-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/40880/#review111170
-----------------------------------------------------------

Ship it!


Looks good to me. Testing something on the containerizer sounds more accurate 
than waiting for the slave reregistered ack to be received from the master, 
since we're really focused on the executor reregistering with the slave rather 
than the slave with the master.

- Adam B


On Dec. 3, 2015, 11:02 a.m., Joseph Wu wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/40880/
> -----------------------------------------------------------
> 
> (Updated Dec. 3, 2015, 11:02 a.m.)
> 
> 
> Review request for mesos, Bernd Mathiske, Artem Harutyunyan, and Till 
> Toenshoff.
> 
> 
> Bugs: MESOS-4047
>     https://issues.apache.org/jira/browse/MESOS-4047
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> `MesosContainerizerSlaveRecoveryTest.ResourceStatistics` has very similar 
> logic for restarting an agent, re-registering the executor, and even getting 
> `ResourceStatistics`.  But 
> `MesosContainerizerSlaveRecoveryTest.ResourceStatistics` is stable.
> 
> This patch updates the flaky test's wait-for-agent-recovery logic to match 
> the stable test.
> 
> 
> Diffs
> -----
> 
>   src/tests/containerizer/memory_pressure_tests.cpp 
> e18b971c4df26c9b9c103ca73bdad4fd400d6c02 
> 
> Diff: https://reviews.apache.org/r/40880/diff/
> 
> 
> Testing
> -------
> 
> On Ubuntu 14:
> `make check`
> `sudo bin/mesos-tests.sh 
> --gtest_filter="*MemoryPressureMesosTest.CGROUPS_ROOT_SlaveRecovery*" 
> --gtest_repeat=-1 --gtest_break_on_failure`
> 
> ^ Ran the above until satisfied.
> 
> 
> Thanks,
> 
> Joseph Wu
> 
>

Reply via email to