Is it possible to Clone Interpreter in zeppelin

[#1]

Is it possible to clone the zeppelin interpreter with all settings programmatically(REST?) or using UI(I did not see that as an option). I created a new one manually but wanted to check if there was a better way?. I want to have 2 or more interpreters setup pointing @ different versions of spark 1.4 and 1.5. But i can easily see this extending to other components.

[#2 · Accepted solution]

would copying and modifying the interpreter file under /incubator-zeppelin/interpreter folder helps?

[#3]

@nshawa@hortonworks.com after talking to @Ali Bajwa it is not possible to point to a different interpreter @ this point. As it is compiled to a specific version of spark. So you cant dynamically chose which spark version you want to use. I filed a JIRA for this feature: https://issues.apache.org/jira/browse/ZEPPELIN-392

[#4]

make perfect since, wonder if it will work with backward compatibility though,right not I ended up with different Zeppelin folders pointed at different Spark versions

[#5]

Thats what I am doing now, I have sandbox 2.3.2 using out of the box ambari zeppelin pointing to 1.4 and setting up now local zeppelin to point to 1.5.1

[#6]

This should now be solved, starting Zeppelin 0.5.5 you dont need to rebuild for different Spark/Hadoop versions...

enjoy 🙂

[#7]

@Ned Shawa nice! Any insight when we would officially support Zeppelin in HDP?

[#8]

After talking to @Ali Bajwa it is not possible to point to a different interpreter @ this point. As it is compiled to a specific version of spark. So you cant dynamically chose which spark version you want to use. I filed a JIRA for this feature: https://issues.apache.org/jira/browse/ZEPPELIN-392