You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@livy.apache.org by "Saisai Shao (Jira)" <ji...@apache.org> on 2020/03/26 06:11:00 UTC

[jira] [Assigned] (LIVY-751) Livy server should allow to customize LIVY_CLASSPATH

     [ https://issues.apache.org/jira/browse/LIVY-751?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Saisai Shao reassigned LIVY-751:
--------------------------------

    Assignee: Shingo Furuyama

> Livy server should allow to customize LIVY_CLASSPATH
> ----------------------------------------------------
>
>                 Key: LIVY-751
>                 URL: https://issues.apache.org/jira/browse/LIVY-751
>             Project: Livy
>          Issue Type: Improvement
>          Components: Server
>    Affects Versions: 0.7.0
>            Reporter: Shingo Furuyama
>            Assignee: Shingo Furuyama
>            Priority: Minor
>             Fix For: 0.8.0
>
>          Time Spent: 20m
>  Remaining Estimate: 0h
>
> What we want to do is - specify LIVY_CLASSPATH at the time of booting livy server to use the other version of hadoop, which is not included in livy artifact.
> The background is - we are trying to use livy 0.7.0-incubating and spark 2.4.5 with YARN HDP2.6.4, but we encountered an error due to the incompatibility of livy included hadoop and HDP2.6.4 hadoop. We came up with a workaround that 1. remove hadoop from livy artifact by building with `mvn -Dhadoop.scope=provided` and 2. add HDP2.6.4 hadoop jars to the classpath for livy server by `hadoop classpath` command at the time of booting livy server. However, bin/livy-server is not allowing change LIVY_CLASSPATH.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)