[
https://issues.apache.org/jira/browse/MESOS-7506?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16260729#comment-16260729
]
Alexander Rukletsov commented on MESOS-7506:
--------------------------------------------
{noformat}
Commit: 95decd404438abd422794524e01d72a889821566 [95decd4]
Author: Andrei Budnik <[email protected]>
Date: 21 November 2017 at 14:30:47 GMT+1
Committer: Alexander Rukletsov <[email protected]>
Commit Date: 21 November 2017 at 14:34:31 GMT+1
Fixed `wait()` and `destroy()` in composing containerizer.
Previously, `wait()` and `destroy()` methods of composing containerizer
returned a future that might be set to `READY` state while the internal
state of composing containerizer is not yet cleaned up.
This patch adds a `termination` promise to `Container` struct,
which is used to return a future from `wait()` and `destroy()` methods.
This promise is set to `READY` state iff related container is
completely destroyed.
`_destroy()` callback is subscribed to a future from `wait()`, which
is called on related containerizer, to propagate a value to the
`termination` promise and do the cleanup.
Review: https://reviews.apache.org/r/63887/
{noformat}
{noformat}
Commit: 84365a140c3730e2d6579ad500118d6749d2f87f [84365a1]
Author: Andrei Budnik <[email protected]>
Date: 21 November 2017 at 14:31:37 GMT+1
Committer: Alexander Rukletsov <[email protected]>
Commit Date: 21 November 2017 at 14:34:32 GMT+1
Updated composing containerizer tests.
Review: https://reviews.apache.org/r/63888/
{noformat}
> Multiple tests leave orphan containers.
> ---------------------------------------
>
> Key: MESOS-7506
> URL: https://issues.apache.org/jira/browse/MESOS-7506
> Project: Mesos
> Issue Type: Bug
> Components: containerization
> Environment: Ubuntu 16.04
> Fedora 23
> other Linux distros
> Reporter: Alexander Rukletsov
> Assignee: Andrei Budnik
> Labels: containerizer, flaky-test, mesosphere
> Attachments: KillMultipleTasks-badrun.txt,
> ROOT_IsolatorFlags-badrun.txt, ResourceLimitation-badrun.txt,
> ResourceLimitation-badrun2.txt,
> RestartSlaveRequireExecutorAuthentication-badrun.txt,
> TaskWithFileURI-badrun.txt
>
>
> I've observed a number of flaky tests that leave orphan containers upon
> cleanup. A typical log looks like this:
> {noformat}
> ../../src/tests/cluster.cpp:580: Failure
> Value of: containers->empty()
> Actual: false
> Expected: true
> Failed to destroy containers: { da3e8aa8-98e7-4e72-a8fd-5d0bae960014 }
> {noformat}
> All currently affected tests:
> {noformat}
> ROOT_DOCKER_DockerAndMesosContainerizers/DefaultExecutorTest.KillTask/0
> ROOT_DOCKER_DockerAndMesosContainerizers/DefaultExecutorTest.TaskWithFileURI/0
> ROOT_DOCKER_DockerAndMesosContainerizers/DefaultExecutorTest.ResourceLimitation/0
> ROOT_DOCKER_DockerAndMesosContainerizers/DefaultExecutorTest.KillMultipleTasks/0
> SlaveTest.RestartSlaveRequireExecutorAuthentication
> LinuxCapabilitiesIsolatorFlagsTest.ROOT_IsolatorFlags
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
