> On March 18, 2016, 11:24 p.m., Ben Mahler wrote: > > src/tests/slave_tests.cpp, lines 3317-3323 > > <https://reviews.apache.org/r/44994/diff/2/?file=1306236#file1306236line3317> > > > > It seems fine to expect this but arguably if what we care about in this > > test is **how** the agent chooses the shutdown timeout, a FUTURE_DISPATCH > > on Slave::shutdownExecutorTimeout seems sufficient and closer reflects what > > is being tested? This avoids the need for the previous changes to the > > TestContainerizer AFAICT?
I agree it is cleaner and better expresses the intention. But for that, we should promote `Slave::shutdownExecutorTimeout()` to public. I was thinking to avoid this by updating the test containerizer. I'm with replacing it with a `FUTURE_DISPATCH` on `Slave::shutdownExecutorTimeout()`. - Alexander ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/44994/#review124308 ----------------------------------------------------------- On March 18, 2016, 5:13 p.m., Alexander Rukletsov wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/44994/ > ----------------------------------------------------------- > > (Updated March 18, 2016, 5:13 p.m.) > > > Review request for mesos and Ben Mahler. > > > Bugs: MESOS-4949 > https://issues.apache.org/jira/browse/MESOS-4949 > > > Repository: mesos > > > Description > ------- > > See summary. > > > Diffs > ----- > > src/tests/slave_tests.cpp ea1d776077bf638885db8421194aa4427c772169 > > Diff: https://reviews.apache.org/r/44994/diff/ > > > Testing > ------- > > On Mac OS 10.10.4: > `make check` > `GLOG_v=2 GTEST_FILTER="*SlaveTest*" ./bin/mesos-tests.sh --gtest_repeat=100 > --gtest_break_on_failure` > > > Thanks, > > Alexander Rukletsov > >
