listenbrainz-server

Server for the ListenBrainz project.

Installation

These instructions are meant to get you started quickly with development process. Installation process in production environment might be different.

First, install ChefDK, then:

$ kitchen create    # creates vm for first time
$ kitchen converge  # provisions/installs from chef cookbooks
$ kitchen login     # ssh in to vm

MessyBrainz and ListenStore

ListenStore is our interface to cassandra. It's bundled as part of this repository, but you must install it so that python can find it.

$ cd listenstore
$ python setup.py develop

You can use setup.py develop to symlink the ListenStore directory into site-packages (letting you make changes easily), or use setup.py install to install it. If you use setup.py install and make any changes to ListenStore you must run install again.

MessyBrainz is an intermediate database to map unknown metadata to stable UUIDs, and then map these intermediate identifiers to MBIDs ( http://messybrainz.org, https://github.com/metabrainz/messybrainz-server ).

For now you must download and install MessyBrainz separately:

$ git clone https://github.com/metabrainz/messybrainz-server.git
$ cd messybrainz-server
$ pip install -r requirements.txt
$ python setup.py develop
$ cp config.py.sample config.py
$ python manage.py init_db

Future updates will map a local copy of MessyBrainz into the Vagrant server.

Python dependencies

Database should be ready. Before running any Python scripts, you need to install Python dependencies:

$ cd ~/listenbrainz
$ pip install -r requirements.txt

Configuration file

Copy the file config.py.sample to config.py and edit the postgres settings. For now you will need to allow your system user to connect as the postgres admin user with no password.

Register for a MusicBrainz application at https://musicbrainz.org/account/applications. During registration set the callback url to http://<your_host>/login/musicbrainz/post. If you don't change anything during setup, your_host will be 10.1.2.3:8080

Database initialization

To initialize the database (create user, tables, etc.) run this command:

$ python manage.py init_db

After that server should be ready to go.

Running

To run the webserver use this command:

$ python manage.py runserver

It should be accessible at http://10.1.2.3:8080/.

For information about running listenstore module see /listenstore/README.md file.

Documentation

Documentation for the ListenBrainz API is available at https://listenbrainz.readthedocs.org. You can build the documentation yourself:

$ cd docs
$ make clean html

Fixing problems

`LeaderNotAvailableError`

After setting up the server and trying to submit your listens to the server, this error might appear:

Kafka listens write error: <class 'kafka.common.LeaderNotAvailableError'>

One way to fix it is to delete all data directories created for Kafka and Zookeeper.

First, stop both Kafka and Zookeeper services:

$ sudo service kafka stop
$ sudo service zookeeper stop

Wipe Kafka's data directory:

$ sudo rm /var/lib/kafka/kafka-logs/*

and Zookeeper's:

$ sudo rm /var/lib/zookeeper/*

Start them again:

$ sudo service kafka start
$ sudo service zookeeper start

Finally, create a listens topic in Kafka:

$ sudo /opt/kafka/bin/kafka-topics.sh --create --zookeeper localhost:2181 --topic "listens" --partitions 1 --replication-factor 1

This should fix the issue. See this page for more info about described method: https://stackoverflow.com/a/24028480

diogocp/listenbrainz-server