You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "roncenzhao (JIRA)" <ji...@apache.org> on 2017/03/19 14:13:41 UTC

[jira] [Commented] (SPARK-20019) spark can not load alluxio fileSystem after adding jar

    [ https://issues.apache.org/jira/browse/SPARK-20019?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15931702#comment-15931702 ] 

roncenzhao commented on SPARK-20019:
------------------------------------

The root cause of this exception is:

The `SessionState.addJar()` only set the new jarClassLoader into current thread contextClassLoader and the classLoader of  `sparkContext.hadoopConfiguration` is the old one which does not include the new jar path.

In the function `InsertIntoHiveTable.getStagingDir()`, it throws the `classNotFoundException` when `inputPath.getFileSystem(hadoopConf)` is called. Because it loads the FileSystem using the `classLoader` of `hadoopConf`.

> spark can not load alluxio fileSystem after adding jar
> ------------------------------------------------------
>
>                 Key: SPARK-20019
>                 URL: https://issues.apache.org/jira/browse/SPARK-20019
>             Project: Spark
>          Issue Type: Bug
>          Components: SQL
>    Affects Versions: 2.1.0
>            Reporter: roncenzhao
>         Attachments: exception_stack.png
>
>
> The follwing sql cannot load alluxioSystem and it throws `ClassNotFoundException`.
> ```
> add jar /xxx/xxx/alluxioxxx.jar;
> set fs.alluxio.impl=alluxio.hadoop.FileSystem;
> select * from alluxionTbl;
> ```



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

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