> On June 16, 2015, 5:09 p.m., Niklas Nielsen wrote: > > src/master/allocator/sorter/drf/sorter.cpp, line 177 > > <https://reviews.apache.org/r/35473/diff/2/?file=984869#file984869line177> > > > > This isn't just fixing the check's then but is introducing a fast path? > > Should we capture this in the review description? Or maybe comment here?
Niq, this is not just a fast path. The invariant here is that we don't create a key (i.e., slaveId) in 'resources' unless there is a NON EMPTY resources associated with that slave. And we'll delete the key if the corresponding resources become empty. - Jie ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/35473/#review88090 ----------------------------------------------------------- On June 15, 2015, 10:03 p.m., Jie Yu wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/35473/ > ----------------------------------------------------------- > > (Updated June 15, 2015, 10:03 p.m.) > > > Review request for mesos, Ben Mahler, Niklas Nielsen, and Vinod Kone. > > > Bugs: MESOS-2627 > https://issues.apache.org/jira/browse/MESOS-2627 > > > Repository: mesos > > > Description > ------- > > Removed a few incorrect CHECKs in DRF sorter. > > See details in the ticket for the reason why this is needed. > > > Diffs > ----- > > src/master/allocator/sorter/drf/sorter.cpp > ac05afdc7d408735dd796faa68c943e75540aaa7 > src/tests/examples_tests.cpp 41a9265e91ea54b657afbd32c22e204c5f64ca17 > > Diff: https://reviews.apache.org/r/35473/diff/ > > > Testing > ------- > > sudo make check > > > Thanks, > > Jie Yu > >
