> On Oct. 7, 2016, 11:51 p.m., Anand Mazumdar wrote: > > src/tests/master_maintenance_tests.cpp, lines 1169-1171 > > <https://reviews.apache.org/r/52620/diff/2/?file=1526600#file1526600line1169> > > > > We prefer to keep expectations close to the actual business logic for > > readability. Can you move this down to L1192?
I'm dropping this issue as discussed on IRC. The reason for it is that the framework will receive an inverse offer as soon as it subscribes. That's because as soon as the offer is sent out, the resources are "allocated", which causes the allocator to immediately send an inverse offer. > On Oct. 7, 2016, 11:51 p.m., Anand Mazumdar wrote: > > src/tests/master_maintenance_tests.cpp, lines 1529-1531 > > <https://reviews.apache.org/r/52620/diff/2/?file=1526600#file1526600line1529> > > > > Similar comment as per previous test. Move this expectation. Dropping for the same reason as in previous test. - Ilya ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/52620/#review151890 ----------------------------------------------------------- On Oct. 10, 2016, 8:18 p.m., Ilya Pronin wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/52620/ > ----------------------------------------------------------- > > (Updated Oct. 10, 2016, 8:18 p.m.) > > > Review request for mesos and Anand Mazumdar. > > > Bugs: MESOS-4948 > https://issues.apache.org/jira/browse/MESOS-4948 > > > Repository: mesos > > > Description > ------- > > Updated InverseOffers* maintenance tests to use the new scheduler mock. > > > Diffs > ----- > > src/tests/master_maintenance_tests.cpp > 77eb405ab7314da906bed9ec1d0018c24928d8d8 > > Diff: https://reviews.apache.org/r/52620/diff/ > > > Testing > ------- > > Tested on OS X 10.11 and Ubuntu 14.04. > ``` > make check GTEST_FILTER='MasterMaintenanceTest.*' GTEST_REPEAT=1000 > GTEST_BREAK_ON_FAILURE=1 > ``` > Noticed a segfault caused by getenv / setenv race (MESOS-3475), but it's > unrelated to this change and is reproducible on current master. > > > Thanks, > > Ilya Pronin > >
