> On July 18, 2019, 3:56 a.m., Andrei Sekretenko wrote:
> > src/master/quota_handler.cpp
> > Lines 664 (patched)
> > <https://reviews.apache.org/r/71111/diff/1/?file=2155880#file2155880line664>
> >
> >     Does this mean that we might rescind an offer even if `consumption + 
> > offered` for the offer's role is within guarantees of that role? 
> >     
> >     If yes - do we really have to be THAT greedy?...
> 
> Andrei Sekretenko wrote:
>     Edit: s/is within guarantees/is going to be within guarantees/

Yes, it is inaccurate. But calculating a more correct amount would be 
expensive. And I do not feel it is necessary. See more details in my message in 
the public slack: https://mesos.slack.com/archives/C1NEKP60K/p1563498805199200


- Meng


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/71111/#review216717
-----------------------------------------------------------


On July 17, 2019, 9:30 p.m., Meng Zhu wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/71111/
> -----------------------------------------------------------
> 
> (Updated July 17, 2019, 9:30 p.m.)
> 
> 
> Review request for mesos, Andrei Sekretenko and Benjamin Mahler.
> 
> 
> Bugs: MESOS-9812
>     https://issues.apache.org/jira/browse/MESOS-9812
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> Outstanding offers need to be rescinded as needed to ensure
> roles' requested guarantees can be satisfied.
> 
> For simplicity, and also due to the race between the master and
> the allocator, we pessimistically assume that what seems like
> "available" resources in the allocator are all gone. We greedily
> rescind offers until we can satisfy the guarantees.
> 
> Also added a test.
> 
> 
> Diffs
> -----
> 
>   src/master/quota_handler.cpp 2a3ca5640c850ec8a8aca7abed45386b5b5e1125 
>   src/tests/master_quota_tests.cpp 64a9b31a38af23fb16a4afd04f6d3413c7d1ffe9 
> 
> 
> Diff: https://reviews.apache.org/r/71111/diff/2/
> 
> 
> Testing
> -------
> 
> make check
> 
> 
> Thanks,
> 
> Meng Zhu
> 
>

Reply via email to