> On March 13, 2014, 10:11 p.m., Adam B wrote: > > src/slave/slave.cpp, line 3413 > > <https://reviews.apache.org/r/18403/diff/4/?file=519536#file519536line3413> > > > > Initializing resources in the constructor is great, but perhaps > > Executor::finalize(ExecInfo) should set resources = info.resources(); Then > > you wouldn't even need to set resources in the other constructor, since the > > call to finalize would do it for you. > > Or maybe finalize needs to do a MergeFrom, in case the TaskInfo > > call-path already set resources before the executor could finish launching.
Great point! Thanks for catching this. Should be fixed in the latest patch. > On March 13, 2014, 10:11 p.m., Adam B wrote: > > src/slave/slave.cpp, lines 3074-3075 > > <https://reviews.apache.org/r/18403/diff/3-4/?file=506155#file506155line3074> > > > > This error message is confusing/inaccurate. The problem is that the > > executor is already known (executors does contain executorId). Somebody's > > trying to launch a task for an executor that's already running on that > > slave. Can you reword the error to clarify what the problem is? True - I added more context to the error message in the new patch. - Niklas ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/18403/#review37157 ----------------------------------------------------------- On March 17, 2014, 10:37 a.m., Niklas Nielsen wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/18403/ > ----------------------------------------------------------- > > (Updated March 17, 2014, 10:37 a.m.) > > > Review request for mesos, Adam B, Benjamin Hindman, Ben Mahler, Ian Downes, > Till Toenshoff, and Vinod Kone. > > > Bugs: MESOS-922 > https://issues.apache.org/jira/browse/MESOS-922 > > > Repository: mesos-git > > > Description > ------- > > This patch delegates the choice of executor to the containerizer by removing > executorInfo dependencies up until Containerizer::launch(). > Containerizer::launch() now returns a future to the executor info that is > being run and the slave creates the corresponding executor structure when > launch completes. > This means message handling from the running executor to the slave in the > interim where the executor structure has not created, need to be enqueued > until executor is ready. So far, registerExecutor() and reregisterExecutor() > has been split into two continuations to deal with this issue. > > > Diffs > ----- > > src/slave/containerizer/containerizer.hpp d9ae326 > src/slave/containerizer/mesos_containerizer.hpp ee1fd30 > src/slave/containerizer/mesos_containerizer.cpp c819c97 > src/slave/http.cpp 594032d > src/slave/slave.hpp 01b80df > src/slave/slave.cpp d8d3e0f > src/tests/containerizer.hpp 5686398 > src/tests/containerizer.cpp bfb9341 > > Diff: https://reviews.apache.org/r/18403/diff/ > > > Testing > ------- > > make check > > > Thanks, > > Niklas Nielsen > >
