Home Assistant - Monitor Performance of Docker Containers

Within my Home Assistant dashboard I wanted to see information about my Docker containers. For example CPU use, memory use and status (running/not running). After some research I came across this fork of the ‘Docker Monitor’ integration.

Monitoring Docker can also be done with Telegraf. The advantage of using Telegraf is that you can for example configure your own bucket within InfluxDB with a short retention and that many metrics are available. You can then use the data in Home Assistant with the help of the InfluxDB sensor

Mount the Docker Daemon

First make sure the Docker daemon is mounted within the Home Assistant Docker container. In summary, you need to add the following to your Docker run script. For me this is homeassistant_run.sh:

-v /var/run/docker.sock:/var/run/docker.sock \

Install the Docker Monitor Integration

The installation of the integration can be done via HACS or possibly via a git clone command. But I made the choice to install the integration with a simple copy job.

Go to the config directory of your Home Assistant Docker container:

cd config

Then make sure the custom_components/docker_monitor directory exists:

mkdir -p custom_components/docker_monitor

Go to the Docker Monitor directory and download the files (raw format!):

# Go to the docker monitor directory
cd custom_components/docker_monitor
# Download the files
wget https://raw.githubusercontent.com/guillaumelamirand/docker-monitor/master/custom_components/docker_monitor/__init__.py && wget https://raw.githubusercontent.com/guillaumelamirand/docker-monitor/master/custom_components/docker_monitor/const.py && wget https://raw.githubusercontent.com/guillaumelamirand/docker-monitor/master/custom_components/docker_monitor/manifest.json && wget https://raw.githubusercontent.com/guillaumelamirand/docker-monitor/master/custom_components/docker_monitor/sensor.py

Exit Nano (CTRL-X) and save the changes.

Configuration

Now go back to the config directory of your Home Assistant Docker container and edit configuration.yaml:

sudo nano configuration.yaml

Add the docker_monitor configuration:

# Example configuration.yaml entry
docker_monitor:
  containers:
    - homeassistant
    - mariadb
    - influxdb
  monitored_conditions:
    - docker_version
    - container_status
    - container_memory_usage
    - container_memory_percentage_usage
    - container_cpu_percentage_usage

Adjust the following:

containers:
Add your own Docker container names

monitored_conditions:
Add your own monitored conditions

Exit Nano (CTRL-X) and save the changes.

Groups (optional)

Personally, I found it useful to group the sensors by the name of the app.

Add the following to the configuration.yaml:

group: !include groups.yaml

Create a groups.yaml in the config directory:

sudo nano groups.yaml

Now add the groups, for example for the homeassistant, mariadb and influxdb containers:

app_homeassistant:
  name: Home Assistant
  entities:
    - binary_sensor.updater
    - sensor.docker_monitor_homeassistant_cpu_use
    - sensor.docker_monitor_homeassistant_memory_use
    - sensor.docker_monitor_homeassistant_memory_use_percent
    - sensor.docker_monitor_homeassistant_status
app_mariadb:
  name: MariaDB
  entities:
    - sensor.docker_monitor_mariadb_cpu_use
    - sensor.docker_monitor_mariadb_memory_use
    - sensor.docker_monitor_mariadb_memory_use_percent
    - sensor.docker_monitor_mariadb_status
app_influxdb:
  name: InfluxDB
  entities:
    - sensor.docker_monitor_influxdb_cpu_use
    - sensor.docker_monitor_influxdb_memory_use
    - sensor.docker_monitor_influxdb_memory_use_percent
    - sensor.docker_monitor_influxdb_status

I also added the binary_sensor.updater to the Home Assistant group.

Exit Nano (CTRL-X) and save the changes.

Using Docker Monitor

Now go to the Home Assistant UI and check your configuration via Configuration > Settings > Server Controls > Check Configuration (the button).

When the configuration is valid restart the server. After restarting you should find your sensors (and groups if configured) in the Overview when this dashboard is still maintained by Home Assistant (which is default):

Read other notes