-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/23520/#review47910
-----------------------------------------------------------


Looks good - I added a couple of suggestions/questions and wonder how you 
tested it :-)


src/launcher/executor.cpp
<https://reviews.apache.org/r/23520/#comment84136>

    The health check program may initiateTaskKill without the task status being 
unhealthy. How about setting this variable to 'healthy' unconditionally?



src/launcher/executor.cpp
<https://reviews.apache.org/r/23520/#comment84137>

    unhealthy could be an option type, so you can encode both "killed healthy" 
and "killed unhealthy"



src/launcher/executor.cpp
<https://reviews.apache.org/r/23520/#comment84139>

    Maybe make this an option type?



src/tests/health_check_tests.cpp
<https://reviews.apache.org/r/23520/#comment84135>

    How did you test this if the test was disabled?


- Niklas Nielsen


On July 15, 2014, 1:30 p.m., Timothy Chen wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/23520/
> -----------------------------------------------------------
> 
> (Updated July 15, 2014, 1:30 p.m.)
> 
> 
> Review request for mesos, Benjamin Hindman and Niklas Nielsen.
> 
> 
> Repository: mesos-git
> 
> 
> Description
> -------
> 
> Currently after the health check receives configured amount of task fails it 
> will initiate a kill task.
> However currently the task kill status update doesn't set the healthy field 
> so it's unclear was it killed because health check or other reasons.
> This patch sets the health check field when the task is killed because of 
> failing health checks.
> 
> 
> Diffs
> -----
> 
>   src/launcher/executor.cpp a573637 
>   src/tests/health_check_tests.cpp 44711fd 
> 
> Diff: https://reviews.apache.org/r/23520/diff/
> 
> 
> Testing
> -------
> 
> make check
> 
> 
> Thanks,
> 
> Timothy Chen
> 
>

Reply via email to