[
https://issues.apache.org/jira/browse/SPARK-13876?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Michael Armbrust resolved SPARK-13876.
--------------------------------------
Resolution: Fixed
Fix Version/s: 2.0.0
Resolved by https://github.com/apache/spark/pull/11646
> Strategy for planning scans of files
> ------------------------------------
>
> Key: SPARK-13876
> URL: https://issues.apache.org/jira/browse/SPARK-13876
> Project: Spark
> Issue Type: Sub-task
> Components: SQL
> Reporter: Michael Armbrust
> Assignee: Michael Armbrust
> Priority: Critical
> Fix For: 2.0.0
>
>
> We should specialize the logic for planning scans over sets of files.
> Requirements:
> - remove the need to have RDD, broadcastedHadoopConf and other distributed
> concerns in the public API of org.apache.spark.sql.sources.FileFormat
> - Partition column appending should be delegated to the format to avoid an
> extra copy / devectorization when appending partition columns
> - Should minimize the amount of data that is shipped to each executor (i.e.
> it does not send the whole list of files to every worker in the form of a
> hadoop conf)
> - should natively support bucketing files into partitions, and thus does not
> require coalescing / creating a UnionRDD with the correct partitioning.
> - Small files should be automatically coalesced into fewer tasks
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
