hadoop-yarn-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Konstantinos Karanasos (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (YARN-5202) Dynamic Overcommit of Node Resources - POC
Date Tue, 14 Jun 2016 01:35:09 GMT

    [ https://issues.apache.org/jira/browse/YARN-5202?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15328763#comment-15328763

Konstantinos Karanasos commented on YARN-5202:

I think it is definitely useful to have a first contained solution that people can try out
to see the effects of resource over-commitment in practice.

bq. Make sure the nodes protect themselves - i.e. if memory utilization on a node gets too
high, preempt something - preferably something from a preemptable queue
You can take a look at YARN-2883 where we added support for queuing of containers at the NMs.
The idea is to use opportunistic containers (YARN-2882) when queuing additional tasks or over-committing
resources, which will be the first candidates for killing/preempting when needed (this is
what YARN-1011 is using).
However, if you check the internals of YARN-2882, you will see that we keep both a queued
guaranteed and a queued opportunistic data structure, which means you can also queue guaranteed
This way you don't need to care about container types, while you can use the queue to store
some additional containers and start/stop them as needed, based on the current resource utilization.

> Dynamic Overcommit of Node Resources - POC
> ------------------------------------------
>                 Key: YARN-5202
>                 URL: https://issues.apache.org/jira/browse/YARN-5202
>             Project: Hadoop YARN
>          Issue Type: Improvement
>          Components: nodemanager, resourcemanager
>    Affects Versions: 3.0.0-alpha1
>            Reporter: Nathan Roberts
>            Assignee: Nathan Roberts
>         Attachments: YARN-5202.patch
> This Jira is to present a proof-of-concept implementation (collaboration between [~jlowe]
and myself) of a dynamic over-commit implementation in YARN.  The type of over-commit implemented
in this jira is similar to but not as full-featured as what's being implemented via YARN-1011.
YARN-1011 is where we see ourselves heading but we needed something quick and completely transparent
so that we could test it at scale with our varying workloads (mainly MapReduce, Spark, and
Tez). Doing so has shed some light on how much additional capacity we can achieve with over-commit
approaches, and has fleshed out some of the problems these approaches will face.
> Primary design goals:
> - Avoid changing protocols, application frameworks, or core scheduler logic,  - simply
adjust individual nodes' available resources based on current node utilization and then let
scheduler do what it normally does
> - Over-commit slowly, pull back aggressively - If things are looking good and there is
demand, slowly add resource. If memory starts to look over-utilized, aggressively reduce the
amount of over-commit.
> - Make sure the nodes protect themselves - i.e. if memory utilization on a node gets
too high, preempt something - preferably something from a preemptable queue
> A patch against trunk will be attached shortly.  Some notes on the patch:
> - This feature was originally developed against something akin to 2.7.  Since the patch
is mainly to explain the approach, we didn't do any sort of testing against trunk except for
basic build and basic unit tests
> - The key pieces of functionality are in {{SchedulerNode}}, {{AbstractYarnScheduler}},
and {{NodeResourceMonitorImpl}}. The remainder of the patch is mainly UI, Config, Metrics,
Tests, and some minor code duplication (e.g. to optimize node resource changes we treat an
over-commit resource change differently than an updateNodeResource change - i.e. remove_node/add_node
is just too expensive for the frequency of over-commit changes)
> - We only over-commit memory at this point. 

This message was sent by Atlassian JIRA

To unsubscribe, e-mail: yarn-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: yarn-issues-help@hadoop.apache.org

View raw message