This is a python script, installed into Docker, which facilitates scraping Arris SB6183 cablemodems for their relevant RF data and surfacing that to prometheus.
There are three methods one could use to run this software:
- Run the script on a machine locally
- Run the docker container on a docker host
- Deploy to Kubernetes via Helm
One can install this app by running pip install . in the source directory. If you're curious about prerequisites, check setup.py. Once installed, executing py_arris_exporter will start a python webserver on port 9393. Any request uri to the service will return a set of prometheus metrics.
You can also run this app as a docker container by executing docker run -d -p9393:9393 petergrace/arris_exporter which will spawn a container and daemonize, exposing the port on your local system at port 9393. This is better if you don't want to commingle your python environment installs.
If you are leveraging kubernetes, there is a helm chart under the helm/ subdirectory of this repository where you can deploy petergrace/arris_exporter directly to your kubernetes cluster. It will create a service called arris-exporter on a ClusterIP that you can then target with your prometheus install.
Here's a snippet of my scrape config for prometheus inside of my kubernetes environment:
- job_name: 'arris'
static_configs:
- targets: ['arris-exporter:9393']
As you can see, it doesn't have many bells or whistles; it just reports the data I wanted to see to prometheus.
In the grafana/ subdirectory, I have included my grafana dashboard json that I use to query the datapoints from prometheus, to make the screenshot shown above.