Faust-Example Docker
Example running Faust with Docker and dynamic configurations.
Read more about Faust here: https://github.com/robinhood/faust
Configurable...Configs (explained poorly)
By configuring different entrypoints in setup.py, you can define config classes that hold important settings you need
to use in your Faust processor.
e.g. in this example, we have 2 classes defined by 2 entrypoints, video and channel. By passing one of those to the
Faust cli as an environment variable, we can then load that config class in our app.py code and get access to whichever necessary settings.
Example: CONFIG=video faust -A example.app worker -lINFO (see run.sh for how it's used with Docker.)
Then in app.py we call a special function load_config that uses the pkg_resources library to load this class that we defined
as an entrypoint in setup.py. Simply initialize that class that's returned and you have access to whatever settings you need.
Extending Docker:
There's an example Dockerfile here that defines some environment variables toward the end, and defines run.sh as the
command that gets called with docker run.
After building the docker image, you can specify which config class to load like so:
docker run -e WORKER=example.app -e WORKER_PORT=6066 -e CONFIG=video faust-example:{tag}