[
https://issues.apache.org/jira/browse/HADOOP-4665?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12664639#action_12664639
]
Joydeep Sen Sarma commented on HADOOP-4665:
-------------------------------------------
Dhruba and I looked at this together and got stuck on getAllowedLocalityLevel()
- why subtract nodeLocalWait from rackLocalWait
- why getting config variables each time
- if we were not rack/node local last time - why don't we wait for locality
next time? Seems like once we lose locality - we are anyway going to run up a
deficit and then schedule a boatload of non-local tasks. seems like the whether
we want to wait for locality or not should be based on how much deficit we are
incurring and whether it's still warranted to wait for locality (as opposed to
whether we were able to schedule the last task locally)
thoughts?
> Add preemption to the fair scheduler
> ------------------------------------
>
> Key: HADOOP-4665
> URL: https://issues.apache.org/jira/browse/HADOOP-4665
> Project: Hadoop Core
> Issue Type: New Feature
> Components: contrib/fair-share
> Reporter: Matei Zaharia
> Attachments: fs-preemption-v0.patch
>
>
> Task preemption is necessary in a multi-user Hadoop cluster for two reasons:
> users might submit long-running tasks by mistake (e.g. an infinite loop in a
> map program), or tasks may be long due to having to process large amounts of
> data. The Fair Scheduler (HADOOP-3746) has a concept of guaranteed capacity
> for certain queues, as well as a goal of providing good performance for
> interactive jobs on average through fair sharing. Therefore, it will support
> preempting under two conditions:
> 1) A job isn't getting its _guaranteed_ share of the cluster for at least T1
> seconds.
> 2) A job is getting significantly less than its _fair_ share for T2 seconds
> (e.g. less than half its share).
> T1 will be chosen smaller than T2 (and will be configurable per queue) to
> meet guarantees quickly. T2 is meant as a last resort in case non-critical
> jobs in queues with no guaranteed capacity are being starved.
> When deciding which tasks to kill to make room for the job, we will use the
> following heuristics:
> - Look for tasks to kill only in jobs that have more than their fair share,
> ordering these by deficit (most overscheduled jobs first).
> - For maps: kill tasks that have run for the least amount of time (limiting
> wasted time).
> - For reduces: similar to maps, but give extra preference for reduces in the
> copy phase where there is not much map output per task (at Facebook, we have
> observed this to be the main time we need preemption - when a job has a long
> map phase and its reducers are mostly sitting idle and filling up slots).
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
