> On Oct. 8, 2014, 10:42 a.m., Vinod Kone wrote: > > src/tests/slave_tests.cpp, lines 1071-1075 > > <https://reviews.apache.org/r/23912/diff/6-7/?file=714593#file714593line1071> > > > > This has not been moved !?
I have been searching through the previous review comments, but have not been able to determine which one you are referring to. However, I strongly suspect that you want this code to be moved downwards, after driver.launchTasks(), just before driver.killTask. Correct? > On Oct. 8, 2014, 10:42 a.m., Vinod Kone wrote: > > src/slave/slave.cpp, line 1355 > > <https://reviews.apache.org/r/23912/diff/6-7/?file=714590#file714590line1355> > > > > I missed this before, but you also should erase the executorid if this > > is the last task on this executor. additionally, you want to remove the > > framework is this is the last executor. see _runTask() for details. Yes, thanks! I was wondering about this, but not enough, not aware how eagerly the slave "wants" to remove its framework info :-) The next patch will support that. The fix will actually be located in _runTask(). - Bernd ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/23912/#review55834 ----------------------------------------------------------- On Oct. 8, 2014, 3:57 a.m., Bernd Mathiske wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/23912/ > ----------------------------------------------------------- > > (Updated Oct. 8, 2014, 3:57 a.m.) > > > Review request for mesos. > > > Bugs: MESOS-947 > https://issues.apache.org/jira/browse/MESOS-947 > > > Repository: mesos-git > > > Description > ------- > > Fixes MESOS-947 "Slave should properly handle a killTask() that arrives > between runTask() and _runTask()". > > Slave::killTask() did not check for task in question combination to be > "pending" (i.e. Slave::runTask had happened, but Slave::_runTask had not yet) > and then erroneously assumed that Slave::runTask() had not been executed. The > task was then marked "LOST" instead of "KILLED". But Slave::runTask had > already scheduled Slave::_runTask to follow. Now the entry for being > "pending" is removed, and the task is marked "KILLED", and _runTask gets > informed about this. It checks whether the task in question is currently > "pending" and if it is not, then it infers that the task has been killed and > does not erroneously try to complete launching it. > > > Diffs > ----- > > src/slave/slave.hpp 28697102047b972ecb3b6b627ee089b430549fc0 > src/slave/slave.cpp e56dcbd80114730949a0d4b553470802a4d38281 > src/tests/mesos.hpp 957e2233cc11c438fd80d3b6d1907a1223093104 > src/tests/mesos.cpp 3dcb2acd5ad4ab5e3a7b4fe524ee077558112773 > src/tests/slave_tests.cpp 69be28f6e82b99e23424bd2be8294f715d8040d4 > > Diff: https://reviews.apache.org/r/23912/diff/ > > > Testing > ------- > > Wrote a unit test that reliably created the situation described in the > ticket. Observed that TASK_LOST and the listed log output occurred. This > pointed directly to the lines in killTask() where the problem is rooted. Ran > the test after fixing, it succeeded. Checked the log. It looks like a "clean > kill" now :-) > > > Thanks, > > Bernd Mathiske > >
