> On Feb. 27, 2016, 2:11 a.m., Alexander Rojas wrote:
> > src/master/allocator/mesos/hierarchical.hpp, line 360
> > <https://reviews.apache.org/r/43882/diff/6/?file=1272157#file1272157line360>
> >
> >     There is a second parameter in the `Timer` object which allows you to 
> > specify a window, once this window is enabled we can get statistics of the 
> > run beyond how long the last allocation took place (note that allocation 
> > varies a lot since sometimes it goes through a few elements and other times 
> > goes throught the whole cluster).
> 
> Benjamin Bannier wrote:
>     Yes, the timer I added here can only be sampled and provides no stats. Do 
> you have a suggestion how many `allocation_intervals` the window should be?

I'm not really sure, I think it would be better if you consult with your 
shepherd on this matter.


- Alexander


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/43882/#review121029
-----------------------------------------------------------


On Feb. 29, 2016, 8:59 p.m., Benjamin Bannier wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/43882/
> -----------------------------------------------------------
> 
> (Updated Feb. 29, 2016, 8:59 p.m.)
> 
> 
> Review request for mesos, Alexander Rukletsov and Ben Mahler.
> 
> 
> Bugs: MESOS-4721
>     https://issues.apache.org/jira/browse/MESOS-4721
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> Added allocation metrics for allocation time.
> 
> 
> Diffs
> -----
> 
>   docs/monitoring.md 827f7073204fcf8575ca980a5571c8be4f5e4110 
>   src/master/allocator/mesos/hierarchical.hpp 
> 3043888630b066505410d3b32c5b3f813cc458c1 
>   src/master/allocator/mesos/hierarchical.cpp 
> 24fa50f62dec10ed43089297473cc386d6ba2f78 
>   src/tests/hierarchical_allocator_tests.cpp 
> 5f771f02db9bd098f3cd36730cd84bf2f5e87a33 
> 
> Diff: https://reviews.apache.org/r/43882/diff/
> 
> 
> Testing
> -------
> 
> make check (OS X)
> 
> I confirmed that this does not lead to general performance regressions in the 
> allocator; this is partially expected since the added code only inserts 
> metrics in the allocator while the actual work is perform asynchronously. 
> These tests where performed with 
> `HierarchicalAllocator_BENCHMARK_Test.DeclineOffers` on an optimized build 
> under OS X using clang(trunk) as compiler.
> 
> 
> Thanks,
> 
> Benjamin Bannier
> 
>

Reply via email to