[
https://issues.apache.org/jira/browse/HIVE-7858?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14166210#comment-14166210
]
Ferdinand Xu commented on HIVE-7858:
------------------------------------
Hi [~brocknoland], I am afraid not. The Hive side just passes codec type to get
a real writer. And compression information is encapsulated in the real writer
from parquet side. Adding tests needs some changes on parquet side since the
ParquetOutput is not exposing the codec information as publicity.
https://github.com/apache/incubator-parquet-mr/blob/0148455170be07f89bd6b9230960a6cd510c7ca6/parquet-hadoop/src/main/java/parquet/hadoop/ParquetRecordWriter.java#L64
> Parquet compression should be configurable via table property
> -------------------------------------------------------------
>
> Key: HIVE-7858
> URL: https://issues.apache.org/jira/browse/HIVE-7858
> Project: Hive
> Issue Type: Improvement
> Reporter: Brock Noland
> Assignee: Ferdinand Xu
> Attachments: HIVE-7858.patch, HIVE-7858.patch
>
>
> ORC supports the orc.compress table property:
> https://cwiki.apache.org/confluence/display/Hive/LanguageManual+ORC
> {noformat}
> create table Addresses (
> name string,
> street string,
> city string,
> state string,
> zip int
> ) stored as orc tblproperties ("orc.compress"="NONE");
> {noformat}
> I think it'd be great to support the same for Parquet.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
