[
https://issues.apache.org/jira/browse/HIVE-15928?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15880047#comment-15880047
]
Lefty Leverenz commented on HIVE-15928:
---------------------------------------
Doc note: This adds configuration parameter *hive.druid.select.distribute* and
amends the description of *hive.druid.select.threshold*, which was created by
HIVE-14217 (also in 2.2.0). They need to be documented in the wiki.
* [Configuration Properties -- Query and DDL Execution |
https://cwiki.apache.org/confluence/display/Hive/Configuration+Properties#ConfigurationProperties-QueryandDDLExecution]
* [Druid Integration |
https://cwiki.apache.org/confluence/display/Hive/Druid+Integration]
Added a TODOC2.2 label.
> Parallelization of Select queries in Druid handler
> --------------------------------------------------
>
> Key: HIVE-15928
> URL: https://issues.apache.org/jira/browse/HIVE-15928
> Project: Hive
> Issue Type: Sub-task
> Components: Druid integration
> Affects Versions: 2.2.0
> Reporter: Jesus Camacho Rodriguez
> Assignee: Jesus Camacho Rodriguez
> Labels: TODOC2.2
> Fix For: 2.2.0
>
> Attachments: HIVE-15928.01.patch, HIVE-15928.02.patch,
> HIVE-15928.patch
>
>
> Even if we split a Select query along its time dimension, parallelization is
> limited as all queries will hit the broker node. Instead, we can interrogate
> the broker to get the Druid nodes that contain the data, and query those
> nodes directly.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
