mesos-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Alexander Rukletsov <ruklet...@gmail.com>
Subject Re: Review Request 64247: Fixed a flaky test case in reservation tests.
Date Wed, 06 Dec 2017 11:09:12 GMT

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/64247/#review192975
-----------------------------------------------------------


Fix it, then Ship it!





src/tests/reservation_tests.cpp
Line 942 (original)
<https://reviews.apache.org/r/64247/#comment271412>

    Let's resume here to finish the test with the running clock.


- Alexander Rukletsov


On Dec. 6, 2017, 10:46 a.m., Jan Schlicht wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/64247/
> -----------------------------------------------------------
> 
> (Updated Dec. 6, 2017, 10:46 a.m.)
> 
> 
> Review request for mesos, Alexander Rukletsov, Benjamin Bannier, and Jie Yu.
> 
> 
> Bugs: MESOS-8289
>     https://issues.apache.org/jira/browse/MESOS-8289
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> ReservationTest.MasterFailover could fail when parameterized with
> the 'RESOURCE_PROVIDER' capability. After the simulated master
> failover, the re-registered framework could receive an offer before
> the re-registered agent sends a 'UPDATE_SLAVE' message. On receiving
> this message, the master would rescind the offer and send out a new
> one. The test expected a single offer to be send. It has been changed
> to accept the first offer and ignore subsequent ones.
> 
> Also, the test case has been updated to run with paused clocks.
> 
> 
> Diffs
> -----
> 
>   src/tests/reservation_tests.cpp 7f1bb398c33b3b8cad371b49bdc4bc2454f6440d 
> 
> 
> Diff: https://reviews.apache.org/r/64247/diff/2/
> 
> 
> Testing
> -------
> 
> src/mesos-tests
> 
> `src/mesos-tests --gtest_filter=ResourceProviderCapability/ReservationTest.MasterFailover/1
--gtest_repeat=2000 --gtest_break_on_failure` while running `stress --cpu 8 --io 8` in the
background (on a 4 core machine)
> 
> 
> Thanks,
> 
> Jan Schlicht
> 
>


Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message