----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/26709/#review56602 -----------------------------------------------------------
src/docker/docker.cpp <https://reviews.apache.org/r/26709/#comment96985> A 30 second timeout seems pretty arbitrary. I would suggest allowing the user to specify this as a arg. Something on the lines of: 1. If a slave system environment value (eg: `DOCKER_STOP_TIMEOUT`)is set we use that. 2. If user provides a `-t` option, instead of using the environment variable. 3. If nothing is set, kill as usual. The only reason why i suggest these is beacuse a lot of people may already have the expectation that killing a task will kill the container. This way, the upgrade process is a smooth upgrade flow. - Ankur Chauhan On Oct. 14, 2014, 9:56 p.m., Ryan Thomas wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/26709/ > ----------------------------------------------------------- > > (Updated Oct. 14, 2014, 9:56 p.m.) > > > Review request for mesos. > > > Bugs: MESOS-1925 > https://issues.apache.org/jira/browse/MESOS-1925 > > > Repository: mesos-git > > > Description > ------- > > Change docker kill to docker stop to allow graceful shutdown of containers. > > > Diffs > ----- > > src/docker/docker.cpp e09b51c4f5101c3a8e77caf12b208c88f47fbcb2 > > Diff: https://reviews.apache.org/r/26709/diff/ > > > Testing > ------- > > > Thanks, > > Ryan Thomas > >
