[
https://issues.apache.org/jira/browse/CRUNCH-375?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13966567#comment-13966567
]
Josh Wills commented on CRUNCH-375:
-----------------------------------
My half-baked thinking on this topic was a PStream interface that lived above
PCollection and supported some of the same functions, but a different PStream
interface that is parallel to PCollection (but supported DoFns and PTypes)
might make more sense.
> Implement Crunch on top of Apache Storm
> ---------------------------------------
>
> Key: CRUNCH-375
> URL: https://issues.apache.org/jira/browse/CRUNCH-375
> Project: Crunch
> Issue Type: Wish
> Components: Core
> Reporter: Corey J. Nolet
> Assignee: Josh Wills
> Labels: proposal
>
> I'm proposing this as a way for Apache Storm users to be able to write their
> pipelines and run them both over Storm and MapReduce.
> There was movement on implementing Pig on Storm which established the ability
> to load/store through Hadoop input/output formats on Storm.
> I think this feature would be well-accepted into the community.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
