[MESOS-8617] Tests using default executor occasionally fail.

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 1.5.1, 1.6.0
• Labels: flaky-test
• Assignee: Meng Zhu
• Reporter: Alex R
• Created: 27/Feb/18 15:21
• Updated: 07/Apr/18 07:29
• Resolved: 06/Apr/18 06:44

Description

Task transition expectation can be violated resulting in a failing test, e.g.:

../../src/tests/master_tests.cpp:4134: Failure
      Expected: TASK_RUNNING
To be equal to: status1->state()
      Which is: TASK_LOST

List of known affected tests:

MasterTest.TasksEndpoint
LauncherAndIsolationParam/PersistentVolumeDefaultExecutor.ROOT_TaskGroupsSharingViaSandboxVolumes
SlaveTest.TerminalTaskContainerizerUpdateFailsWithLost
SlaveTest.KillQueuedTaskGroup
MesosContainerizer/DefaultExecutorTest.ROOT_LaunchGroupFailure

Also it looks like the example frameworks are affected as well:

ExamplesTest.JavaFramework
ExamplesTest.TestHTTPFramework
ExamplesTest.V0JavaFramework

Issue Links

• is caused by: MESOS-8624 Valid tasks may be explicitly dropped by agent due to race conditions
• is duplicated by: MESOS-8603 SlaveTest.TerminalTaskContainerizerUpdateFailsWithGone and SlaveTest.TerminalTaskContainerizerUpdateFailsWithLost are flaky