You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@zeppelin.apache.org by AhyoungRyu <gi...@git.apache.org> on 2016/09/14 15:01:41 UTC

[GitHub] zeppelin issue #1339: [ZEPPELIN-1332] Remove spark-dependencies & suggest ne...

Github user AhyoungRyu commented on the issue:

    https://github.com/apache/zeppelin/pull/1339
  
    @bzz @Leemoonsoo 
    Sorry for my late response. I spent some time to test various cases on the different OS.
    I think it's ready for review(CI is green at last!). 
    I updated the PR description accordingly. Hope it helps you to remind the purpose of this PR :)
    
    Here is the list of changes after my initial commits
    
     - Directory of Spark bin
    I changed the dir of Spark bin from `interpreter/spark/` to `local-spark`. Since `mvn clean` will remove `interpreter/`, users will see \u201cDo you want to download local Spark?\u201d whenever they re-build and restart Zeppelin. So I think creating a new dir(`local-spark`) would be better in this case.
    
     - Supporting Windows users
    I wanted to create `download-spark.cmd` for Windows. But sadly we can\u2019t use many shell commands such as `curl`, `tar` and `sed` in the batch script. It was hard to find 100% compatible commands for Windows. Maybe we can guide the Windows users to install those commands by themselves, but it\u2019s a bit overdoing i think. Actually `download-spark` script is only for downloading the latest version of Spark and set `SPARK_HOME`. So I updated some docs to explain this as a alternative way. 
    
     - Documentations
    I think it\u2019s quite big change that ppl need to enter \u201cYes/No\u201d when they start Zeppelin. Even though it\u2019s only one time. So I updated `README.md`, `install.md` and `spark.md`. The below screenshot is `spark.md`. 
    ![screen shot 2016-09-13 at 8 22 34 pm](https://cloud.githubusercontent.com/assets/10060731/18516345/8b0820ac-7ad3-11e6-8581-11e3cb12c57a.png)
    
    
    And as @bzz said before, 
    >How about EMR\Dataproc\Juju\BigTop users, will the proposed change affect them? 
    
    Do we need to provide this local Spark mode for them? Actually it's my question.. :D
    And please feel free to point me for anything if it's needed.


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastructure@apache.org or file a JIRA ticket
with INFRA.
---