> On April 22, 2016, 9:49 p.m., Jie Yu wrote:
> > src/slave/containerizer/mesos/isolators/docker/volume/isolator.hpp, line 109
> > <https://reviews.apache.org/r/44454/diff/5/?file=1357489#file1357489line109>
> >
> >     Please use a Sequence for each DockerVolume. umount A and mount B can 
> > happen in parallel as the ordering does not matter.
> 
> Guangya Liu wrote:
>     One question is if define `sequence` for each volume, then can we 
> guarantee that two different volumes can always keep the order?
>     
>     e.g.
>     Volume 1) driver: convoy, name: dvd1
>     Volume 2) driver: flocker, name: dvd2
>     
>     One container using the above two different volumes, can we guarantee 
> volume 1 mount point return first than volume 2 if letting differene sequence 
> manage those two volumes?
> 
> Jie Yu wrote:
>     why does that matter?
> 
> Guangya Liu wrote:
>     For such case:
>     
>     Volume 1) driver: convoy, name: dvd1
>     Volume 2) driver: flocker, name: dvd2
>     
>     Volume 1 map to /tmp/abc1
>     Volume 2 map to /tmp/abc2
>     
>     If the return value of volume 1 and 2 mount point is not right, then 
> volume 1 mount point will mount to /tmp/abc1 and volume 2 will mount to 
> /tmp/abc2
>     
>     I'm just not sure if using different volume sequence can guarantee the 
> order?
> 
> Guangya Liu wrote:
>     Correct a typo: If the return value of volume 1 and 2 mount point is not 
> right, then volume 1 mount point will mount to `/tmp/abc2` and volume 2 will 
> mount to `/tmp/abc1`

I don't think that's an issue. I don't think a driver will return an incorrect 
mount point for a volume (otherwise, it's a bug).


- Jie


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/44454/#review130192
-----------------------------------------------------------


On April 22, 2016, 4:11 p.m., Guangya Liu wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/44454/
> -----------------------------------------------------------
> 
> (Updated April 22, 2016, 4:11 p.m.)
> 
> 
> Review request for mesos, David vonThenen, Gilbert Song, haosdent huang, and 
> Jie Yu.
> 
> 
> Bugs: MESOS-5013
>     https://issues.apache.org/jira/browse/MESOS-5013
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> Enabled Sequence mount for prepare().
> 
> 
> Diffs
> -----
> 
>   src/slave/containerizer/mesos/isolators/docker/volume/isolator.hpp 
> bedc687cc280d0b721fb84801039fd3614364cca 
>   src/slave/containerizer/mesos/isolators/docker/volume/isolator.cpp 
> 915e5ae755a55a02b7dfcda88165f27346cad955 
> 
> Diff: https://reviews.apache.org/r/44454/diff/
> 
> 
> Testing
> -------
> 
> make
> make check
> 
> 
> Thanks,
> 
> Guangya Liu
> 
>

Reply via email to