[
https://issues.apache.org/jira/browse/SPARK-22167?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16186008#comment-16186008
]
holdenk commented on SPARK-22167:
---------------------------------
So for some reason the R directory in the hadoop 2.7 build looks like:
holden@holden:~/repos/spark/spark-2.1.2-bin-hadoop2.7$ ls R
check-cran.sh CRAN_RELEASE.md create-docs.sh DOCUMENTATION.md
install-dev.bat install-dev.sh log4j.properties pkg README.md run-tests.sh
WINDOWS.md
holden@holden:~/repos/spark/spark-2.1.2-bin-hadoop2.7$
I think there is a race condition which only shows up on my laptop where the
Spark directory is modified in one of the previous build steps before copying
into the hadoop-2.7 version.
> Spark Packaging w/R distro issues
> ---------------------------------
>
> Key: SPARK-22167
> URL: https://issues.apache.org/jira/browse/SPARK-22167
> Project: Spark
> Issue Type: Bug
> Components: Build, SparkR
> Affects Versions: 2.1.2
> Reporter: holdenk
> Assignee: holdenk
> Priority: Blocker
>
> The Spark packaging for Spark R in 2.1.2 did not work as expected, namely the
> R directory was missing from the hadoop-2.7 bin distro. This is the version
> we build the PySpark package for so it's possible this is related.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
