[
https://issues.apache.org/jira/browse/HADOOP-3412?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12609364#action_12609364
]
Matei Zaharia commented on HADOOP-3412:
---------------------------------------
While the idea of splitting up the TaskScheduler and JobQueue is
great, isn't having a getJobQueue() in TaskScheduler needlessly
complicating the interface? Now, when somebody wants to write a new
scheduling algorithm, they may have to write two classes, a
TaskScheduler and a JobQueue, and figure out how they'll interact. I'd
suggest having the add(JobInProgress) and remove(JobInProgress)
methods directly in the TaskScheduler, and removing getJobQueue(), so
that the JobTracker only calls methods of the TaskScheduler object.
The DefaultTaskScheduler can include a JobQueue that it passes the add/
remove methods to, but people writing new schedulers should have the
option to do something different.
Here are two concrete cases where this would be useful:
- Some TaskSchedulers, even if they use a JobQueue, might want to run
code in response to the add() and remove() methods (updating data
structures, deciding to preempt existing jobs because a new one, etc).
Right now this is not possible unless the JobQueue tells them it was
updated.
- Some schedulers might not logically have a job queue. For example,
we might imagine a scheduler that sorts tasks, not jobs, based on
locality and on whether the task is speculative.
Matei
> Refactor the scheduler out of the JobTracker
> --------------------------------------------
>
> Key: HADOOP-3412
> URL: https://issues.apache.org/jira/browse/HADOOP-3412
> Project: Hadoop Core
> Issue Type: Improvement
> Components: mapred
> Reporter: Brice Arnould
> Assignee: Brice Arnould
> Priority: Minor
> Fix For: 0.19.0
>
> Attachments: JobScheduler.patch, JobScheduler_v2.patch,
> JobScheduler_v3.patch, JobScheduler_v3b.patch, JobScheduler_v4.patch,
> JobScheduler_v5.patch, JobScheduler_v6.1.patch, JobScheduler_v6.2.patch,
> JobScheduler_v6.3.patch, JobScheduler_v6.4.patch, JobScheduler_v6.patch,
> JobScheduler_v7.1.patch, JobScheduler_v7.patch, RackAwareJobScheduler.java,
> SimpleResourceAwareJobScheduler.java
>
>
> First I would like warn you that my proposition is assumed to be very naive.
> I just hope that reading it won't make you lose time.
> h4. The aim
> It seems to me that improving Hadoop scheduling could be very profitable.
> But, it is hard to implement and compare schedulers, because the scheduling
> logic is mixed within the rest of the JobTracker.
> This bug is the first step of an attempt to improve the Hadoop scheduler. It
> re-implements the current scheduling algorithm in a separate class called
> JobScheduler. This new class is instantiated in the JobTracker.
> h4. Bug fixed as a side effects
> This patch probably cannot be submited as it is.
> A first difficulty is that it does not have exactly the same behaviour than
> the current JobTracker. More precisely, it doesn't re-implement things like
> code that seems to be never called or concurency problems.
> I wrote TOCONFIRM where my proposition differ from the current
> implementation, so you can find them easily.
> I know that fixing bugs silently is bad. So, independently of what you decide
> about this patch, I will open issues for bugs that you confirm.
> h4. Other side effects
> Another side effect of this patch is to add documentation about each step of
> the scheduling. I hope that it will help future improvement by lowering the
> level required to contribute to the scheduler.
> It also reduces the complexity and the granularity of the JobTracker (making
> it more parallel).
> h4. The future
> If you feel that this is a step the right direction, I will try to propose a
> JobSchedulerInterface that many JobSchedulers could implement and to propose
> alternatives to the current « FifoJobScheduler ». If some of you have ideas
> about that please tell ^^ I will also open issues for things marked as FIXME
> in the patch.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
