> On June 11, 2013, 10:03 p.m., Vinod Kone wrote:
> > hadoop/mesos/src/java/org/apache/hadoop/mapred/MesosScheduler.java, lines 
> > 114-148
> > <https://reviews.apache.org/r/11119/diff/5/?file=303828#file303828line114>
> >
> >     Are you missing a diff here? I don't see this resolved per my earlier 
> > comment?
> 
> Brenden Matthews wrote:
>     It's in later commits:
>     
>     https://reviews.apache.org/r/11603/
>     https://reviews.apache.org/r/11489/
>     
>     I can drop the other request and roll it in to this one if you want.  
> It's just very difficult for me to keep track of all the different changes.
> 
> Vinod Kone wrote:
>     Either that, or dicard this if those two reviews subsume these changes.
> 
> Brenden Matthews wrote:
>     What's the procedure for capturing dependencies with review board?  Most 
> of my patches depend on previous ones.  They need to be added sequentially.  
> Here's the current state:
>     
>     https://github.com/airbnb/mesos/commits/testing
> 
> Vinod Kone wrote:
>     Unfortunately, I don't think there is a good way to capture dependencies. 
> We have plans to augment our postreviews.py script to add some metadata 
> (e.g., parent review id)  in the commit message.
>     
>     For this particular one, I suggested to drop this only if other reviews 
> subsumed the changes. Ofcourse, if there are some changes that are only in 
> this patch, you cannot discard this.

I this case I think it's better to keep these as separate changes because there 
are several other patches between them.  I want to a) keep the change history 
and b) avoid a merge nightmare.

Another problem is that this patch depends on others, in particular, this, 
which does not have a shipit:

https://reviews.apache.org/r/11118/

Can you take a look?


- Brenden


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/11119/#review21749
-----------------------------------------------------------


On June 11, 2013, 10:12 p.m., Brenden Matthews wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/11119/
> -----------------------------------------------------------
> 
> (Updated June 11, 2013, 10:12 p.m.)
> 
> 
> Review request for mesos.
> 
> 
> Description
> -------
> 
> Don't prematurely kill TaskTrackers.
> 
> We assign the JobID to the internal Mesos tracker (for the TaskTracker)
> rather than the TaskID.
> 
> In the case where a TaskTracker has only map tasks assigned to it and
> they have all completed, we mustn't terminate the tracker until the
> entire job has finished, including the map tasks.
> 
> Review: https://reviews.apache.org/r/11119
> 
> 
> Diffs
> -----
> 
>   hadoop/mesos/src/java/org/apache/hadoop/mapred/MesosScheduler.java 
> afe401f5265e3d9494af7eace42eec45943184a3 
> 
> Diff: https://reviews.apache.org/r/11119/diff/
> 
> 
> Testing
> -------
> 
> Used in production at airbnb.
> 
> make -j10 check && cd hadoop && make hadoop-2.0.0-mr1-cdh4.2.1 && make 
> hadoop-0.20.205.0 && make hadoop-0.20.2-cdh3u3
> 
> 
> Thanks,
> 
> Brenden Matthews
> 
>

Reply via email to