-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/24316/#review50039
-----------------------------------------------------------


Looks good! Can you do me one favor, can you test it on a machine with memsw 
turned on and run out tests with --cgroups_limit_swap set to true?


src/slave/containerizer/isolators/cgroups/mem.hpp
<https://reviews.apache.org/r/24316/#comment87550>

    We don't use const ref for primitive types
    
    s/const bool&/bool/



src/slave/containerizer/isolators/cgroups/mem.cpp
<https://reviews.apache.org/r/24316/#comment87551>

    indent 2 spaces please.


- Jie Yu


On Aug. 8, 2014, 10:42 a.m., Anton Lindström wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/24316/
> -----------------------------------------------------------
> 
> (Updated Aug. 8, 2014, 10:42 a.m.)
> 
> 
> Review request for mesos.
> 
> 
> Bugs: MESOS-1662
>     https://issues.apache.org/jira/browse/MESOS-1662
> 
> 
> Repository: mesos-git
> 
> 
> Description
> -------
> 
> Changed memory.limit_in_bytes to memory.memsw.limit_in_bytes where memory is 
> limited. This makes Mesos limit both memory and swap as described in the Jira 
> ticket.
> 
> Using --limit-swap, it's possible to use memory.memsw.limit_in_bytes. If it 
> doesn't exist it will fallback on memory.limit_in_bytes.
> 
> 
> Diffs
> -----
> 
>   src/linux/cgroups.hpp 9dfba6eb9059728386d053f8c2e92edab0caa295 
>   src/linux/cgroups.cpp 39a4874ee0030110d7a61f1b73653af116d66549 
>   src/slave/containerizer/isolators/cgroups/mem.hpp 
> 6869ed4a9fa9a1a719e695ac83aeec88ead0efc0 
>   src/slave/containerizer/isolators/cgroups/mem.cpp 
> e8160efc4f5baa06470a8a3c774bf97d5b9e83cc 
>   src/slave/flags.hpp 841de2364085c66e4272734d5b11dbf8efa845a5 
> 
> Diff: https://reviews.apache.org/r/24316/diff/
> 
> 
> Testing
> -------
> 
> make check
> 
> 
> Thanks,
> 
> Anton Lindström
> 
>

Reply via email to