----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/13791/ -----------------------------------------------------------
(Updated Aug. 26, 2013, 7:48 p.m.) Review request for mesos, Benjamin Hindman and Vinod Kone. Changes ------- Vinod's review. Repository: mesos-git Description ------- Before slave recovery, the executor driver self-terminated upon disconnection with the slave. With slave recovery, the executor driver waits forever upon disconnection with the slave. This adds a timeout (default of 15 minutes) for the executor to wait before self-terminating. The slave now has a time limit on how long it can stay down before the executor drivers self-terminate. This is configurable via a flag. Note that this timeout is essential for the process isolator (where it's possible for processes to escape), or when the slave never comes back up. Diffs (updated) ----- src/exec/exec.cpp ca61892127cd5f977658bbbf3a67cfa82d12dddf src/launcher/launcher.hpp 637c9bcdfd9c3ee4c071cc46ba8fd274a06873cf src/launcher/launcher.cpp 004d90e4a21aa9c96a115327c98c3a949eee57c2 src/launcher/main.cpp 5674afb7eeded167af97a953d174f4045860a4c8 src/slave/cgroups_isolator.cpp d4ccd114bdcafcaff2e5b12b3881e46daa46f932 src/slave/constants.hpp 901fdf220a902de9241511393530eb19fdfc3244 src/slave/constants.cpp e8d16ca3307249a8b49720eaf8dcb0e7555fca7a src/slave/flags.hpp 616be9b3ecb6fc4165be99da580a1c9876e51d81 src/slave/process_isolator.cpp 24a7fb2be63003d50b24848061cc7be313319eb9 src/tests/slave_recovery_tests.cpp 548e8c09875e4d911ac626b15a2556ff4fd8ff4b Diff: https://reviews.apache.org/r/13791/diff/ Testing ------- Added a test to verify self-termination. Thanks, Ben Mahler
