Tasmota Plug Monitoring

This repository contains the necessary knowledge and resources to setup and monitor energy consumption measured by Tasmota-driven smart plugs using InfluxDB, Grafana, and MQTT.

The setup consists of the following components:

• InfluxDB as a time series database to store the metrics.
• Grafana as a visualization tool to create dashboards and panels.
• Mosquitto as a MQTT broker to receive the metrics from the Tasmota plugs.
• A custom script that listens to the MQTT broker and writes the metrics to the InfluxDB.

The dashboard is designed to automatically detect new devices and include them in the visualization.

Grafana dashboard of collected data

Requirements

• Docker
• Docker Compose
• Tasmota plug(s), e.g. https://www.amazon.de/dp/B0CHMMKZCQ

Setup and Configuration

1. Connect the Tasmota smart plug(s) to your network by following the instructions provided by the manufacturer. You can find the IP address of the plug by checking the DHCP leases of your router. You may also want to assign a static IP address to the plug(s).

2. Install the latest version of Tasmota on your devices.

3. Create a password file for mosquitto by running the official docker image to leverage the mosquitto_passwd command.

   docker run -it eclipse-mosquitto:latest sh
   $ mosquitto_passwd -c passwd_file mqtt_user
   Password: ********
   Reenter password: ********
   
   $ cat passwd_file
   mqtt_user:................

   Copy the content of passwd_file into mosquitto/passwd_file on your host system.

   (Do not change the path to passwd_file file within mosquitto/mosquitto.conf!)

4. Run the docker-compose file to start the services, but first, ensure that the required environment variables are set.

   The host must match the IP address of the machine running the Docker containers. For simplicity, we take the same password for MQTT and InfluxDB.

   Note: The MQTT password must match the password used when generating the passwd_file.

   The InfluxDB token must be created in the InfluxDB web UI in the next step.

   # InfluxDB
   export INFLUXDB_HOST="http://192.168.2.141:8086"
   export INFLUXDB_USERNAME="influxdb_user"
   export INFLUXDB_PASSWORD="influxdb_password"
   export INFLUXDB_DATABASE="smart_home"
   export INFLUXDB_ORG="smart_home"
   export INFLUXDB_TOKEN="You have to create a token in the influxdb IU"
   
   # MQTT
   export MQTT_HOST="192.168.2.141"
   export MQTT_USERNAME="mqtt_user"
   export MQTT_PASSWORD="${INFLUXDB_PASSWORD}"

   docker compose up

5. Open the InfluxDB web UI at http://192.168.2.141:8086/, login using the credentials and create a new API token at "Load Data" -> "API Tokens".

6. Stop the services started using Docker Compose and update the INFLUXDB_TOKEN with the newly created token.

   docker compose down
   export INFLUXDB_TOKEN="<the new token>"
   docker compose up -d

7. Setup the Tasmota devices to use the MQTT broker. The IP address must match the host of the MQTT broker. The username and password must match the MQTT_USERNAME and MQTT_PASSWORD environment variables. For this go to the IP address of the Tasmota smart plugs, go to "Configuration" -> "Configure MQTT" and enter the values. You may also want to rename the "Topic" to distinguish the devices. You can verify the connection by reviewing the logs of the MQTT service container and the Tasmota smart plug logs. Additionally, you can use tools like mqtt-explorer to verify the connection.

8. Increase the frequency of publishing metrics by setting TelePeriod 10 in the Tasmota console ("Tools" -> "Console"). This will ensure that the metrics are updated every 10 seconds.

9. Login to the Grafana web UI at http://192.168.2.141:3000/ using the default credentials admin and admin. You can change the password in the settings. Add the InfluxDB as a data source at "Connections" -> "Data Sources".

10. Now its time to import the Dashboard provided in this repository via copy-pasting the content of the grafana/dashboard.json file into the "Dashboards" -> "New" -> "Import" section of the Grafana UI.

11. You may have to adjust the price_per_kWh variable in some of the panels to match your electricity price.

Notes

• This is a basic setup and can be extended with more features like alerts, additional metrics, and more.
• The setup is designed to be run on a local network and is not secured for public access. You may want to add more security features like SSL certificates, further authentication, and authorization.
• The setup is tested and running on Ubuntu 24.04. It should work on other systems as well.
• For any questions, issues, or feature requests please open an issue in this repository.