{% include JB/setup %}
Apache Zeppelin provides Interpreter Installation mechanism for whom downloaded Zeppelin netinst
binary package, or just want to install another 3rd party interpreters.
Apache Zeppelin provides several interpreters as community managed interpreters. If you downloaded netinst
binary package, you need to install by using below commands.
./bin/install-interpreter.sh --all
./bin/install-interpreter.sh --name md,shell,jdbc,python
You can get full list of community managed interpreters by running
./bin/install-interpreter.sh --list
Zeppelin support both Scala 2.10 and 2.11 for several interpreters as below:
If you install one of these interpreters only with --name
option, installer will download interpreter built with Scala 2.11 by default. If you want to specify Scala version, you will need to add --artifact
option. Here is the example of installing flink interpreter built with Scala 2.10.
./bin/install-interpreter.sh --name flink --artifact org.apache.zeppelin:zeppelin-flink_2.10:0.7.3
Spark distribution package has been built with Scala 2.10 until 1.6.2. If you have SPARK_HOME
set pointing to Spark version earlier than 2.0.0, you need to download Spark interpreter packaged with Scala 2.10. To do so, use follow command:
rm -rf ./interpreter/spark ./bin/install-interpreter.sh --name spark --artifact org.apache.zeppelin:zeppelin-spark_2.10:0.7.3
You can also install 3rd party interpreters located in the maven repository by using below commands.
./bin/install-interpreter.sh --name interpreter1 --artifact groupId1:artifact1:version1
The above command will download maven artifact groupId1:artifact1:version1
and all of it's transitive dependencies into interpreter/interpreter1
directory.
After restart Zeppelin, then create interpreter setting and bind it with your notebook.
./bin/install-interpreter.sh --name interpreter1,interpreter2 --artifact groupId1:artifact1:version1,groupId2:artifact2:version2
--name
and --artifact
arguments will recieve comma separated list.
You can also find the below community managed interpreter list in conf/interpreter-list
file.