> On Aug. 9, 2017, 6:04 p.m., Benjamin Mahler wrote: > > src/slave/slave.cpp > > Lines 4987-4989 (original), 4987-4989 (patched) > > <https://reviews.apache.org/r/61524/diff/1/?file=1793980#file1793980line4991> > > > > Perhaps a note here saying that the master injects the executor for > > tasks within a task group?
In the interest of time, I followed up with patches to add a new test and update comments: https://reviews.apache.org/r/61534/ https://reviews.apache.org/r/61535/ - Greg ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/61524/#review182502 ----------------------------------------------------------- On Aug. 9, 2017, 1:14 p.m., Andrei Budnik wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/61524/ > ----------------------------------------------------------- > > (Updated Aug. 9, 2017, 1:14 p.m.) > > > Review request for mesos, Benjamin Mahler, Greg Mann, and Vinod Kone. > > > Bugs: MESOS-7871 > https://issues.apache.org/jira/browse/MESOS-7871 > > > Repository: mesos > > > Description > ------- > > Previously, `getExecutorInfo` was called only in `runTask`, so it was > validating an invariant related to launching a single task. The > invariant is that a task should have either CommandInfo or ExecutorInfo > set but not both. Now `getExecutorInfo` is also called to calculate > allocated resources which might be related to a task group, which might > violate the invariant. > > > Diffs > ----- > > src/slave/slave.cpp 7381530515f86faf4c3e8f82bcd9483f6cf0498b > > > Diff: https://reviews.apache.org/r/61524/diff/1/ > > > Testing > ------- > > sudo make check > > > Thanks, > > Andrei Budnik > >
