You are viewing a plain text version of this content. The canonical link for it is here.
Posted to reviews@mesos.apache.org by Benjamin Mahler <bm...@apache.org> on 2019/04/04 23:00:05 UTC

Review Request 70395: Enhanced HierarchicalAllocatorTest.QuotaWithNestedRoleReservation.

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/70395/
-----------------------------------------------------------

Review request for mesos and Meng Zhu.


Bugs: MESOS-9688 and MESOS-9691
    https://issues.apache.org/jira/browse/MESOS-9688
    https://issues.apache.org/jira/browse/MESOS-9691


Repository: mesos


Description
-------

This test missed an issue in https://reviews.apache.org/r/70393/
because it was only checking that the quota consumption for "a"
was >= the subrole "a/b" reservation. Updated this test to ensure
that the quota consumption is exactly equal to the subrole "a/b"
reservation.


Diffs
-----

  src/tests/hierarchical_allocator_tests.cpp c7edee2f49bfac0796c9506265d3bb766416ee63 


Diff: https://reviews.apache.org/r/70395/diff/1/


Testing
-------

make check


Thanks,

Benjamin Mahler


Re: Review Request 70395: Enhanced HierarchicalAllocatorTest.QuotaWithNestedRoleReservation.

Posted by Benjamin Mahler <bm...@apache.org>.

> On April 5, 2019, 1:12 a.m., Meng Zhu wrote:
> > src/tests/hierarchical_allocator_tests.cpp
> > Lines 6129-6132 (original), 6129-6133 (patched)
> > <https://reviews.apache.org/r/70395/diff/1/?file=2137550#file2137550line6129>
> >
> >     I am confused here:
> >     
> >     "Add 1 agent cpus(a/b):10;mem(a/b):1000"
> >     
> >     Let's just mention both agents:
> >     
> >     
> >       // Test:
> >       //   Add agent1 cpus(a/b):10;mem(a/b):100
> >       //   Add agent2 cpus:10;mem:1000
> >       //   Expect "a" to be allocated cpus:1;mem:100 to reach
> >       //     its guarantee (since it already has a consumption of
> >       //     cpus:1;mem:100 from the "a/b" reservation).

The first agent is part of the setup and not part of what's being tested here, but I can make this clearer by:

* Saying that a "second agent is added"
* Marking off the setup vs test aspects

I think that should make it clearer, let me know!


- Benjamin


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/70395/#review214394
-----------------------------------------------------------


On April 4, 2019, 11 p.m., Benjamin Mahler wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/70395/
> -----------------------------------------------------------
> 
> (Updated April 4, 2019, 11 p.m.)
> 
> 
> Review request for mesos and Meng Zhu.
> 
> 
> Bugs: MESOS-9688 and MESOS-9691
>     https://issues.apache.org/jira/browse/MESOS-9688
>     https://issues.apache.org/jira/browse/MESOS-9691
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> This test missed an issue in https://reviews.apache.org/r/70393/
> because it was only checking that the quota consumption for "a"
> was >= the subrole "a/b" reservation. Updated this test to ensure
> that the quota consumption is exactly equal to the subrole "a/b"
> reservation.
> 
> 
> Diffs
> -----
> 
>   src/tests/hierarchical_allocator_tests.cpp c7edee2f49bfac0796c9506265d3bb766416ee63 
> 
> 
> Diff: https://reviews.apache.org/r/70395/diff/1/
> 
> 
> Testing
> -------
> 
> make check
> 
> 
> Thanks,
> 
> Benjamin Mahler
> 
>


Re: Review Request 70395: Enhanced HierarchicalAllocatorTest.QuotaWithNestedRoleReservation.

Posted by Meng Zhu <mz...@mesosphere.io>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/70395/#review214394
-----------------------------------------------------------


Fix it, then Ship it!





src/tests/hierarchical_allocator_tests.cpp
Lines 6129-6132 (original), 6129-6133 (patched)
<https://reviews.apache.org/r/70395/#comment300605>

    I am confused here:
    
    "Add 1 agent cpus(a/b):10;mem(a/b):1000"
    
    Let's just mention both agents:
    
      // Test:
      //   Add agent1 cpus(a/b):10;mem(a/b):100
      //   Add agent2 cpus:10;mem:1000
      //   Expect "a" to be allocated cpus:1;mem:100 to reach
      //     its guarantee (since it already has a consumption of
      //     cpus:1;mem:100 from the "a/b" reservation).


