> On May 6, 2013, 11:44 p.m., Ben Mahler wrote:
> > hadoop/mesos/src/java/org/apache/hadoop/mapred/MesosScheduler.java, line 139
> > <https://reviews.apache.org/r/10920/diff/1/?file=287627#file287627line139>
> >
> >     I don't see any idle check here.. is this comment valid?

This is a left-over comment from some other work.


> On May 6, 2013, 11:44 p.m., Ben Mahler wrote:
> > hadoop/mesos/src/java/org/apache/hadoop/mapred/MesosScheduler.java, lines 
> > 114-117
> > <https://reviews.apache.org/r/10920/diff/1/?file=287627#file287627line114>
> >
> >     Is this possible? Looks like a LOG.fatal rather than a continue?

It's certainly possible.  You can have many different jobs/tasks on the same 
trackers.


> On May 6, 2013, 11:44 p.m., Ben Mahler wrote:
> > hadoop/mesos/src/java/org/apache/hadoop/mapred/MesosScheduler.java, lines 
> > 137-147
> > <https://reviews.apache.org/r/10920/diff/1/?file=287627#file287627line137>
> >
> >     Looks like this no longer needs to be in the TaskStatus loop? Can you 
> > update the code here, it doesn't look like the TaskStatus loop is needed 
> > anymore?

I don't quite see how it isn't needed.  You could move the logic elsewhere, but 
it still has to happen.


- Brenden


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/10920/#review20232
-----------------------------------------------------------


On May 7, 2013, 7 p.m., Brenden Matthews wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/10920/
> -----------------------------------------------------------
> 
> (Updated May 7, 2013, 7 p.m.)
> 
> 
> Review request for mesos.
> 
> 
> Description
> -------
> 
> From 5b29cf70eb7b6ee0f45148668ed8848bf9a941a4 Mon Sep 17 00:00:00 2001
> From: Brenden Matthews <[email protected]>
> Date: Fri, 26 Apr 2013 11:42:18 -0700
> Subject: [PATCH 10/19] 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.
> ---
>  .../org/apache/hadoop/mapred/MesosScheduler.java   |  371 
> ++++++++++----------
>  1 file changed, 189 insertions(+), 182 deletions(-)
> 
> 
> Diffs
> -----
> 
>   hadoop/mesos/src/java/org/apache/hadoop/mapred/MesosScheduler.java afe401f 
> 
> Diff: https://reviews.apache.org/r/10920/diff/
> 
> 
> Testing
> -------
> 
> Used in production at airbnb.
> 
> 
> Thanks,
> 
> Brenden Matthews
> 
>

Reply via email to