mesos-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Vinod Kone <vinodk...@gmail.com>
Subject Re: Review Request 50422: Future-proofed some slave removal tests.
Date Wed, 27 Jul 2016 23:54:21 GMT

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/50422/#review143855
-----------------------------------------------------------




src/tests/slave_tests.cpp (line 2381)
<https://reviews.apache.org/r/50422/#comment209793>

    s/only/removed only/ ?



src/tests/slave_tests.cpp (line 2425)
<https://reviews.apache.org/r/50422/#comment209794>

    Are you sure pausing the clock at the beginning is safe? IIRC, the scheduler driver does
reliable registration which relies on timers and with esp auth enabled there is no guarantee
that auth and registration go through in the first try?
    
    Also, please run these tests in a loop if you haven't already.



src/tests/slave_tests.cpp (line 2521)
<https://reviews.apache.org/r/50422/#comment209796>

    ditto. see above comment.



src/tests/slave_tests.cpp (line 2536)
<https://reviews.apache.org/r/50422/#comment209797>

    should've been done in the previous review?



src/tests/slave_tests.cpp (line 2564)
<https://reviews.apache.org/r/50422/#comment209798>

    not necessary because this is automatically handled. you didn't do it in the above changed
test as well?


- Vinod Kone


On July 25, 2016, 10:22 p.m., Neil Conway wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/50422/
> -----------------------------------------------------------
> 
> (Updated July 25, 2016, 10:22 p.m.)
> 
> 
> Review request for mesos and Vinod Kone.
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> These tests relied on the implementation detail that when an agent is
> removed from the list of registered agents, the master sends a
> ShutdownSlaveMessage to the agent. That will change in the future
> (MESOS-4049). To prepare for this future planned behavior, adjust these
> tests to be more robust by instead looking for the invocation of the
> `slaveLost` scheduler callback.
> 
> 
> Diffs
> -----
> 
>   src/tests/slave_tests.cpp b9fa85dc1ae0922a100786fcb01156b90a013d2a 
> 
> Diff: https://reviews.apache.org/r/50422/diff/
> 
> 
> Testing
> -------
> 
> make check
> 
> Validated that when the "cancel pending slave removal on receipt of ping" code is removed,
`CancelSlaveRemoval` still fails.
> 
> 
> Thanks,
> 
> Neil Conway
> 
>


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