> On March 7, 2013, 6:25 p.m., Vinod Kone wrote: > > src/slave/process_based_isolation_module.cpp, line 332 > > <https://reviews.apache.org/r/9795/diff/1/?file=267811#file267811line332> > > > > I think it makes more sense to return an error here instead of > > returning a wrong value to the user!
I don't think I agree, because gathering all the children, and obtaining the status of each children does not happen atomically. Therefore, if there are any processes that go away during this time, we can't stat them! Maybe it's sufficiently rare, but I'd rather get an approximate answer when this happens, rather than give up, no? Say there are 50 processes, and while we stat, one goes away. I'd rather get the usage for the remaining 49 processes than nothing at all. > On March 7, 2013, 6:25 p.m., Vinod Kone wrote: > > src/slave/process_based_isolation_module.cpp, line 318 > > <https://reviews.apache.org/r/9795/diff/1/?file=267811#file267811line318> > > > > s/children/descendants/ I think it's clear that it's recursive children, no? - Ben ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/9795/#review17556 ----------------------------------------------------------- On March 7, 2013, 3:18 a.m., Ben Mahler wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/9795/ > ----------------------------------------------------------- > > (Updated March 7, 2013, 3:18 a.m.) > > > Review request for mesos, Benjamin Hindman and Vinod Kone. > > > Description > ------- > > See subject. > > > This addresses bug MESOS-370. > https://issues.apache.org/jira/browse/MESOS-370 > > > Diffs > ----- > > src/slave/process_based_isolation_module.cpp > ff98d105af675dfc66070feaa43b42c1aa438fd8 > > Diff: https://reviews.apache.org/r/9795/diff/ > > > Testing > ------- > > make check and added tests in a subsequent review > > > Thanks, > > Ben Mahler > >
