[
https://issues.apache.org/jira/browse/HADOOP-9639?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13765749#comment-13765749
]
Sangjin Lee commented on HADOOP-9639:
-------------------------------------
Thanks Maysam for great feedback! Sorry my response was delayed.
1) We did think about using Zookeeper for the coordination. But we shied away
from it, as it may introduce another component to support this feature. To us
the trade-off didn't seem to be worth it.
2) The condition under which we query whether the app is active is if the app
just got in between the time the first modification time check and writing the
cleaner lock. So this check needs to be timely. I also think that the overall
volume of these queries would be pretty low...
3) I think arguments can be made for either approach. But one issue with not
deleting incorrect file is that all subsequent attempts at using this file will
result in another temp jar in that directory. If this is a frequently used
file, it would result in a continuous increase in the number of temp files in
that directory.
I hope these answer your questions. Thanks!
> truly shared cache for jars (jobjar/libjar)
> -------------------------------------------
>
> Key: HADOOP-9639
> URL: https://issues.apache.org/jira/browse/HADOOP-9639
> Project: Hadoop Common
> Issue Type: New Feature
> Components: filecache
> Affects Versions: 2.0.4-alpha
> Reporter: Sangjin Lee
> Assignee: Sangjin Lee
> Attachments: shared_cache_design.pdf, shared_cache_design_v2.pdf
>
>
> Currently there is the distributed cache that enables you to cache jars and
> files so that attempts from the same job can reuse them. However, sharing is
> limited with the distributed cache because it is normally on a per-job basis.
> On a large cluster, sometimes copying of jobjars and libjars becomes so
> prevalent that it consumes a large portion of the network bandwidth, not to
> speak of defeating the purpose of "bringing compute to where data is". This
> is wasteful because in most cases code doesn't change much across many jobs.
> I'd like to propose and discuss feasibility of introducing a truly shared
> cache so that multiple jobs from multiple users can share and cache jars.
> This JIRA is to open the discussion.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
