----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/20429/ -----------------------------------------------------------
Review request for mesos, Ben Mahler, Jie Yu, and Vinod Kone. Repository: mesos-git Description ------- A process that is killed will move out of a cgroup, so cgroups::destroy will see the cgroup as empty and complete, but the process may not have been reaped. This patch ensures that all processes are reaped before the launcher declares a container destroyed. Diffs ----- src/slave/containerizer/cgroups_launcher.hpp db61107141f60dd13cfa13278a7daab16f0d108c src/slave/containerizer/cgroups_launcher.cpp 39f0e4c3baaed3403da160ba63dada4a53d9af09 src/slave/containerizer/launcher.hpp dee526f254d56a7a974a70522cd0ca059814fb6d src/slave/containerizer/launcher.cpp c83327b2f13f0511c8e4e0e9902bc6a1e1328283 src/slave/containerizer/mesos_containerizer.hpp ee1fd3010b4a178a6abcbd813af5eba4d564be00 src/slave/containerizer/mesos_containerizer.cpp 1ce41d71eb13582b46d240d48f0508bdc3e7ef10 Diff: https://reviews.apache.org/r/20429/diff/ Testing ------- make check # Linux Thanks, Ian Downes
