> On May 5, 2014, 4:18 p.m., Vinod Kone wrote:
> > src/linux/cgroups.cpp, lines 1713-1715
> > <https://reviews.apache.org/r/20818/diff/4/?file=572874#file572874line1713>
> >
> >     How come we are not doing retires anymore?

Retries are done for freeze and thaw. See additional comments below in response 
to Jie's question about the SIGSTOP logic.


> On May 5, 2014, 4:18 p.m., Vinod Kone wrote:
> > src/linux/cgroups.cpp, lines 1636-1638
> > <https://reviews.apache.org/r/20818/diff/4/?file=572874#file572874line1636>
> >
> >     IIRC, we added SIGSTOP and SIGKILL before freeze because the earlier 
> > freeze, kill, thaw sequence didn't always work. Have the semantics changed 
> > so that this is guaranteed to work?

See response to Jie's comment, below.


- Ian


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/20818/#review42216
-----------------------------------------------------------


On May 5, 2014, 1:16 p.m., Ian Downes wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/20818/
> -----------------------------------------------------------
> 
> (Updated May 5, 2014, 1:16 p.m.)
> 
> 
> Review request for mesos, Benjamin Hindman, Ben Mahler, Chi Zhang, Jie Yu, 
> and Vinod Kone.
> 
> 
> Bugs: MESOS-759 and MESOS-976
>     https://issues.apache.org/jira/browse/MESOS-759
>     https://issues.apache.org/jira/browse/MESOS-976
> 
> 
> Repository: mesos-git
> 
> 
> Description
> -------
> 
>     internal::Destroyer now uses a single pass of freeze, kill, thaw, reap
>     to kill all processes in a freezer cgroup. cgroups::destroy will not
>     complete until all processes have been reaped. A timeout is used for
>     each step so destroy will eventually fail rather than constantly
>     retrying.
> 
>     Added a test for destroying a freezer cgroup containing a stopped
>     process.
> 
> 
> Diffs
> -----
> 
>   src/linux/cgroups.hpp 5a5735721fb9f051eee661edb08d1cdaa163d0f3 
>   src/linux/cgroups.cpp 8202c282f580d027a60ded2081962e96e4860f60 
>   src/slave/containerizer/isolators/cgroups/cpushare.cpp 
> b494a9236210245383e20fa9ab3dbac01e42f8dd 
>   src/slave/containerizer/linux_launcher.cpp 
> 530e0bd64d71bad761a2eab3d6e2f2179a167b4b 
>   src/tests/cgroups_tests.cpp 6ba9de622953e158feadaa9950618b0b13c9e832 
> 
> Diff: https://reviews.apache.org/r/20818/diff/
> 
> 
> Testing
> -------
> 
> make check # Linux
> 
> 
> Thanks,
> 
> Ian Downes
> 
>

Reply via email to