You are viewing a plain text version of this content. The canonical link for it is here.
Posted to reviews@mesos.apache.org by Benjamin Hindman <be...@berkeley.edu> on 2015/04/23 02:33:17 UTC
Review Request 33459: Fix containerizers to only recover containers
they would launch.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/33459/
-----------------------------------------------------------
Review request for mesos, Joris Van Remoortere, Michael Park, and Timothy Chen.
Bugs: MESOS-2605
https://issues.apache.org/jira/browse/MESOS-2605
Repository: mesos
Description
-------
It's possible that a containerizer might recover a container that it
can't properly handle, and later when any updates are attempted on
that container the slave will actually destroy it which could leave an
orphaned container because the containerizer doesn't actually know how
to destroy it. See MESOS-2605.
Diffs
-----
src/slave/containerizer/docker.cpp f9fb07806e3b7d7d2afc1be3b8756eac23b32dcd
src/slave/containerizer/mesos/containerizer.cpp 0bd26ea58ef56bad3b215711a8a38ca65ecd8ebd
Diff: https://reviews.apache.org/r/33459/diff/
Testing
-------
Thanks,
Benjamin Hindman
Re: Review Request 33459: Fix containerizers to only recover
containers they would launch.
Posted by Jie Yu <yu...@gmail.com>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/33459/#review81251
-----------------------------------------------------------
Subsumed by this https://reviews.apache.org/r/33257?
- Jie Yu
On April 23, 2015, 12:33 a.m., Benjamin Hindman wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/33459/
> -----------------------------------------------------------
>
> (Updated April 23, 2015, 12:33 a.m.)
>
>
> Review request for mesos, Joris Van Remoortere, Michael Park, and Timothy Chen.
>
>
> Bugs: MESOS-2605
> https://issues.apache.org/jira/browse/MESOS-2605
>
>
> Repository: mesos
>
>
> Description
> -------
>
> It's possible that a containerizer might recover a container that it
> can't properly handle, and later when any updates are attempted on
> that container the slave will actually destroy it which could leave an
> orphaned container because the containerizer doesn't actually know how
> to destroy it. See MESOS-2605.
>
>
> Diffs
> -----
>
> src/slave/containerizer/docker.cpp f9fb07806e3b7d7d2afc1be3b8756eac23b32dcd
> src/slave/containerizer/mesos/containerizer.cpp 0bd26ea58ef56bad3b215711a8a38ca65ecd8ebd
>
> Diff: https://reviews.apache.org/r/33459/diff/
>
>
> Testing
> -------
>
>
> Thanks,
>
> Benjamin Hindman
>
>
Re: Review Request 33459: Fix containerizers to only recover
containers they would launch.
Posted by Benjamin Hindman <be...@berkeley.edu>.
> On April 23, 2015, 12:37 a.m., Timothy Chen wrote:
> > I already have a fix out for this:
> > https://reviews.apache.org/r/33257/
> >
> > My patch also handles docker container tasks, since the docker containerizer uses the command executor in docker tasks, and without more handling we can't tell if it's from mesos or docker.
Cool! Well in that case I'll discard this review. Please update MESOS-2605 and MESOS-2601 with this proposed review.
- Benjamin
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/33459/#review81252
-----------------------------------------------------------
On April 23, 2015, 12:33 a.m., Benjamin Hindman wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/33459/
> -----------------------------------------------------------
>
> (Updated April 23, 2015, 12:33 a.m.)
>
>
> Review request for mesos, Joris Van Remoortere, Michael Park, and Timothy Chen.
>
>
> Bugs: MESOS-2605
> https://issues.apache.org/jira/browse/MESOS-2605
>
>
> Repository: mesos
>
>
> Description
> -------
>
> It's possible that a containerizer might recover a container that it
> can't properly handle, and later when any updates are attempted on
> that container the slave will actually destroy it which could leave an
> orphaned container because the containerizer doesn't actually know how
> to destroy it. See MESOS-2605.
>
>
> Diffs
> -----
>
> src/slave/containerizer/docker.cpp f9fb07806e3b7d7d2afc1be3b8756eac23b32dcd
> src/slave/containerizer/mesos/containerizer.cpp 0bd26ea58ef56bad3b215711a8a38ca65ecd8ebd
>
> Diff: https://reviews.apache.org/r/33459/diff/
>
>
> Testing
> -------
>
>
> Thanks,
>
> Benjamin Hindman
>
>
Re: Review Request 33459: Fix containerizers to only recover
containers they would launch.
Posted by Timothy Chen <tn...@apache.org>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/33459/#review81252
-----------------------------------------------------------
I already have a fix out for this:
https://reviews.apache.org/r/33257/
My patch also handles docker container tasks, since the docker containerizer uses the command executor in docker tasks, and without more handling we can't tell if it's from mesos or docker.
- Timothy Chen
On April 23, 2015, 12:33 a.m., Benjamin Hindman wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/33459/
> -----------------------------------------------------------
>
> (Updated April 23, 2015, 12:33 a.m.)
>
>
> Review request for mesos, Joris Van Remoortere, Michael Park, and Timothy Chen.
>
>
> Bugs: MESOS-2605
> https://issues.apache.org/jira/browse/MESOS-2605
>
>
> Repository: mesos
>
>
> Description
> -------
>
> It's possible that a containerizer might recover a container that it
> can't properly handle, and later when any updates are attempted on
> that container the slave will actually destroy it which could leave an
> orphaned container because the containerizer doesn't actually know how
> to destroy it. See MESOS-2605.
>
>
> Diffs
> -----
>
> src/slave/containerizer/docker.cpp f9fb07806e3b7d7d2afc1be3b8756eac23b32dcd
> src/slave/containerizer/mesos/containerizer.cpp 0bd26ea58ef56bad3b215711a8a38ca65ecd8ebd
>
> Diff: https://reviews.apache.org/r/33459/diff/
>
>
> Testing
> -------
>
>
> Thanks,
>
> Benjamin Hindman
>
>