mesos-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Michael Park <mp...@apache.org>
Subject Re: Review Request 59383: Added benchmark for allocator perf with many suppressed frameworks.
Date Wed, 24 May 2017 06:36:49 GMT

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/59383/#review175903
-----------------------------------------------------------


Fix it, then Ship it!





src/tests/hierarchical_allocator_tests.cpp
Lines 5500-5501 (patched)
<https://reviews.apache.org/r/59383/#comment249250>

    Is this intentional?
    
    It seems like it is. As in, `used` is not supposed to be accumulated.
    
    Is so, can we just do
    
    ```
    hashmap<FrameworkID, Resources> used {
      {frameworks[i % frameworkCount].id(), allocation}
    };
    ```
    
    so that the intentions is clearer?
    or even just pass it inline.



src/tests/hierarchical_allocator_tests.cpp
Lines 5537 (patched)
<https://reviews.apache.org/r/59383/#comment249251>

    Can we do `i++` here just to stay consistent with the one only 4 lines above? It seems
the rest of this file uses `i++` too.


- Michael Park


On May 22, 2017, 11:33 p.m., Neil Conway wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/59383/
> -----------------------------------------------------------
> 
> (Updated May 22, 2017, 11:33 p.m.)
> 
> 
> Review request for mesos, Benjamin Mahler, James Peach, Michael Park, and Jiang Yan Xu.
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> This covers the case where the vast majority (99%) of frameworks have
> suppressed offers.
> 
> 
> Diffs
> -----
> 
>   src/tests/hierarchical_allocator_tests.cpp 6dee2296d5a14185dbf7eee17968b20148839bfd

> 
> 
> Diff: https://reviews.apache.org/r/59383/diff/1/
> 
> 
> Testing
> -------
> 
> 
> Thanks,
> 
> Neil Conway
> 
>


Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message