[
https://issues.apache.org/jira/browse/MAPREDUCE-548?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12733130#action_12733130
]
Devaraj Das commented on MAPREDUCE-548:
---------------------------------------
I think what Matei did is okay. Basically, the tasksInited atomic boolean is
used to take care of cases where job is failed/killed while initTasks is in
progress (HADOOP-4236), and, for getting the setup/cleanup tasks without trying
to lock the JobInProgress (HADOOP-5285). I don't think obtainNewMap* methods
need to check the tasksInited flag. In those methods, the check for (jobStatus
!= JobStatus,RUNNING) should be sufficient since these methods lock the
JobInProgress object, and, we invoke these methods only after initTasks has
run. It will be nice if someone validates this..
But having the checks for tasksInited in those methods is harmless except for
the extra check..
> Global scheduling in the Fair Scheduler
> ---------------------------------------
>
> Key: MAPREDUCE-548
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-548
> Project: Hadoop Map/Reduce
> Issue Type: New Feature
> Reporter: Matei Zaharia
> Assignee: Matei Zaharia
> Fix For: 0.21.0
>
> Attachments: fs-global-v0.patch, hadoop-4667-v1.patch,
> hadoop-4667-v1b.patch, hadoop-4667-v2.patch, HADOOP-4667_api.patch,
> mapreduce-548-v1.patch, mapreduce-548-v2.patch, mapreduce-548-v3.patch,
> mapreduce-548.patch
>
>
> The current schedulers in Hadoop all examine a single job on every heartbeat
> when choosing which tasks to assign, choosing the job based on FIFO or fair
> sharing. There are inherent limitations to this approach. For example, if the
> job at the front of the queue is small (e.g. 10 maps, in a cluster of 100
> nodes), then on average it will launch only one local map on the first 10
> heartbeats while it is at the head of the queue. This leads to very poor
> locality for small jobs. Instead, we need a more "global" view of scheduling
> that can look at multiple jobs. To resolve the locality problem, we will use
> the following algorithm:
> - If the job at the head of the queue has no node-local task to launch, skip
> it and look through other jobs.
> - If a job has waited at least T1 seconds while being skipped, also allow it
> to launch rack-local tasks.
> - If a job has waited at least T2 > T1 seconds, also allow it to launch
> off-rack tasks.
> This algorithm improves locality while bounding the delay that any job
> experiences in launching a task.
> It turns out that whether waiting is useful depends on how many tasks are
> left in the job - the probability of getting a heartbeat from a node with a
> local task - and on whether the job is CPU or IO bound. Thus there may be
> logic for removing the wait on the last few tasks in the job.
> As a related issue, once we allow global scheduling, we can launch multiple
> tasks per heartbeat, as in HADOOP-3136. The initial implementation of
> HADOOP-3136 adversely affected performance because it only launched multiple
> tasks from the same job, but with the wait rule above, we will only do this
> for jobs that are allowed to launch non-local tasks.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
