This is an old uni project extended to become a Loki prototype setup. I have modified the readme to this use case. The old README.md can be found in the root as README.md.old
Make sure you have the antenna connected before running the scripts, otherwise it can seriously harm the LoPy4 Lora chip Main.py and Boot.py are MicroPython scripts to be run on a LoPy4 development board with the PyTrack Extension Board.
The scripts connects to the LoraWAN network The Things Network via MQTT Mosquitto.
The server consists of:
- mysql MySQL image with preloaded database schema
- mqtt MQTT is a lightweight message-queue protocol used for IoT devices due to low overhead.
- grafana Grafana is used for visualizing metrics and data
- loki Grafana Loki with Loki Canary instance. Loki ingests logs Promtail, indexes and stores the logs for querying by Grafana
- promtail Promtail takes logs from filesystem and passes the raw log changes to Loki for indexing and long term storage.
- prometheus Prometheus is used to collect metrics from services. Pull-based approach.
- alertmanager AlertManager is used to alert teammembers about issues identified throgh Loki or Prometheus
- bot The bot image is used to generate traffic, to simulate an IoT device.
All the components are included in the docker-compose file.
From within the server directory, start the server with
docker-compose pull
docker-compose up
In the server directory you will find four folders for building the docker images.
For convinience i included a script build_docker_images.sh
The server is intended to be used with docker-compose up and docker-compose down
The file contains a volume for grafana and mysql, so those containers have persistent storage
Due to a bug in the DigitalOcean vagrant plugin, create a docker droplet 18.04 with hostname coffeeserver manually.
After this, the current directory (server) can be synced to the digital ocean droplet with vagrant provision
Find Grafana Dashboard here http://localhost:3000/
default user is and pass admin
Look for Loki Logs and Coffee Dashboad under Dashboards -> Manage Dashboards
Find Prometheus instance here http://localhost:9090/
Find AlertManager instance here http://localhost:9093/
Find Promtail instance here http://localhost:8080/