Topics

#custom-component #docker #hacs #home-assistant #homeassistant #python

Language breakdown

Python 100.0%

▸ Files click to expand

README

Custom Monitor Docker component for Home Assistant

About

This repository contains the Monitor Docker component I developed for monitoring my Docker environment from Home-Assistant. It is inspired by the Sander Huisman Docker Monitor, where I switched mainly from threads to asyncio and added my own wishes/functionality. Feel free to use the component and report bugs if you find them. If you want to contribute, please report a bug or pull request, and I will reply as soon as possible.

Monitor Docker

The Monitor Docker allows you to monitor Docker and container statistics and turn on/off containers. It can connect to the Docker daemon locally or remotely. When Home Assistant is used within a Docker container, the Docker daemon should be mounted as follows -v /var/run/docker.sock:/var/run/docker.sock.

Docker run Example

docker run -d \ ... \ -v /var/run/docker.sock:/var/run/docker.sock \   homeassistant/home-assistant

docker-compose.yaml Example

services:   hass:     image: homeassistant/home-assistant ...     volumes:       - /var/run/docker.sock:/var/run/docker.sock ...

NOTE: Making /var/run/docker.sock read-only has no effect because it is a socket (not a file).

Raspberry Pi (Raspbian)

Using a Raspberry Pi with Raspbian it could happen no memory is reported. In such case, the Docker API does not report it to Monitor Docker. Making the following changes normally fixes the problem:

Open the file /boot/cmdline.txt
Add the following to the end of the existing line cgroupenable=cpuset cgroupmemory=1 cgroup_enable=memory
Reboot your Raspberry Pi

NOTE: Add the line to the existing line, do not replace it

Ubuntu / Debian

Also on Ubuntu/Debian, no memory may be shown, the following changes could solve your problem:

Open the file /etc/default/grub
Modify the GRUBCMDLINELINUXDEFAULT="" to GRUBCMDLINELINUXDEFAULT="quiet cgroup_enable=memory swapaccount=1"
Run sudo update-grub
Reboot your Ubuntu/Debian

NOTE: This is untested, use at your own risk!

Installation

HACS - Recommended

Have HACS installed, this will allow you to easily manage and track updates.
Search for 'Monitor Docker'.
Click Install below the found integration.
Configure using the configuration instructions below.
If applicable, add the volume /var/run/docker.sock to your Home Assistant container.
Restart Home-Assistant.

Manual

Copy directory customcomponents/monitordocker to your <config dir>/custom_components directory.
Configure with the config below.
If applicable, add the volume /var/run/docker.sock to your Home Assistant container.
Restart Home-Assistant.

Configuration

To use the monitor_docker in your installation, add the following to your configuration.yaml file:

# Example configuration.yaml entry
monitor_docker:
  - name: Docker
    containers:
      - appdaemon
      - db-dsmr
      - db-hass
      - deconz
      - dsmr
      - hass
      - influxdb
      - mosquitto
      - nodered
      - unifi
    rename:
      appdaemon: AppDaemon
      db-dsmr: "Database DSMR-Reader"
      db-hass: Database Home Assistant
      deconz: DeCONZ
      dsmr: "DSMR-Reader"
      hass: Home Assistant
      influxdb: InfluxDB
      mosquitto: Mosquitto
      nodered: "Node-RED"
      unifi: UniFi
    monitored_conditions:
      - version
      - containers_running
      - containers_total
      - state
      - status
      - memory

Important NOTE: The rename functionality works with regular expression. If you got containers with roughly the same name, it could match the wrong one. Examples:

appdaemon: AppDaemon - Will match anything with "appdaemon" 
^appdaemon: AppDaemon - Will match if it starts with "appdaemon" 
^appdaemon$: AppDaemon - Only match if it exactly matches "appdaemon", thus "appdaemon-2" will not match

