Sure @lindong28, I believe the issue we were seeing was that
`producer.send(record)` would fail silently because we were not calling
`producer.send(record).get()`, but
`producer.partitionsFor(topic).asScala.find(_.partition() ==
0).get.leader().id() != leaderServerId` would return true anyway. Because the
producer send was never retried, the test below times out
```
// Consumer should receive some messages
TestUtils.waitUntilTrue(() => {
consumer.poll(0).count() > 0
}, "Expected some messages", 3000L)
```
This patch ensures that both
`producer.partitionsFor(topic).asScala.find(_.partition() ==
0).get.leader().id() != leaderServerId` passes and we successfully send a
record before breaking out of the `waitUntilTrue` loop and progressing to the
rest of the test.
[ Full content available at: https://github.com/apache/kafka/pull/5695 ]
This message was relayed via gitbox.apache.org for [email protected]
