[
https://issues.apache.org/jira/browse/HIVE-10257?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Dong Chen updated HIVE-10257:
-----------------------------
Attachment: HIVE-10257-parquet.2.patch
Sure, Thanks!
Patch updated.
Since this renaming has been fixed in HIVE-10256, I updated this patch based on
the code there. So this patch has to be merged after that one.
Sorroy for the inconvenience.
> Ensure Parquet Hive has null optimization
> -----------------------------------------
>
> Key: HIVE-10257
> URL: https://issues.apache.org/jira/browse/HIVE-10257
> Project: Hive
> Issue Type: Sub-task
> Components: Tests
> Reporter: Dong Chen
> Assignee: Dong Chen
> Attachments: HIVE-10257-parquet.1.patch, HIVE-10257-parquet.2.patch,
> HIVE-10257-parquet.patch
>
>
> In Parquet statistics, a boolean value {{hasNonNullValue}} is used for each
> column chunk. Hive could use this value to skip a column, avoid null-checking
> logic, and speed up vectorization like HIVE-4478 (in the future, Parquet
> vectorization is not completed yet).
> In this Jira we could check whether this null optimization works, and make
> changes if any.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
