[
https://issues.apache.org/jira/browse/PHOENIX-687?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14247826#comment-14247826
]
maghamravikiran commented on PHOENIX-687:
-----------------------------------------
Closing this as it is a duplicate of PHOENIX-1454
> Support better integration with Map-reduce
> ------------------------------------------
>
> Key: PHOENIX-687
> URL: https://issues.apache.org/jira/browse/PHOENIX-687
> Project: Phoenix
> Issue Type: Task
> Reporter: James Taylor
> Assignee: mravi
>
> Idea from the mailing list
> [here](https://groups.google.com/forum/_!topic/phoenix-hbase-user/sYpUR_zYZYI)
> and
> [here](https://groups.google.com/forum/_!topic/phoenix-hbase-user/gif-97F7Sn0),
> to support a SELECT query to determine the output of the Map function and/or
> an UPSERT SELECT in the Reduce function.
> Given that Phoenix already parallelizes your query in a similar way to a
> Map-reduce job, it's unclear to me what benefit this functionality would
> provide. Would the SELECT statement be running over only a single regions
> worth of data? What can you not do with an UPSERT SELECT that you can do with
> Map-reduce? I'm sure there are plenty of things, but it would be good to list
> the top ten to identify what's missing from Phoenix.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
