> On Feb. 28, 2015, 1:59 a.m., Dominic Hamon wrote:
> > src/slave/containerizer/isolators/cgroups/mem.hpp, line 22
> > <https://reviews.apache.org/r/30546/diff/2/?file=881463#file881463line22>
> >
> > why is this header treated like something installed while the other
> > slave headers are treated as local dev headers?
> >
> > consistency rules here.. this should be lower down and
> > "slave/isolator.hpp" i believe.
introduced from the following comment.
commit 0082eb6d1a4cb87b2fb3672524a02d04e02854f9
Author: Kapil Arya <ka...@mesosphere.io>
Date: Tue Feb 10 09:54:37 2015 -0800
Out of tree build 5: Exposed slave/containerizer/isolator.hpp as
mesos/slave/isolator.hpp
Expose "slave/state.hpp" and "slave/containerizer/isolator.hpp" to
allow modules to include them directly from the Mesos install
location.
Review: https://reviews.apache.org/r/29603
> On Feb. 28, 2015, 1:59 a.m., Dominic Hamon wrote:
> > src/slave/containerizer/isolators/cgroups/mem.cpp, line 432
> > <https://reviews.apache.org/r/30546/diff/2/?file=881464#file881464line432>
> >
> > instead of a list, why not a vector where it is indexed by 'LOW',
> > 'MEDIUM' and 'CRITICAL'?
> >
> > then in the continuation you don't have to assume the ordering in the
> > list.
list doesnt allow indexing, unfortunately. changed to use tuple.
> On Feb. 28, 2015, 1:59 a.m., Dominic Hamon wrote:
> > src/slave/containerizer/isolators/cgroups/mem.cpp, line 436
> > <https://reviews.apache.org/r/30546/diff/2/?file=881464#file881464line436>
> >
> > a separate failure for each listener might help debugging down the road.
fixed.
> On Feb. 28, 2015, 1:59 a.m., Dominic Hamon wrote:
> > src/slave/containerizer/isolators/cgroups/mem.cpp, line 442
> > <https://reviews.apache.org/r/30546/diff/2/?file=881464#file881464line442>
> >
> > await or collect?
await allows faiure, which works better with my case.
> On Feb. 28, 2015, 1:59 a.m., Dominic Hamon wrote:
> > src/slave/containerizer/isolators/cgroups/mem.cpp, line 444
> > <https://reviews.apache.org/r/30546/diff/2/?file=881464#file881464line444>
> >
> > the general rule is this would be _pressureListen
changed to _usage
> On Feb. 28, 2015, 1:59 a.m., Dominic Hamon wrote:
> > src/slave/containerizer/isolators/cgroups/mem.cpp, line 668
> > <https://reviews.apache.org/r/30546/diff/2/?file=881464#file881464line668>
> >
> > cbegin is only in c++11 .. is it definitely in g++-4.6?
avoided using iterator anymore.
> On Feb. 28, 2015, 1:59 a.m., Dominic Hamon wrote:
> > src/slave/containerizer/isolators/cgroups/mem.cpp, line 672
> > <https://reviews.apache.org/r/30546/diff/2/?file=881464#file881464line672>
> >
> > iterator->isReady() is clearer
fixed.
> On Feb. 28, 2015, 1:59 a.m., Dominic Hamon wrote:
> > src/tests/slave_recovery_tests.cpp, line 3571
> > <https://reviews.apache.org/r/30546/diff/2/?file=881465#file881465line3571>
> >
> > s/> >/>>/
fixed.
- Chi
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/30546/#review74647
-----------------------------------------------------------
On Feb. 28, 2015, 1:43 a.m., Chi Zhang wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/30546/
> -----------------------------------------------------------
>
> (Updated Feb. 28, 2015, 1:43 a.m.)
>
>
> Review request for mesos, Dominic Hamon, Ian Downes, and Jie Yu.
>
>
> Bugs: MESOS-2136
> https://issues.apache.org/jira/browse/MESOS-2136
>
>
> Repository: mesos
>
>
> Description
> -------
>
> MemIsolator: expose memory pressures for containers.
>
>
> Diffs
> -----
>
> include/mesos/mesos.proto 9df972d750ce1e4a81d2e96cc508d6f83cad2fc8
> src/slave/containerizer/isolators/cgroups/mem.hpp
> a00f723de61b9bccd76a2948b6d14fde7a993a8d
> src/slave/containerizer/isolators/cgroups/mem.cpp
> 6299ca4ba01b65daa3d75c64150e2738e32b841e
> src/tests/slave_recovery_tests.cpp 53adae0118a26e6d25a9ff20c6374cc8e73275b1
>
> Diff: https://reviews.apache.org/r/30546/diff/
>
>
> Testing
> -------
>
>
> Thanks,
>
> Chi Zhang
>
>