Configuration variables

| Parameter | Type | Description | | --------------------------- | -------------------------- | --------------------------------------------------------------------- | | name | string (Required) | Client name of Docker daemon. Defaults to Docker. | | url | string (Optional) | Host URL of Docker daemon. Defaults to unix://var/run/docker.sock. Remote Docker daemon via TCP socket is also supported, use e.g. http://ip:2375. Do NOT add a slash add the end, this will invalidate the URL. For TLS support see the Q&A section. SSH is not supported. | | scaninterval | timeperiod (Optional) | Update interval. Defaults to 10 seconds. | | retry | time_period (Optional) | Retry interval when a TCP error is detected. Defaults to 60 seconds. | | certpath | string (Optional) | If a TCP socket is used, you can define your Docker certificate path, forcing Monitor Docker to enable TLS. The filenames must be ca.pem, cert.pem and key.pem| | containers | list (Optional) | Array of containers to monitor. Defaults to all containers. | | containers_exclude | list (Optional) | Array of containers to be excluded from monitoring, when all containers are included. | | monitored_conditions | list (Optional) | Array of conditions to be monitored. Defaults to all conditions. | | rename | dictionary (Optional) | Dictionary of containers to rename. Renaming is done on the name in HA Lovelace, not the entity name (see rename_entity). Default no renaming. | | rename_entity | boolean (Optional) | If rename is enabled, it changes the name in HA Lovelace, not the entity name. Enable this setting to also rename the entity name (Default: False) | | sensorname | string (Optional) | Sensor string to format the name used in Home Assistant. Defaults to {name} {sensor}, where {name} is the container name and {sensor} is e.g. Memory, Status, Network speed Up | | switchname | string (Optional) | Switch string to format the name used in Home Assistant. Defaults to {name}, where {name} is the container name. | | switchenabled | boolean / list (Optional) | Enable/Disable the switch entity for containers (Default: True Enabled switch for all containers, False: Disabled switch for all containers). Or specify a list of containers for which to enable switch entities. | | buttonenabled | boolean (Optional) | Enable/Disable the button entity for containers (Default: False Enabled button for all containers, False: Disabled button for all containers). Or specify a list of containers for which to enable button entities. | | precision_cpu | integer (Optional) | Precision of CPU usage percentage (Default: 2) | | precisionmemorymb | integer (Optional) | Precision of memory usage in MB (Default: 2) | | precisionmemorypercentage | integer (Optional) | Precision of memory usage in percentage (Default: 2) | | precisionnetworkkb | integer (Optional) | Precision of network bandwidth in kB (Default: 2) | | precisionnetworkmb | integer (Optional) | Precision of network usage in MB (Default: 2) |

| Monitored Conditions | Description | Unit | | --------------------------------- | ------------------------------- | ----- | | version | Docker version | - | | containers_total | Total number of containers | - | | containers_running | Number of running containers | - | | containers_paused | Number of paused containers | - | | containers_stopped | Number of stopped containers | - | | containerscpupercentage | CPU Usage. The CPU usage depends on the number of CPU cores, e.g. if you have 8 cores, this value can have a maximum of 800% | % | | containers1cpupercentage | CPU Usage, between 0-100% | % | | containers_memory | Memory usage | MB | | containersmemorypercentage | Memory usage | % | | images | Number of images | - | | state | Container state. This is created, restarting, running, removing, paused, exited or dead | - | | status | Container status. E.g. Up 13 days, Up 5 hours, Exited (0) 11 hours ago | - | | health | Container health if available | - | | uptime | Container start time | - | | image | Container image | - | | cpu_percentage | CPU usage. The CPU usage depends on the number of CPU cores, e.g. if you have 8 cores, this value can have a maximum of 800% | % | | 1cpu_percentage | CPU Usage, between 0-100% | % | | memory | Memory usage | MB | | memory_percentage | Memory usage | % | | networkspeedup | Network speed upstream. Not available when using network mode is 'host' | kB/s | | networkspeeddown | Network speed downstream. Not available when using network mode is 'host' | kB/s | | networktotalup | Network total upstream. Not available when using network mode is 'host' | MB | | networktotaldown | Network total downstream. Not available when using network mode is 'host' | MB | | allinone | This is a special condition and when used, it will only create 1 sensor per container with all the monitored conditions as attribute value. NOTE: If you use this sensor, all other sensors are NOT created, just 1 sensor |- |

Debugging

It is possible to debug the Monitor Docker component, this can be done by adding the following lines to the configuration.yaml file:

logger:
  logs:
    customcomponents.monitordocker: debug

Q&A

Here are some possible questions/errors with their answers.

Question: Does this integration work with the HASS or supervisord installers?

Answer: Yes, with an external docker container. Home Assistant supervised does not expose the Docker UNIX/TCP socket. However, you can use an external docker container named docker-socket-proxy. Start this docker with the following docker-compose code. It exposes the socket over TCP and monitor_docker can listen to it.

