[
https://issues.apache.org/jira/browse/HIVE-10029?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14498716#comment-14498716
]
Siddharth Seth commented on HIVE-10029:
---------------------------------------
Yes, to the most part. We'll likely a need a follow up to provide data to the
pre-emption queue.
> LLAP: Scheduling of work from different queries within the daemon
> -----------------------------------------------------------------
>
> Key: HIVE-10029
> URL: https://issues.apache.org/jira/browse/HIVE-10029
> Project: Hive
> Issue Type: Sub-task
> Reporter: Siddharth Seth
> Fix For: llap
>
>
> The current implementation is a simple queue - whichever query wins the race
> to submit work to a daemon will execute first.
> A policy around this may be useful - potentially a fair share, or a first
> query in gets all slots approach.
> Also, prioritiy associated with work within a query should be considered.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
