API to assert some messages ebing read. But they seem to have be failing now. What the tests do is to read the end offsets for Topic Partitions and keep consuming messages. The moment the last consumed offset exceeds the end offset, the test exits. Basically this block of code: https://github.com/apache/kafka/blob/512fd6e5cbc49371e9e761bc1973342f639abeb4/connect/runtime/src/test/java/org/apache/kafka/connect/util/clusters/EmbeddedKafkaCluster.java#L552-L580 In this case, the assertion fails because even though the last consumed offsets zooms past the endOffset read upfront, the number of records aren't what we are expecting it to be (even though we have already ensured that those many number of records are produced successfully). We can go ahead and look at other ways of `fixing` this tests but wanted to understand if there's a fundamental thing which has changed or we are missing? PS: this is the flaky test in question https://github.com/apache/kafka/blob/512fd6e5cbc49371e9e761bc1973342f639abeb4/connect/runtime/src/test/java/org/apache/kafka/connect/integration/ExactlyOnceSourceIntegrationTest.java#L376 -- 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: jira-unsubscr...@kafka.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org