openHAB Docker Containers

Table of Contents

• openHAB Docker Containers
  • Introduction
  • Image variants
  • Usage
    • Starting with Docker named volumes (for beginners)
      • Running from command line
      • Running from compose-file.yml
    • Running openHAB with libpcap support
    • Running on Windows and macOS
    • Starting with Docker mounting a host directory (for advanced user)
    • Automating Docker setup using ansible (for advanced user)
    • Accessing the console
    • Startup modes
  • Environment variables
    • User and group identifiers
    • Java cryptographic strength policy
  • Parameters
    • Passing devices with symlinks
  • Executing shell scripts before openHAB is started
  • Upgrading
  • Common problems
  • Building the images
  • Contributing

Introduction

Repository for building Docker containers for openHAB (Home Automation Server). Comments, suggestions and contributions are welcome!

Docker Image

Image variants

• For specific versions use:

  • openhab/openhab:<version>
  • openhab/openhab:<version>-<distribution>

• For the latest stable release use:

  • openhab/openhab
  • openhab/openhab:latest
  • openhab/openhab:latest-<distribution>

• For the latest release that has a milestone or stable maturity use:

  • openhab/openhab:milestone
  • openhab/openhab:milestone-<distribution>

• For the latest snapshot release use:

  • openhab/openhab:snapshot
  • openhab/openhab:snapshot-<distribution>

Versions:

• Stable: Thoroughly tested semi-annual official releases of openHAB. Use the stable version for your production environment if you do not need the latest enhancements and prefer a robust system.
  • 3.1.0 (Release notes)
  • 3.0.2 (Release notes)
• Snapshot: Usually 1 or 2 days old and include the latest code. Use these for testing out very recent changes using the latest code. Be aware that some snapshots might be unstable so use these in production at your own risk!
  • 3.2.0-snapshot

Distributions:

• debian for Debian 10 "buster" (default when not specified in tag) (Dockerfile)
• alpine for Alpine 3.13 (Dockerfile)

The Alpine images are substantially smaller than the Debian images but may be less compatible because OpenJDK is used (see Prerequisites for known disadvantages).

If you are unsure about what your needs are, you probably want to use openhab/openhab:3.1.0.

Prebuilt Docker Images can be found here: Docker Images

Platforms:

The following Docker platforms are supported (automatically determined):

• linux/amd64
• linux/arm64
• linux/arm/v7

There is no linux/arm/v7 Alpine image for openHAB 3 because the openjdk11 package is unavailable for this platform.

Usage

Important: To be able to use UPnP for discovery the container needs to be started with --net=host.

Important: In the container openHAB runs with user "openhab" (id 9001) by default. See user configuration section below!

The following will run openHAB in demo mode on the host machine:

docker run --name openhab --net=host openhab/openhab:3.1.0

NOTE: Although this is the simplest method to getting openHAB up and running, but it is not the preferred method. To properly run the container, please specify a host volume for the directories.

Starting with Docker named volumes (for beginners)

Following configuration uses Docker named data volumes. These volumes will survive, if you delete or upgrade your container. It is a good starting point for beginners. The volumes are created in the Docker volume directory. You can use docker inspect openhab to locate the directories (e.g. /var/lib/docker/volumes) on your host system. For more information visit Manage data in containers.

Running from command line

docker run \
  --name openhab \
  --net=host \
  -v /etc/localtime:/etc/localtime:ro \
  -v /etc/timezone:/etc/timezone:ro \
  -v openhab_addons:/openhab/addons \
  -v openhab_conf:/openhab/conf \
  -v openhab_userdata:/openhab/userdata \
  -e "EXTRA_JAVA_OPTS=-Duser.timezone=Europe/Berlin" \
  -d \
  --restart=always \
  openhab/openhab:3.1.0

Running from compose-file.yml

Create the following docker-compose.yml for use of local directories and start the container with docker-compose up -d

version: '2.2'

services:
  openhab:
    image: "openhab/openhab:3.1.0"
    restart: always
    network_mode: host
    volumes:
      - "/etc/localtime:/etc/localtime:ro"
      - "/etc/timezone:/etc/timezone:ro"
      - "./openhab_addons:/openhab/addons"
      - "./openhab_conf:/openhab/conf"
      - "./openhab_userdata:/openhab/userdata"
    environment:
      OPENHAB_HTTP_PORT: "8080"
      OPENHAB_HTTPS_PORT: "8443"
      EXTRA_JAVA_OPTS: "-Duser.timezone=Europe/Berlin"

