-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/45520/#review126372
-----------------------------------------------------------



Why /var/run/netns could be in the same mount peer group with its parent? At 
least on fedora21 this is not the case.

Also, why do you fix two bugs in one patch? I know you don't care about bisect, 
but even so this is still not a good practice at all.

- Cong Wang


On March 31, 2016, 1:47 a.m., Jie Yu wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/45520/
> -----------------------------------------------------------
> 
> (Updated March 31, 2016, 1:47 a.m.)
> 
> 
> Review request for mesos, Ian Downes and Cong Wang.
> 
> 
> Bugs: MESOS-4662
>     https://issues.apache.org/jira/browse/MESOS-4662
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> Fixed the bind mount root issue in port mapping isolator. This patch fixed 
> two issues:
> 1) no long assume /var/run/netns is a realpath
> 2) made sure /var/run/netns is a shared mount in its own mount peer group
> 
> 
> Diffs
> -----
> 
>   src/slave/containerizer/mesos/isolators/network/port_mapping.hpp 
> 0fe2f486eb733acf738c1c61fc44f820d7401afc 
>   src/slave/containerizer/mesos/isolators/network/port_mapping.cpp 
> 323c84a3d960a196d8ba87f753814e9d43a07957 
>   src/tests/containerizer/port_mapping_tests.cpp 
> e062daa9fcfc776144b48325daa1f1284c5e59a4 
> 
> Diff: https://reviews.apache.org/r/45520/diff/
> 
> 
> Testing
> -------
> 
> sudo make check on Fedora23
> 
> 
> Thanks,
> 
> Jie Yu
> 
>

Reply via email to