> On Sept. 29, 2015, 9:04 p.m., Ian Downes wrote:
> > src/slave/containerizer/isolators/filesystem/linux.cpp, line 121
> > <https://reviews.apache.org/r/38858/diff/2/?file=1087334#file1087334line121>
> >
> >     Suggestion, refactor to avoid the nesting?
> >     
> >     if (workDirMount.isNone() {
> >       // mount it and get the fs::MountInfoTable::Entry
> >     }
> >     
> >     // Check it's a slave mount.
> >     
> >     // Check it's a shared mount.

This is tricky because we need to deal with different system defaults. FOr 
instance on fedora-22, all mounts are shared mounts by default. On Ubuntu14.04, 
all mounts are private mounts by default.

Also, the command `mount --make-shared` is not idempotent. If the mount is 
already shared, calling this command again will create a new peer group!


- Jie


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/38858/#review101011
-----------------------------------------------------------


On Sept. 29, 2015, 7:48 p.m., Jie Yu wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/38858/
> -----------------------------------------------------------
> 
> (Updated Sept. 29, 2015, 7:48 p.m.)
> 
> 
> Review request for mesos, Timothy Chen and Jiang Yan Xu.
> 
> 
> Bugs: MESOS-3539
>     https://issues.apache.org/jira/browse/MESOS-3539
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> Ensured that slave's work_dir is a shared mount in its own peer group when 
> LinuxFilesystemIsolator is used.
> 
> This addressed the TODO in the current code.
> 
> 
> Diffs
> -----
> 
>   src/slave/containerizer/isolators/filesystem/linux.cpp 
> 8fa929f12838044762be82287e1b71440cccc6d9 
>   src/tests/containerizer/filesystem_isolator_tests.cpp 
> 41d098c7d66eba0befc3708c6660a6a3e5f0d2de 
> 
> Diff: https://reviews.apache.org/r/38858/diff/
> 
> 
> Testing
> -------
> 
> sudo make check
> 
> 
> Thanks,
> 
> Jie Yu
> 
>

Reply via email to