phoenix-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "nicolas maillard (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (PHOENIX-1093) Phoenix on Tez
Date Fri, 28 Nov 2014 13:46:12 GMT

    [ https://issues.apache.org/jira/browse/PHOENIX-1093?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14228297#comment-14228297
] 

nicolas maillard commented on PHOENIX-1093:
-------------------------------------------

A lot of this work is being tackled by other means so I will close this unless someone wants
to revisit.

> Phoenix on Tez
> --------------
>
>                 Key: PHOENIX-1093
>                 URL: https://issues.apache.org/jira/browse/PHOENIX-1093
>             Project: Phoenix
>          Issue Type: New Feature
>            Reporter: nicolas maillard
>
> Hello
> In case of complex queries and or joins would it make sense to have a phoenix intermediate
server spin up a tez dag to resolve a query before sending back to the client.The coprocessors
output could be seen as tez inputs. 
> So in the event of resulting datasets not fitting in memory for say complex joins. The
results phoenix servers could become inputs for a tez job. Tez being a framework it could
be embedded in this phoenix intermediate server,which would in turn decide to spin up a tez
job or not.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message