> On Feb. 23, 2018, 5:38 p.m., Gilbert Song wrote: > > src/docker/docker.cpp > > Lines 151 (patched) > > <https://reviews.apache.org/r/65683/diff/5/?file=1963585#file1963585line151> > > > > Should we do `killtree()` even if the status is not pending (e.g., > > ready or failure)? > > > > If the semantic (caller invokes `.discard()` -> no longer care about > > the subprocess) is something we want to define for the library, should we > > always killtree regardless whatever status?
I think it's incorrect for us to kill the process if `s.status()` is no longer pending. In that case, the process has actually terminated, so the PID could be some completely unrelated process which was spawned recently. - Greg ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/65683/#review198188 ----------------------------------------------------------- On Feb. 23, 2018, 11:07 p.m., Greg Mann wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/65683/ > ----------------------------------------------------------- > > (Updated Feb. 23, 2018, 11:07 p.m.) > > > Review request for mesos, Andrei Budnik, Gilbert Song, Michael Park, and > Vinod Kone. > > > Bugs: MESOS-8575 > https://issues.apache.org/jira/browse/MESOS-8575 > > > Repository: mesos > > > Description > ------- > > Updated discard handling of the Docker 'inspect' call. > > > Diffs > ----- > > src/docker/docker.hpp d9e71f8841a868082170d28fc4f3d495e2eb1e61 > src/docker/docker.cpp 3df370e988fce12d323ff6b441da15dab27bdd28 > > > Diff: https://reviews.apache.org/r/65683/diff/6/ > > > Testing > ------- > > > Thanks, > > Greg Mann > >
