You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "Apache Spark (JIRA)" <ji...@apache.org> on 2015/04/30 20:29:08 UTC
[jira] [Commented] (SPARK-7281) No option for AM native library
path in yarn-client mode.
[ https://issues.apache.org/jira/browse/SPARK-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14522000#comment-14522000 ]
Apache Spark commented on SPARK-7281:
-------------------------------------
User 'vanzin' has created a pull request for this issue:
https://github.com/apache/spark/pull/5813
> No option for AM native library path in yarn-client mode.
> ---------------------------------------------------------
>
> Key: SPARK-7281
> URL: https://issues.apache.org/jira/browse/SPARK-7281
> Project: Spark
> Issue Type: Bug
> Components: YARN
> Affects Versions: 1.3.1, 1.4.0
> Reporter: Marcelo Vanzin
> Priority: Minor
>
> There are many options to control the AM in yarn-client mode, but the option for the native library path is missing. Without that, users have to change the process's environment directly (using {{spark.yarn.appMasterEnv.LD_LIBRARY_PATH}} on Linux, for example) to add things to the native library path.
> Depending on the cluster's configuration, native libraries are needed in the AM (for example when using the JNI-based group mapper in Hadoop).
> There's also no option to control the AM's classpath, but since the AM is not running user code, that seems less important.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org