> On Oct. 12, 2016, 10:52 p.m., Vinod Kone wrote: > > src/launcher/default_executor.cpp, line 700 > > <https://reviews.apache.org/r/52755/diff/1/?file=1531440#file1531440line700> > > > > can you add a comment on how this can happen when we are in `CONNECTED` > > state? it's not obvious to me.
Sure, will do. It can happen in `CONNECTED` state if the agent explicitly asks the executor to shut down if it's not able to subscribe within the re-register timeout for some reason. - Anand ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/52755/#review152413 ----------------------------------------------------------- On Oct. 12, 2016, 12:15 a.m., Anand Mazumdar wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/52755/ > ----------------------------------------------------------- > > (Updated Oct. 12, 2016, 12:15 a.m.) > > > Review request for mesos and Vinod Kone. > > > Bugs: MESOS-6363 > https://issues.apache.org/jira/browse/MESOS-6363 > > > Repository: mesos > > > Description > ------- > > Previously, the default executor used to crash with a failed assertion > when the executor library injected a shutdown event when it noticed > a disconnection with the agent for non-checkpointed frameworks and > upon recovery timeout for checkpointed frameworks. This change > modifies it to commit suicide minus the failed assertion. > > > Diffs > ----- > > src/launcher/default_executor.cpp 2454bd7df608254af72af01460782f5ab78a19c1 > > Diff: https://reviews.apache.org/r/52755/diff/ > > > Testing > ------- > > Manual testing (It's hard to test this directly owing to the executor library > immediately invoking the `connected()` callback since the libprocess instance > initialized in tests is always active!) > > Filed MESOS-6373 to address this. > > > Thanks, > > Anand Mazumdar > >
