Spark development¶
The ListenBrainz Spark environment is used for computing statistics and computing recommendations. If you’re just working on adding a feature to the ListenBrainz webserver, you do not need to set up the Spark development environment. However, if you’re looking to add a new stat or improve our fledgling recommender system, you’ll need both the webserver and the spark development environment.
This guide should explain how to develop and test new features for ListenBrainz that use Spark.
Set up the webserver¶
The spark environment is dependent on the webserver. Follow the steps in the guide to set up the webserver environment.
Create listenbrainz_spark/config.py¶
The spark environment needs a config.py in the listenbrainz_spark/ dir. Create it by copying from the sample config file.
cp listenbrainz_spark/config.py.sample listenbrainz_spark/config.py
Initialize ListenBrainz Spark containers¶
Run the following command to build the spark containers.
./develop.sh spark build
The first time you build the containers, you also need to format the namenode
container.
./develop.sh spark format
Note
You can run ./develop.sh spark format
any time that you want to delete all of the
data that is loaded in spark. This will shut down the spark docker cluster, remove
the docker volumes used to store the data, and recreate the HDFS filesystem.
Your development environment is now ready. Now, let’s actually see ListenBrainz Spark in action!
Bring containers up¶
First, ensure that you are running the main ListenBrainz development environment:
./develop.sh up
Start the ListenBrainz Spark environment:
./develop.sh spark up
This will also bring up the spark reader container which is described in detail here.
Import data into the spark environment¶
We provide small data dumps that are helpful for working with real ListenBrainz data. Download and import a data dump into your spark environment using the following commands in a separate terminal.
./develop.sh spark run spark_reader python manage.py spark request_import_incremental
Now, you are all set to begin making changes and seeing them in real-time inside of your development environment!
Once you are done with your work, shut down the containers using the following command.
./develop.sh spark down
Note
You’ll need to run ./develop.sh spark down
every time you restart your environment, otherwise hadoop errors out.
Working with request_consumer¶
The ListenBrainz webserver and spark cluster interact with each other via the request consumer. For a more detailed guide on working with the request consumer, read this document.
Test your changes with unit tests¶
Unit tests are an important part of ListenBrainz Spark. It helps make it easier for developers to test changes and help prevent easily avoidable mistakes later on. Before committing new code or making a pull request, run the unit tests on your code.
./test.sh spark
This builds and runs the containers needed for the tests. This script configures test-specific data volumes so that test data is isolated from your development data.
When the tests complete, you will see if your changes are valid or not. These tests are a helpful way to validate new changes without a lot of work.