----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/50857/#review145549 -----------------------------------------------------------
Still hunting down the rogue `ConnectionProcess` that sometimes crashes the test during finalization... src/tests/scheduler_tests.cpp (line 1526) <https://reviews.apache.org/r/50857/#comment211829> This is already called in `MesosTest::SetUp()`. src/tests/scheduler_tests.cpp (lines 1540 - 1551) <https://reviews.apache.org/r/50857/#comment211830> Since we're already reinitializing in the `SetUp`, we don't need to reinitialize in `TearDown`. Instead, change this cleanup to `TearDownTestCase`. - Joseph Wu On Aug. 10, 2016, 3:37 p.m., Greg Mann wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/50857/ > ----------------------------------------------------------- > > (Updated Aug. 10, 2016, 3:37 p.m.) > > > Review request for mesos, Joseph Wu and Vinod Kone. > > > Bugs: MESOS-3753 > https://issues.apache.org/jira/browse/MESOS-3753 > > > Repository: mesos > > > Description > ------- > > This patch modifies the test `SchedulerTest.Teardown` to > be parametrized by both `ContentType` and SSL configuration, > and renames it to `SchedulerSSLTest.RunTaskAndTeardown`. > This allows the test to verify the scheduler's behavior with > SSL both enabled and disabled. > > > Diffs > ----- > > src/tests/scheduler_tests.cpp ccd0f2f602a7a1a9a44b68fd0f59bdc8e0fa58b1 > > Diff: https://reviews.apache.org/r/50857/diff/ > > > Testing > ------- > > This test is currently flaky in the SSL-enabled configurations and will > produce a segfault on my OSX machine after 50-100 repetitions. To reproduce, > try: > > `GTEST_REPEAT=1000 GTEST_BREAK_ON_FAILURE=1 > GTEST_FILTER="ContentTypeAndSSLConfig/SchedulerSSLTest.RunTaskAndTeardown/2" > bin/mesos-tests.sh` > > > Thanks, > > Greg Mann > >
