[ARTEMIS-2016] Wait.waitFor is sometimes called with zero timeout (durationMillis), or with durationMillis < sleepMillis

• Type: Test
• Status: Closed
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: 2.7.0
• Fix Version/s: 2.6.3
• Flags: Patch
• Assignee: Unassigned
• Reporter: Jiri Daněk
• Created: 08/Aug/18 08:13
• Updated: 10/Sep/18 08:18
• Resolved: 31/Aug/18 01:55

Description

timeout = 0 happens in org.apache.activemq.artemis.tests.integration.clientcrash.ClientTestBase#assertActiveConnections(int) and org.apache.activemq.artemis.tests.integration.clientcrash.ClientTestBase#assertActiveSession(int).

durationMillis < sleepMillis happens in org.apache.activemq.artemis.tests.integration.management.QueueControlTest#assertMetrics

I am not sure what the intent of the author of the test was, but the current state seems wrong. The test QueueControlTest#testRemoveAllWithPagingMode is flaky due to insufficient number of retry attempts in QueueControlTest#assertMetrics.

Issue Links

• links to: GitHub Pull Request #2226