[jira] [Commented] (HIVE-9422) LLAP: row-level vectorized SARGs

[Gopal V (JIRA)]

[ 
https://issues.apache.org/jira/browse/HIVE-9422?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14284671#comment-14284671
 ] 

Gopal V commented on HIVE-9422:
---

The ORC readers today apply SARGs for all readers - vectorized and row-readers.

This is for vectorized readers.

> LLAP: row-level vectorized SARGs
> 
>
> Key: HIVE-9422
> URL: https://issues.apache.org/jira/browse/HIVE-9422
> Project: Hive
>  Issue Type: Sub-task
>Reporter: Sergey Shelukhin
>
> When VRBs are built from encoded data, sargs can be applied on low level to 
> reduce the number of rows to process.



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

[jira] [Commented] (HIVE-9422) LLAP: row-level vectorized SARGs

[Sergey Shelukhin (JIRA)]

[ 
https://issues.apache.org/jira/browse/HIVE-9422?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14284666#comment-14284666
 ] 

Sergey Shelukhin commented on HIVE-9422:


what is vectorized? sargs can
1) operate on encoded data (may be arcane given the proliferation of ORC 
encodings)
2) decode data as usual for sarg (as few values as possible), apply sarg, if 
needed decode other columns and put in vrb, otherwise skip
3) decode data into vrbs, then apply sargs.

Do you make emphasis on 3?

> LLAP: row-level vectorized SARGs
> 
>
> Key: HIVE-9422
> URL: https://issues.apache.org/jira/browse/HIVE-9422
> Project: Hive
>  Issue Type: Sub-task
>Reporter: Sergey Shelukhin
>
> When VRBs are built from encoded data, sargs can be applied on low level to 
> reduce the number of rows to process.



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