----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/51470/#review147118 -----------------------------------------------------------
I think we should also bring back this function: `const hashmap<SlaveID, Resources>& total()`. - Anindya Sinha On Aug. 26, 2016, 10:53 p.m., Jiang Yan Xu wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/51470/ > ----------------------------------------------------------- > > (Updated Aug. 26, 2016, 10:53 p.m.) > > > Review request for mesos, Alexander Rukletsov, Anindya Sinha, Michael Park, > and Neil Conway. > > > Repository: mesos > > > Description > ------- > > This is a partial reversal of 5a63f2a5d2a5fe0e5315a5b8f79b75b99a6a5893. > We used to have this field but it turned out to be unnecessary for the > current use cases so it was removed. We however will need it for the > ongoing work to support tasks using shared resources. > > The details can be found in https://reviews.apache.org/r/45961 but in > essence, we now know of a case where the removal of resource quantities > from the sorter depends on the identity. e.g., To determine if the > quantity of a shared resource can be removed from the sorter we need to > know if there are no longer instances of **the same** (hence identity) > shared resource in the sorter. > > > Diffs > ----- > > src/master/allocator/sorter/drf/sorter.hpp > bc6bfb2d5d3b32d55be055a0514861b4e7d889bb > src/master/allocator/sorter/drf/sorter.cpp > ac85b327fc33d34246788e6a8c8bf5a486c61434 > > Diff: https://reviews.apache.org/r/51470/diff/ > > > Testing > ------- > > make check > > > Thanks, > > Jiang Yan Xu > >
