> On July 3, 2014, 12:18 a.m., Chi Zhang wrote: > > src/slave/containerizer/isolators/network/port_mapping.cpp, line 1071 > > <https://reviews.apache.org/r/23246/diff/2/?file=623104#file623104line1071> > > > > is this sufficient to check BIND_MOUNT_ROOT is _self_ mounted?
a little update: - only mount A and B are copied into C at the time of clone. C is touched and mounted in isolate, so in child isolation script context, C is a file but not a mount point, while is fine. we umount A and B and bind mount an empty directory to ContainerC/BIND_MOUNT_ROOT to cover up as before. - Chi ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/23246/#review47256 ----------------------------------------------------------- On July 2, 2014, 9:59 p.m., Jie Yu wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/23246/ > ----------------------------------------------------------- > > (Updated July 2, 2014, 9:59 p.m.) > > > Review request for mesos, Chi Zhang, Ian Downes, and Vinod Kone. > > > Bugs: MESOS-1558 > https://issues.apache.org/jira/browse/MESOS-1558 > > > Repository: mesos-git > > > Description > ------- > > See summary. The idea is to use --make-rshared for /var/run/netns on host and > --make-rslave for /var/run/netns in each container. > > > Diffs > ----- > > src/slave/containerizer/isolators/network/port_mapping.cpp a326653 > > Diff: https://reviews.apache.org/r/23246/diff/ > > > Testing > ------- > > make check > sudo make check > > The regression tests is in another review by Chi Zhang: > https://reviews.apache.org/r/23221/ > > > Thanks, > > Jie Yu > >