> On May 2, 2016, 9:39 a.m., Gilbert Song wrote:
> > src/slave/containerizer/mesos/isolators/filesystem/linux.cpp, lines 490-500
> > <https://reviews.apache.org/r/46807/diff/1/?file=1365199#file1365199line490>
> >
> >     Not yours, but since we may have more volume source type. Could we just 
> > check if do not have host path, then VLOG and skip it?
> >     
> >     (For here, we only handle host_path type, and image type in which case 
> > we mutate the _executorInfo to set host path).

Hum, I prefer not to do so. I think all volumes except for those that have 
'source' should be handled by filesystem isolator.


> On May 2, 2016, 9:39 a.m., Gilbert Song wrote:
> > src/slave/containerizer/mesos/isolators/filesystem/linux.cpp, line 583
> > <https://reviews.apache.org/r/46807/diff/1/?file=1365199#file1365199line583>
> >
> >     But we created the mount point at 'target' if container_path is an 
> > absolute path. I dont understand here.
> >     
> >     This is the only issue. All others LGTM!!

the container_path here is relative due to the if check above.


- Jie


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/46807/#review131278
-----------------------------------------------------------


On April 29, 2016, 10:21 p.m., Jie Yu wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/46807/
> -----------------------------------------------------------
> 
> (Updated April 29, 2016, 10:21 p.m.)
> 
> 
> Review request for mesos, Gilbert Song, Ian Downes, Joseph Wu, and Timothy 
> Chen.
> 
> 
> Bugs: MESOS-5306 and MESOS-5307
>     https://issues.apache.org/jira/browse/MESOS-5306
>     https://issues.apache.org/jira/browse/MESOS-5307
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> This patch simplifies the way we deal with rootfs for command tasks.
> Here are the major simplifications:
> 
> 1) We no longer do a bind mount (<rootfs> -> <sandbox>/.rootfs) for
> command tasks. All isolators can now assume that for both custom
> executors or command tasks, if they change root filesystems, the root
> filesystems will be specified in 'containerConfig.rootfs'. This will
> simplify isolators' logic.
> 
> 2) The sandbox bind mount will be done in the container's mount
> namespace. Currently, it's done in the host mount namespace. This
> creates a lot of complications, especially during the cleanup and
> recovery path.
> 
> 3) All persistent volumes will be consistently mounted under the
> container's sandbox directory. Currently, depending on whether the
> container changes the root filesystem or not, the persistent volumes
> will be mounted at different locations. This simplifies the cleanup and
> recovery path as well.
> 
> 
> Diffs
> -----
> 
>   docs/container-image.md 843881e9c3d0c434fedff44f4125154dee6d52c6 
>   src/launcher/executor.cpp 9f1d2168bc4ddbce1bcd25ff38dc1c34714eb28b 
>   src/launcher/http_command_executor.cpp 
> 0b4136c040ec9cf585c0d38f8471495a855cd640 
>   src/slave/constants.hpp c24167f273bf57d360895e9167b4249e560c3c59 
>   src/slave/containerizer/mesos/containerizer.hpp 
> 13399f014dcd85defbff79f3b5aa4e7e75d41fd1 
>   src/slave/containerizer/mesos/containerizer.cpp 
> 8d538954d6e1f13e833d75c2eaa37e700278ee0c 
>   src/slave/containerizer/mesos/isolators/filesystem/linux.hpp 
> cb181265482c884b02bdfc576f906aa0dd9f00f2 
>   src/slave/containerizer/mesos/isolators/filesystem/linux.cpp 
> a6e748ae777b6280f5dab0a6fec7305f931f2c22 
>   src/slave/slave.cpp 5aa1530a3f6b2fdb7196fc65b133ab6afaaa6a53 
> 
> Diff: https://reviews.apache.org/r/46807/diff/
> 
> 
> Testing
> -------
> 
> sudo make check
> 
> 
> Thanks,
> 
> Jie Yu
> 
>

Reply via email to