# Proxy the Docker sock so that we can pick up stats for HomeAssistant
    services:
      dockerproxy:
        image: tecnativa/docker-socket-proxy
        container_name: dockerproxy
        privileged: true
        volumes:
          - /var/run/docker.sock:/var/run/docker.sock
        ports:
          - 2375:2375
        environment:
          - BUILD=1
          - COMMIT=1
          - CONFIGS=1
          - CONTAINERS=1
          - DISTRIBUTION=1
          - EXEC=1
          - IMAGES=1
          - INFO=1
          - NETWORKS=1
          - NODES=1
          - PLUGINS=1
          - SERVICES=1
          - SESSSION=1
          - SWARM=1
          - POST=1

Add the following to your configuration.yaml:

monitor_docker:
      - name: Docker
        url: http://<host_ip>:2375

Error: Missing valid dockerhost.Either DOCKERHOST or local sockets are not available.

Answer: Most likely the socket is not mounted properly in your Home Assistant container. Please check if you added the volume /var/run/docker.sock

Error: aiodocker.exceptions.DockerError: DockerError(900, "Cannot connect to Docker Engine via http://10.0.0.1:2376...).

Answer: You are trying to connect via TCP and most likely the remote address is unavailable. Test it with the command docker -H tcp://10.0.0.1:2376 ps if it works (replace 10.0.0.1 with your IP address). Also you can consult the following URL for more help: https://docs.docker.com/config/daemon/remote-access/

Question: Is Docker TCP socket via TLS supported?

Answer: Yes it is. You need to set the URL to e.g. tcp://ip:2376 and the per instance configurationg certpath need to be set The following is a docker-compose example of how to set the environment variables and the volume with the certificates:

services:
  hass:
    image: homeassistant/home-assistant
...
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock
      # The files need to be named "ca.pem", "cert.pem" and "key.pem"
      - ./certs:/certs
...

Question: Can this integration monitor 2 or more Docker instances?

Answer: Yes it can. Just duplicate the entries and give it an unique name and define the url as shown below:

# Example configuration.yaml entry
monitor_docker:
  - name: Docker
    containers:
    ...
  - name: RemoteDocker
    url: tcp://10.0.0.1:2376
    containers:
    ...

Question: Can create, delete or re-create a container be implemented in the integration?

Answer: The used Docker library has no easy (and safe) way to handle such functionality. Please use docker-compose to handle such operations. If anybody can make this fully work in a safe way, I'll be happy to merge the PR

Question: Can you add more security to a switch?

Answer: No, this isn't possible from the integration. You need to do this directly in Lovelace itself, within the card e.g. https://github.com/iantrich/restriction-card

Question: All the reported memory values are 0 (zero), can this be fixed in the integration?

Answer: No, the integration just uses the available information from the API and you should fix your Docker

Question: Is it possible to monitor HASS.IO?

Answer: Yes, please use the Docker Socker Proxy https://github.com/Tecnativa/docker-socket-proxy and configure http://ip:port to connect to the proxy. This has been tested and verified by other users, but I cannot give support on it.

Question: I get a permission denied error?

Answer: In general Docker and HASS.IO are running as root and always can connect to /var/run/docker.sock. If you run in a venv environment or directly with Python, you may need to add the "docker" group to the user used for Home Assistant. The following commands may help you, and it is recommended to reboot after "usermod":

$ sudo usermod -a -G docker <user>
  $ sudo reboot

Question: Can you add the feature to check if there are updates to images in e.g. hub.docker.com?

Answer: Such feature goes outside of the scope of monitor_docker and there are few other options available for this. You can use https://newreleases.io or https://github.com/crazy-max/diun/

Question: Is Docker via SSH supported?

Answer: No, the Docker library used, does not support it. There is a small but, maybe you can get it to work via socat. The following URL may help you: https://serverfault.com/questions/127794/forward-local-port-or-socket-file-to-remote-socket-file/362833#362833

Question: Can the sensors have unique entity identifiers? This is useful for renaming it in the HA GUI

Answer: This is not possible, due to the nature of how this integration works. The docker name needs to be consistent across restart and recreate, this can be only done by overruling the entity identifier as it is working now

Credits

Sanderhuisman

License

Apache License 2.0

🔗 More in this category