[
https://issues.apache.org/jira/browse/YARN-2791?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14276107#comment-14276107
]
Vinod Kumar Vavilapalli commented on YARN-2791:
-----------------------------------------------
Okay folks, I've read the design docs on both YARN-2791 (this JIRA) and
YARN-2139. This is indeed part of YARN-2139, and a direct dup of YARN-2618 and
other tickets.
Yes YARN-2139 is a much larger effort but it encompasses both scheduling and
isolation. The important tickets of YARN-2139 already were created before this
JIRA. I am going to close this as a duplicate in a day unless I see specific
tasks that are not covered under YARN-2139. If there are things that are not
covered indeed, I urge Swapnil Daingade, Santosh Marellaand and Yuliya Feldman
to file sub-tasks under YARN-2791.
As Karthik appealed before, let's have the design discussion over at YARN-2139,
merging things that are only here and missing in that JIRA. Due credit will be
given to all contributors to the design and implementation there.
I am oblivious who contributes code, but let's work together please!
> Add Disk as a resource for scheduling
> -------------------------------------
>
> Key: YARN-2791
> URL: https://issues.apache.org/jira/browse/YARN-2791
> Project: Hadoop YARN
> Issue Type: New Feature
> Components: scheduler
> Affects Versions: 2.5.1
> Reporter: Swapnil Daingade
> Assignee: Yuliya Feldman
> Attachments: DiskDriveAsResourceInYARN.pdf
>
>
> Currently, the number of disks present on a node is not considered a factor
> while scheduling containers on that node. Having large amount of memory on a
> node can lead to high number of containers being launched on that node, all
> of which compete for I/O bandwidth. This multiplexing of I/O across
> containers can lead to slower overall progress and sub-optimal resource
> utilization as containers starved for I/O bandwidth hold on to other
> resources like cpu and memory. This problem can be solved by considering disk
> as a resource and including it in deciding how many containers can be
> concurrently run on a node.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