- Meng Zhu


On April 4, 2019, 4 p.m., Benjamin Mahler wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/70395/
> -----------------------------------------------------------
> 
> (Updated April 4, 2019, 4 p.m.)
> 
> 
> Review request for mesos and Meng Zhu.
> 
> 
> Bugs: MESOS-9688 and MESOS-9691
>     https://issues.apache.org/jira/browse/MESOS-9688
>     https://issues.apache.org/jira/browse/MESOS-9691
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> This test missed an issue in https://reviews.apache.org/r/70393/
> because it was only checking that the quota consumption for "a"
> was >= the subrole "a/b" reservation. Updated this test to ensure
> that the quota consumption is exactly equal to the subrole "a/b"
> reservation.
> 
> 
> Diffs
> -----
> 
>   src/tests/hierarchical_allocator_tests.cpp c7edee2f49bfac0796c9506265d3bb766416ee63 
> 
> 
> Diff: https://reviews.apache.org/r/70395/diff/1/
> 
> 
> Testing
> -------
> 
> make check
> 
> 
> Thanks,
> 
> Benjamin Mahler
> 
>


Re: Review Request 70395: Enhanced HierarchicalAllocatorTest.QuotaWithNestedRoleReservation.

Posted by Meng Zhu <mz...@mesosphere.io>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/70395/#review214425
-----------------------------------------------------------


Fix it, then Ship it!





src/tests/hierarchical_allocator_tests.cpp
Line 6130 (original), 6130 (patched)
<https://reviews.apache.org/r/70395/#comment300626>

    I think this should be:
    
    Add a second agent with cpus:10;mem:1000


- Meng Zhu


On April 4, 2019, 7:47 p.m., Benjamin Mahler wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/70395/
> -----------------------------------------------------------
> 
> (Updated April 4, 2019, 7:47 p.m.)
> 
> 
> Review request for mesos and Meng Zhu.
> 
> 
> Bugs: MESOS-9688 and MESOS-9691
>     https://issues.apache.org/jira/browse/MESOS-9688
>     https://issues.apache.org/jira/browse/MESOS-9691
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> This test missed an issue in https://reviews.apache.org/r/70393/
> because it was only checking that the quota consumption for "a"
> was >= the subrole "a/b" reservation. Updated this test to ensure
> that the quota consumption is exactly equal to the subrole "a/b"
> reservation.
> 
> 
> Diffs
> -----
> 
>   src/tests/hierarchical_allocator_tests.cpp c7edee2f49bfac0796c9506265d3bb766416ee63 
> 
> 
> Diff: https://reviews.apache.org/r/70395/diff/2/
> 
> 
> Testing
> -------
> 
> make check
> 
> 
> Thanks,
> 
> Benjamin Mahler
> 
>


Re: Review Request 70395: Enhanced HierarchicalAllocatorTest.QuotaWithNestedRoleReservation.

Posted by Benjamin Mahler <bm...@apache.org>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/70395/
-----------------------------------------------------------

(Updated April 5, 2019, 2:47 a.m.)


Review request for mesos and Meng Zhu.


Changes
-------

Made the comments clearer per meng's feedback.


Bugs: MESOS-9688 and MESOS-9691
    https://issues.apache.org/jira/browse/MESOS-9688
    https://issues.apache.org/jira/browse/MESOS-9691


Repository: mesos


Description
-------

This test missed an issue in https://reviews.apache.org/r/70393/
because it was only checking that the quota consumption for "a"
was >= the subrole "a/b" reservation. Updated this test to ensure
that the quota consumption is exactly equal to the subrole "a/b"
reservation.


Diffs (updated)
-----

  src/tests/hierarchical_allocator_tests.cpp c7edee2f49bfac0796c9506265d3bb766416ee63 


Diff: https://reviews.apache.org/r/70395/diff/2/

Changes: https://reviews.apache.org/r/70395/diff/1-2/


Testing
-------

make check


Thanks,

Benjamin Mahler