[Bug] Flaky Test RocksDBMessageStoreTest.testMaxOffset

[lizhanhui]

Before Creating the Bug Report

• I found a bug, not just asking a question, which should be created in GitHub Discussions.

• I have searched the GitHub Issues and GitHub Discussions of this repository and believe that this is not a duplicate.

• I have confirmed that this bug belongs to the current repository, not other repositories of RocketMQ.

Runtime platform environment

CI

RocketMQ version

CI

JDK Version

CI

Describe the Bug

RocksDBMessageStoreTest.testMaxOffset:625 expected:<[3]L> but was:<[1]L>

https://github.com/apache/rocketmq/actions/runs/7083756493/job/19276717683

Steps to Reproduce

See CI

What Did You Expect to See?

Test should pass

What Did You See Instead?

Test case failed

Additional Context

No response

[yp969803]

  public void testMaxOffset() throws InterruptedException {
        if (notExecuted()) {
            return;
        }
        int firstBatchMessages = 3;
        int queueId = 0;
        messageBody = storeMessage.getBytes();

        assertThat(messageStore.getMaxOffsetInQueue(messageTopic, queueId)).isEqualTo(0);

        for (int i = 0; i < firstBatchMessages; i++) {
            final MessageExtBrokerInner msg = buildMessage();
            msg.setQueueId(queueId);
            messageStore.putMessage(msg);
        }

        while (messageStore.dispatchBehindBytes() != 0) {
            TimeUnit.MILLISECONDS.sleep(1);
        }

        assertThat(messageStore.getMaxOffsetInQueue(messageTopic, queueId)).isEqualTo(firstBatchMessages);

Didn't find anything wrong in this code dont know why the test is failing

[lizhanhui]

messageStore.dispatchBehindBytes() != 0 does not guarantee that the store has completed building the queue index. Awaitbility is preferred over sleep-loop.