You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "Shingo Furuyama (Jira)" <ji...@apache.org> on 2020/04/13 07:28:00 UTC
[jira] [Commented] (SPARK-31432) bin/sbin scripts should allow to
customize jars dir
[ https://issues.apache.org/jira/browse/SPARK-31432?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17082129#comment-17082129 ]
Shingo Furuyama commented on SPARK-31432:
-----------------------------------------
I will soon send a PR to the master branch. If the PR is merged, I will send it branch-2.4.
> bin/sbin scripts should allow to customize jars dir
> ---------------------------------------------------
>
> Key: SPARK-31432
> URL: https://issues.apache.org/jira/browse/SPARK-31432
> Project: Spark
> Issue Type: Improvement
> Components: Deploy
> Affects Versions: 2.4.5, 3.0.0
> Reporter: Shingo Furuyama
> Priority: Minor
>
> In the script under bin/sbin, it is better that we can specify SPARK_JARS_DIR as same as SPARK_CONF_DIR.
> Our usecase:
> We are trying to employ spark 2.4.5 with YARN in HDP2.6.4. Since there is an incompatible conflict on commons-lang3 between spark 2.4.5 and HDP2.6.4, we tweak the jars by Maven Shade Plugin.
> The jars slightly differ from jars in spark 2.4.5, and we locate it in a directory different from the default. So it is useful for us if we can set SPARK_JARS_DIR for bin/sbin scripts to point the direcotry.
> We can do that without the modification by deploying spark home as many as set of jars, but it is somehow redundant.
> Common usecase:
> I believe there is a similer usecase. For example, deploying spark built for scala 2.11 and 2.12 in a machine and switch jars location by setting SPARK_JARS_DIR.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org