asarkar/zeppelin repository overview

⁠Getting Started

1. Start Zeppelin:

   docker run --rm --name=zeppelin -p 8080:8080 -p 8081:8081 \
   -v ~/metrics:/srv/metrics/:ro \
   -d asarkar/zeppelin:0.7.0
   

   Copy

   This exposes the Web UI on port 8080 and mounts host directory ~/metrics to /srv/metrics/ in the container.

2. Go to http://localhost:8080⁠ and create a new note. Give it a name.

   The option may not be available until the circle on the right upper corner turns green.

3. To load a bunch of JSON files, type the following and run it (using the > button). This uses a built-in Spark interpreter (the default).

   val df = spark.read.json("/srv/metrics/*.json")
   df.registerTempTable("metrics")
   

   Copy

   This creates a temp table named metrics that you can query. To learn more about Sprak SQL, see this⁠.

   1. Spark will automatically infer the schema. To see the schema, run df.printSchema().
   2. To pass arguments to the Scala compiler, for example to view deprecation warnings, go to http://localhost:8080/#/interpreter⁠, click "edit", and add -deprecation to the args property.

4. In another paragraph, start writing queries. For example:

   %sql select age, count(1) from metrics where age < 30 group by age order by age
   

   Copy

   More examples can be found here: Zeppelin Tutorial⁠

   The output can be visualized in tabular or one of various graphical formats.

   Column names with periods in them have to be escaped by enclosing in backticks.

   See actuator-meets-zeppelin⁠ project for examples.