----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/24128/#review49400 -----------------------------------------------------------
We probably wanna go with this patch for 0.20.0. See MESOS-1466 for details. However, for 0.20.0, besides this patch, I'll kill private_resources and add ephemeral ports to 'resources', but not sending it to the master. src/master/flags.hpp <https://reviews.apache.org/r/24128/#comment86460> s/int/size_t/ src/master/master.cpp <https://reviews.apache.org/r/24128/#comment86461> Link the ticket: MESOS-1654 in the TODO. Also, would you please explain briefly why we need this flag with the existing code. src/master/master.cpp <https://reviews.apache.org/r/24128/#comment86462> Could you please add an optional filer for 'resourcesRecovered' (similar to that in 'resourcesUnused'), and use a default filter here. - Jie Yu On Aug. 1, 2014, 7:37 p.m., Dominic Hamon wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/24128/ > ----------------------------------------------------------- > > (Updated Aug. 1, 2014, 7:37 p.m.) > > > Review request for mesos, Benjamin Hindman and Jie Yu. > > > Bugs: MESOS-1654 > https://issues.apache.org/jira/browse/MESOS-1654 > > > Repository: mesos-git > > > Description > ------- > > when network isolation is enabled, we should keep track of how many executors > have launched on a given host to avoid sending offers for those that have hit > the container limit. > > this is a short-term fix for the issue to enable broader testing of network > monitoring and isolation. > > > Diffs > ----- > > src/master/flags.hpp 2b92de37b64c3363915918c65ab092efd5b378d6 > src/master/master.cpp 77f2536c36df9a0a1f6021bd4732baa3ab424459 > src/tests/master_tests.cpp 5a1cf7f1f7a4864f516825cc14504c870b60055e > > Diff: https://reviews.apache.org/r/24128/diff/ > > > Testing > ------- > > make check > > > Thanks, > > Dominic Hamon > >
