You are viewing a plain text version of this content. The canonical link for it is here.
Posted to reviews@mesos.apache.org by Neil Conway <ne...@gmail.com> on 2017/01/25 02:38:46 UTC

Review Request 55911: Fixed flakiness in TaskValidationTest.TaskReusesUnreachableTaskID.

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/55911/
-----------------------------------------------------------

Review request for mesos and Vinod Kone.


Bugs: MESOS-6983
    https://issues.apache.org/jira/browse/MESOS-6983


Repository: mesos


Description
-------

Fixed flakiness in TaskValidationTest.TaskReusesUnreachableTaskID.


Diffs
-----

  src/tests/master_validation_tests.cpp ce10ea4502d0a13faca28d288dd16cd5cb864f6e 

Diff: https://reviews.apache.org/r/55911/diff/


Testing
-------

`make check`

Without this change, this test fails on most platforms in the internal CI. With this change, no failures of this test are observed.


Thanks,

Neil Conway


Re: Review Request 55911: Fixed flakiness in TaskValidationTest.TaskReusesUnreachableTaskID.

Posted by Mesos Reviewbot <re...@mesos.apache.org>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/55911/#review162948
-----------------------------------------------------------



Patch looks great!

Reviews applied: [55911]

Passed command: export OS='ubuntu:14.04' BUILDTOOL='autotools' COMPILER='gcc' CONFIGURATION='--verbose' ENVIRONMENT='GLOG_v=1 MESOS_VERBOSE=1'; ./support/docker-build.sh

- Mesos Reviewbot


On Jan. 25, 2017, 2:38 a.m., Neil Conway wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/55911/
> -----------------------------------------------------------
> 
> (Updated Jan. 25, 2017, 2:38 a.m.)
> 
> 
> Review request for mesos and Vinod Kone.
> 
> 
> Bugs: MESOS-6983
>     https://issues.apache.org/jira/browse/MESOS-6983
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> Fixed flakiness in TaskValidationTest.TaskReusesUnreachableTaskID.
> 
> 
> Diffs
> -----
> 
>   src/tests/master_validation_tests.cpp ce10ea4502d0a13faca28d288dd16cd5cb864f6e 
> 
> Diff: https://reviews.apache.org/r/55911/diff/
> 
> 
> Testing
> -------
> 
> `make check`
> 
> Without this change, this test fails on most platforms in the internal CI. With this change, no failures of this test are observed.
> 
> 
> Thanks,
> 
> Neil Conway
> 
>


Re: Review Request 55911: Fixed flakiness in TaskValidationTest.TaskReusesUnreachableTaskID.

Posted by Neil Conway <ne...@gmail.com>.

> On Jan. 25, 2017, 7:31 p.m., Joseph Wu wrote:
> > If I understand this correctly, the test is failing (when run as ROOT) because:
> > 
> > * Agent 1 starts a container.
> > * Agent 1 is partitioned.
> > * Agent 2 starts.
> > * Agent 2 observes (via global cgroups) Agent 1's container as an orphan.
> > * Agent 2 kills Agent 1's container.
> > * During test teardown, Agent 1 tries to clean up the container destroyed by Agent 2, fails.
> > 
> > To fix this, you are simply cleaning up Agent 1 after you've partitioned it, as it doesn't come back during the lifetime of the test anyway.  Can you add something like this to the review description?

That's right -- will update the review description.


- Neil


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/55911/#review163007
-----------------------------------------------------------


On Jan. 25, 2017, 2:38 a.m., Neil Conway wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/55911/
> -----------------------------------------------------------
> 
> (Updated Jan. 25, 2017, 2:38 a.m.)
> 
> 
> Review request for mesos and Vinod Kone.
> 
> 
> Bugs: MESOS-6983
>     https://issues.apache.org/jira/browse/MESOS-6983
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> Fixed flakiness in TaskValidationTest.TaskReusesUnreachableTaskID.
> 
> 
> Diffs
> -----
> 
>   src/tests/master_validation_tests.cpp ce10ea4502d0a13faca28d288dd16cd5cb864f6e 
> 
> Diff: https://reviews.apache.org/r/55911/diff/
> 
> 
> Testing
> -------
> 
> `make check`
> 
> Without this change, this test fails on most platforms in the internal CI. With this change, no failures of this test are observed.
> 
> 
> Thanks,
> 
> Neil Conway
> 
>


Re: Review Request 55911: Fixed flakiness in TaskValidationTest.TaskReusesUnreachableTaskID.

Posted by Joseph Wu <jo...@mesosphere.io>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/55911/#review163007
-----------------------------------------------------------


Ship it!




If I understand this correctly, the test is failing (when run as ROOT) because:

* Agent 1 starts a container.
* Agent 1 is partitioned.
* Agent 2 starts.
* Agent 2 observes (via global cgroups) Agent 1's container as an orphan.
* Agent 2 kills Agent 1's container.
* During test teardown, Agent 1 tries to clean up the container destroyed by Agent 2, fails.

To fix this, you are simply cleaning up Agent 1 after you've partitioned it, as it doesn't come back during the lifetime of the test anyway.  Can you add something like this to the review description?

- Joseph Wu


On Jan. 24, 2017, 6:38 p.m., Neil Conway wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/55911/
> -----------------------------------------------------------
> 
> (Updated Jan. 24, 2017, 6:38 p.m.)
> 
> 
> Review request for mesos and Vinod Kone.
> 
> 
> Bugs: MESOS-6983
>     https://issues.apache.org/jira/browse/MESOS-6983
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> Fixed flakiness in TaskValidationTest.TaskReusesUnreachableTaskID.
> 
> 
> Diffs
> -----
> 
>   src/tests/master_validation_tests.cpp ce10ea4502d0a13faca28d288dd16cd5cb864f6e 
> 
> Diff: https://reviews.apache.org/r/55911/diff/
> 
> 
> Testing
> -------
> 
> `make check`
> 
> Without this change, this test fails on most platforms in the internal CI. With this change, no failures of this test are observed.
> 
> 
> Thanks,
> 
> Neil Conway
> 
>


Re: Review Request 55911: Fixed flakiness in TaskValidationTest.TaskReusesUnreachableTaskID.

Posted by Neil Conway <ne...@gmail.com>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/55911/
-----------------------------------------------------------

(Updated Jan. 25, 2017, 7:51 p.m.)


Review request for mesos and Vinod Kone.


Changes
-------

Improve commit description.


Bugs: MESOS-6983
    https://issues.apache.org/jira/browse/MESOS-6983


Repository: mesos


Description (updated)
-------

The previous coding of the test partitioned one agent and then launched
a second agent. However, when run as root, the second agent discovered
(via global cgroups) the task that was started by the first agent, which
it then proceeded to kill.

To avoid this, we now shutdown the first agent after partitioning it and
before starting the second agent.


Diffs (updated)
-----

  src/tests/master_validation_tests.cpp 914b3c5749f775cce806f37c26e67d1f0f5351c6 

Diff: https://reviews.apache.org/r/55911/diff/


Testing
-------

`make check`

Without this change, this test fails on most platforms in the internal CI. With this change, no failures of this test are observed.


Thanks,

Neil Conway