----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13455/#review24960 -----------------------------------------------------------
Ship it! src/sched/sched.cpp <https://reviews.apache.org/r/13455/#comment49129> indentation? src/slave/slave.cpp <https://reviews.apache.org/r/13455/#comment49130> s/to set the framework id/to ensure framework id exists/ - Vinod Kone On Aug. 10, 2013, 12:15 a.m., Ben Mahler wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/13455/ > ----------------------------------------------------------- > > (Updated Aug. 10, 2013, 12:15 a.m.) > > > Review request for mesos, Benjamin Hindman and Vinod Kone. > > > Repository: mesos-git > > > Description > ------- > > When a Slave re-registers with the master, it sends the ExecutorInfos for > running executors *and sets the framework_id in the process*. > > This is problematic because the Master expects the ExecutorInfo to be > immutable inside ExecutorInfoChecker. That is, any new tasks launched with > the same ExecutorID will be considered invalid by the master since the slave > mutated the ExecutorInfo by setting framework_id. > > To fix this issue, I've modified the Scheduler Driver to set the framework_id > when it is missing. > > See the following review for more information: > https://reviews.apache.org/r/13382/ > > > Diffs > ----- > > src/master/master.cpp a2f8929987f95e796aba71063223aea7003bf3df > src/sched/sched.cpp 5718fe9280d7e1124ffef2de9e4d47b347f0600a > src/slave/slave.cpp 3b4911844e19c85a97c79b22ef90897b26f26aac > > Diff: https://reviews.apache.org/r/13455/diff/ > > > Testing > ------- > > make check > > > Thanks, > > Ben Mahler > >
