> On Jan. 22, 2016, 8:29 a.m., Ben Mahler wrote: > > src/tests/hierarchical_allocator_tests.cpp, lines 509-515 > > <https://reviews.apache.org/r/42629/diff/1/?file=1205229#file1205229line509> > > > > Looks like you only need the second advance here? Technically, you are > > triggering two batch allocations here, which doesn't appear to be the > > intent? > > > > It would mean we need a Clock::settle after calling recoverResources.
I was thinking about the following: if we remove the first advance, can there be a race when `expire()` is processed *after* the second allocation cycle? I think it can if we do not ensure the filter is set *before* the first batch allocation (your addition). Re: triggering two batch allocations here, it doesn't matter because we are interested only in the next allocation, but I agree it may distract the reader. Marking as "fixed". > On Jan. 22, 2016, 8:29 a.m., Ben Mahler wrote: > > src/tests/hierarchical_allocator_tests.cpp, line 525 > > <https://reviews.apache.org/r/42629/diff/1/?file=1205229#file1205229line525> > > > > Some unicode character here? :facepalm: > On Jan. 22, 2016, 8:29 a.m., Ben Mahler wrote: > > src/tests/hierarchical_allocator_tests.cpp, line 1415 > > <https://reviews.apache.org/r/42629/diff/1/?file=1205229#file1205229line1415> > > > > What is a Quarantee? ;) double :facepalm: ... > On Jan. 22, 2016, 8:29 a.m., Ben Mahler wrote: > > src/tests/hierarchical_allocator_tests.cpp, lines 1513-1522 > > <https://reviews.apache.org/r/42629/diff/1/?file=1205229#file1205229line1513> > > > > This change actually belongs in the previous patch, since your last > > change breaks this test on its own. Correct, thanks Ben! > On Jan. 22, 2016, 8:29 a.m., Ben Mahler wrote: > > src/tests/hierarchical_allocator_tests.cpp, lines 1497-1502 > > <https://reviews.apache.org/r/42629/diff/1/?file=1205229#file1205229line1497> > > > > Let's leave this for now, but it would be great to avoid this > > assumption and explicitly control the allocation interval duration, no? > > When you are advancing the clock below twice, you are triggering two batch > > allocations, when your intention appears to be to trigger only one batch > > allocation. > > > > For now I'll fix this same issue in the OfferFilter test, and I'll > > leave this one as still triggering two allocations since it's an existing > > test. Yeah, may be a bit misleading for the reader. I'll note it down and propose a patch some time in the future. - Alexander ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/42629/#review115801 ----------------------------------------------------------- On Jan. 22, 2016, 1:24 a.m., Alexander Rukletsov wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/42629/ > ----------------------------------------------------------- > > (Updated Jan. 22, 2016, 1:24 a.m.) > > > Review request for mesos and Ben Mahler. > > > Bugs: MESOS-4302 > https://issues.apache.org/jira/browse/MESOS-4302 > > > Repository: mesos > > > Description > ------- > > See summary. > > > Diffs > ----- > > src/tests/hierarchical_allocator_tests.cpp > 953712149bd951789beb29c72779c4ac65aa48dc > > Diff: https://reviews.apache.org/r/42629/diff/ > > > Testing > ------- > > On Mac OS 10.10.4: > > `make check` > > `GTEST_FILTER="HierarchicalAllocatorTest.FilterTimeout" ./bin/mesos-tests.sh > --gtest_repeat=100 --gtest_break_on_failure` passes with the patch and fails > without. > > `GTEST_FILTER="HierarchicalAllocatorTest.*" ./bin/mesos-tests.sh > --gtest_repeat=100 --gtest_break_on_failure` > > > Thanks, > > Alexander Rukletsov > >
