[
https://issues.apache.org/jira/browse/CARBONDATA-3994?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Kunal Kapoor resolved CARBONDATA-3994.
--------------------------------------
Fix Version/s: 2.1.0
Resolution: Fixed
> Skip Order by for map task if it is sort column and use limit pushdown for
> array_contains filter
> ------------------------------------------------------------------------------------------------
>
> Key: CARBONDATA-3994
> URL: https://issues.apache.org/jira/browse/CARBONDATA-3994
> Project: CarbonData
> Issue Type: Bug
> Reporter: Ajantha Bhat
> Assignee: Ajantha Bhat
> Priority: Major
> Fix For: 2.1.0
>
> Time Spent: 6h
> Remaining Estimate: 0h
>
> When the order by column is in sort column, every map task output will be
> already sorted. No need to sort the data again.
> Hence skipping the order at map task by changing plan node from
> {{TakeOrderedAndProject}} --> {{CarbonTakeOrderedAndProjectExec}}
> Also in this scenario collecting the limit at map task and Array_contains()
> will use this limit value for row scan filtering to break scan once limit
> value is reached.
> Also added a carbon property to control this .
> {{carbon.mapOrderPushDown.<db_name>_<table_name>.column}}
> Note: later we can improve this for other filters also to use the limit value.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
