-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/30952/#review72634
-----------------------------------------------------------


Looks nice! Do you want to remove these logics in the drivers (old and new)?


src/master/master.cpp
<https://reviews.apache.org/r/30952/#comment118679>

    Hum, why do you want to move the 'accept' invokation to the bottom? I would 
suggest keep this function the same as before.



src/master/master.cpp
<https://reviews.apache.org/r/30952/#comment118680>

    We usually avoid using non-const references because that makes the 
reasoning not local. See my comments below, this is not needed.



src/master/master.cpp
<https://reviews.apache.org/r/30952/#comment118681>

    I would suggest move this logic to `_accept`. See my comments below.



src/master/master.cpp
<https://reviews.apache.org/r/30952/#comment118682>

    You probably want to move the framework ID filling logic here right before 
the task validation. You need to make a copy of `task`:
    
    ```
    // Make a copy of the original task so that we can
    // fill the missing `framework_id` in ExecutorInfo
    // if needed. This field was added to the API later
    // and thus was made optional.
    TaskInfo _task = task;
    if (task.has_executor() && !task.executor().has_framework_id()) {
      _task.mutable_executor()->mutable_framework_id()->CopyFrom(framework->id);
    }
    
    // Validate the task.
    const Option<Error>& validationError = validation::task::validate(
        _task,
        framework,
        slave,
        _offeredResources);
    ...
    ```



src/master/validation.cpp
<https://reviews.apache.org/r/30952/#comment118683>

    Do you want to add a test, or there is a test already?


- Jie Yu


On Feb. 16, 2015, 8:53 a.m., Isabel Jimenez wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/30952/
> -----------------------------------------------------------
> 
> (Updated Feb. 16, 2015, 8:53 a.m.)
> 
> 
> Review request for mesos, Jie Yu, Joris Van Remoortere, and Vinod Kone.
> 
> 
> Bugs: MESOS-2290
>     https://issues.apache.org/jira/browse/MESOS-2290
> 
> 
> Repository: mesos-incubating
> 
> 
> Description
> -------
> 
> There is a scheduler validation 
> (https://github.com/apache/mesos/blob/master/src/scheduler/scheduler.cpp#L275)
>  missing in master. See motivation on MESOS-2288.
> 
> 
> Diffs
> -----
> 
>   src/master/master.hpp 6a39df0 
>   src/master/master.cpp f10a3cf 
>   src/master/validation.cpp acc35b2 
> 
> Diff: https://reviews.apache.org/r/30952/diff/
> 
> 
> Testing
> -------
> 
> make check and distcheck. Tests will be added with new HTTP API endpoints 
> tests.
> 
> 
> Thanks,
> 
> Isabel Jimenez
> 
>

Reply via email to