> On June 5, 2015, 2:58 p.m., Vinod Kone wrote:
> > src/master/master.cpp, line 3517
> > <https://reviews.apache.org/r/35118/diff/1/?file=980131#file980131line3517>
> >
> >     while you are at it, can you just send slave->totalResources here 
> > instead of oversubscribedResources? this will address my TODO on this 
> > method.

The allocator expects only oversubscribedResources and does a CHECK on it. The 
logic in Allocator::updateSlave(...) needs a little tweaking as well. I'd be 
happy to do it in a subsequent review.

SG?


- Jiang Yan


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/35118/#review86850
-----------------------------------------------------------


On June 5, 2015, 2:09 p.m., Jiang Yan Xu wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/35118/
> -----------------------------------------------------------
> 
> (Updated June 5, 2015, 2:09 p.m.)
> 
> 
> Review request for mesos and Vinod Kone.
> 
> 
> Bugs: MESOS-2776
>     https://issues.apache.org/jira/browse/MESOS-2776
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> - This way Master::Slave::totalResources includes revocable resources, which 
> we need for metrics for revocable resources.
> - Changed updateSlave() argument to use `const Resources& 
> oversubscribedResources` instead of `const std::vector<Resource>& 
> oversubscribedResources` because `Resources` provides convenience methods 
> such as `revocable()`.
> 
> 
> Diffs
> -----
> 
>   src/master/master.hpp deeb0d8c87a13315206556e1d0974cdd13e8224f 
>   src/master/master.cpp be0db42da3c59761aa154439653d715556465256 
> 
> Diff: https://reviews.apache.org/r/35118/diff/
> 
> 
> Testing
> -------
> 
> make check.
> 
> 
> Thanks,
> 
> Jiang Yan Xu
> 
>

Reply via email to