> On March 16, 2020, 8:44 p.m., Greg Mann wrote: > > src/tests/containerizer/cgroups_isolator_tests.cpp > > Lines 465-473 (original), 525-533 (patched) > > <https://reviews.apache.org/r/71953/diff/5/?file=2214045#file2214045line525> > > > > Why are our expectations for CPU consumption changed after your update > > to the test? Aren't we still setting the CFS quota in the same way? > > Qian Zhang wrote: > That's actually a bug in this test: besides the CPU times consumed by the > task, there are also some overhead caused by the executor (e.g. launching > executor will also take some CPU times), so in total the actually CPU usage > will be a bit higher due to the overhead.
Previously in this test we did add some overhead (0.05) when checking the CPU time, however that overhead is not big enough so this test fails occasionally. - Qian ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/71953/#review219956 ----------------------------------------------------------- On March 16, 2020, 5:12 p.m., Qian Zhang wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/71953/ > ----------------------------------------------------------- > > (Updated March 16, 2020, 5:12 p.m.) > > > Review request for mesos, Andrei Budnik and Greg Mann. > > > Bugs: MESOS-10047 > https://issues.apache.org/jira/browse/MESOS-10047 > > > Repository: mesos > > > Description > ------- > > This test is renamed to `ROOT_CGROUPS_CFS_CommandTaskNoLimits`, and > besides CFS quota, now it also verifies CPU shares, memory soft and > hard limits and OOM score adjustment. > > > Diffs > ----- > > src/tests/containerizer/cgroups_isolator_tests.cpp > f72e6cdab417368e63349915114aeed586e0ef0e > > > Diff: https://reviews.apache.org/r/71953/diff/5/ > > > Testing > ------- > > sudo make check > > > Thanks, > > Qian Zhang > >
