spark-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Fernando Pereira (JIRA)" <j...@apache.org>
Subject [jira] [Comment Edited] (SPARK-19256) Hive bucketing support
Date Fri, 02 Feb 2018 08:51:00 GMT

    [ https://issues.apache.org/jira/browse/SPARK-19256?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16345397#comment-16345397
] 

Fernando Pereira edited comment on SPARK-19256 at 2/2/18 8:50 AM:
------------------------------------------------------------------

Thanks a lot for this great contribution to Spark.

I was just wondering, would it make sense to apply this to direct outputs (e.g. write.parquet()),
so that we could keep partitioning information - and again avoid reshuffling data before a
merge? I believe this is most what saveAsTable() does by default in Spark, but to my mind it
would improve the DataFrame write API and make these performance benefits more accessible.

Update:
I've just noticed that it has been considered in [https://github.com/apache/spark/pull/13452.
] [~cloud_fan] [ |https://github.com/apache/spark/pull/13452.]- Is there an Issue to follow
up on this feature? Eventually we could simply store a metadata json file together with the
data files.


was (Author: ferdonline):
Thanks a lot for this great contribution to Spark.

I was just wondering, would it make sense to apply this to direct outputs (e.g. write.parquet()),
so that we could keep partitioning information - and again avoid reshuffling data before a
merge? I believe this is most what saveAsTable() does by default in Spark, but to my mind it
would improve the DataFrame write API and make these performance benefits more accessible.

> Hive bucketing support
> ----------------------
>
>                 Key: SPARK-19256
>                 URL: https://issues.apache.org/jira/browse/SPARK-19256
>             Project: Spark
>          Issue Type: Umbrella
>          Components: SQL
>    Affects Versions: 2.1.0
>            Reporter: Tejas Patil
>            Priority: Minor
>
> JIRA to track design discussions and tasks related to Hive bucketing support in Spark.
> Proposal : https://docs.google.com/document/d/1a8IDh23RAkrkg9YYAeO51F4aGO8-xAlupKwdshve2fc/edit?usp=sharing



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org


Mime
View raw message