[
https://issues.apache.org/jira/browse/PHOENIX-6993?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Abhishek Kothalikar reassigned PHOENIX-6993:
--------------------------------------------
Assignee: Abhishek Kothalikar
> Push limit and order by to skip-scan join uncovered index queries where
> possible
> --------------------------------------------------------------------------------
>
> Key: PHOENIX-6993
> URL: https://issues.apache.org/jira/browse/PHOENIX-6993
> Project: Phoenix
> Issue Type: Improvement
> Components: core
> Reporter: Istvan Toth
> Assignee: Abhishek Kothalikar
> Priority: Major
>
> Whwn using the old skip-scan-join query plan for hinted uncovered indexes for
> topN queries, we are ignoring the limit and order by caluses.
> In the indexed case, we know that there is a 1:1 relation between the index
> table rows and the data tables rows, so - if the order by can be resolved
> from the index table only, we could push down the limit and order by queries
> to the index query, and speed up topN queries a lot.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