Create the following docker-compose.yml for use of Docker volumes and start the container with docker-compose up -d

version: '2.2'

services:
  openhab:
    image: "openhab/openhab:3.1.0"
    restart: always
    network_mode: host
    volumes:
      - "/etc/localtime:/etc/localtime:ro"
      - "/etc/timezone:/etc/timezone:ro"
      - "openhab_addons:/openhab/addons"
      - "openhab_conf:/openhab/conf"
      - "openhab_userdata:/openhab/userdata"
    environment:
      OPENHAB_HTTP_PORT: "8080"
      OPENHAB_HTTPS_PORT: "8443"
      EXTRA_JAVA_OPTS: "-Duser.timezone=Europe/Berlin"

volumes:
  openhab_addons:
    driver: local
  openhab_conf:
    driver: local
  openhab_userdata:
    driver: local

Running openHAB with libpcap support

You can run all openHAB images with libpcap support. This enables you to use the Amazon Dashbutton Binding in the Docker container. For that feature to work correctly, you need to run the image as root user. Create the following docker-compose.yml and start the container with docker-compose up -d

version: '2.2'

services:
  openhab:
    container_name: openhab
    image: "openhab/openhab:3.1.0"
    restart: always
    network_mode: host
    cap_add:
      - NET_ADMIN
      - NET_RAW
    volumes:
      - "/etc/localtime:/etc/localtime:ro"
      - "/etc/timezone:/etc/timezone:ro"
      - "./openhab_addons:/openhab/addons"
      - "./openhab_conf:/openhab/conf"
      - "./openhab_userdata:/openhab/userdata"
    # The command node is very important. It overrides
    # the "gosu openhab tini -s ./start.sh" command from Dockerfile and runs as root!
    command: "tini -s ./start.sh server"

If you could provide a method to run libpcap support in user mode please open a pull request.

Running on Windows and macOS

The host networking driver only works on Linux hosts, and is not supported by Docker on Windows and macOS. On Windows and macOS ports should be exposed by adding port options to commands (-p 8080) or by adding a ports section to docker-compose.yml.

version: '2.2'

services:
  openhab:
    image: "openhab/openhab:3.1.0"
    restart: always
    ports:
      - "8080:8080"
      - "8443:8443"
    volumes:
      - "./openhab_addons:/openhab/addons"
      - "./openhab_conf:/openhab/conf"
      - "./openhab_userdata:/openhab/userdata"
    environment:
      OPENHAB_HTTP_PORT: "8080"
      OPENHAB_HTTPS_PORT: "8443"
      EXTRA_JAVA_OPTS: "-Duser.timezone=Europe/Berlin"

Starting with Docker mounting a host directory (for advanced user)

You can mount a local host directory to store your configuration files. If you followed the beginners guide, you do not need to read this section. The following run command will create the folders and copy the initial configuration files for you.

docker run \
  --name openhab \
  --net=host \
  -v /etc/localtime:/etc/localtime:ro \
  -v /etc/timezone:/etc/timezone:ro \
  -v /opt/openhab/addons:/openhab/addons \
  -v /opt/openhab/conf:/openhab/conf \
  -v /opt/openhab/userdata:/openhab/userdata \
  -e "EXTRA_JAVA_OPTS=-Duser.timezone=Europe/Berlin" \
  openhab/openhab:3.1.0

Automating Docker setup using Ansible (for advanced user)

Here is an example playbook in case you control your environment with Ansible. You can test it by running ansible-playbook -i mycontainerhost, -t openhab run-containers.yml. The :Z at the end of volume lines is for SELinux systems. If run elsewhere, replace it with ro.

- name: ensure containers are running
  hosts: all
  tasks:

  - name: ensure openhab is up
    tags: openhab
    docker_container:
      name: openhab
      image: openhab/openhab:3.1.0
      state: started
      detach: yes
      interactive: yes
      tty: yes
      ports:
        - 8080:8080
        - 8101:8101
        - 5007:5007
      volumes:
        - /etc/localtime:/etc/localtime:ro
        - /etc/timezone:/etc/timezone:ro
        - /opt/openhab/addons:/openhab/addons:Z
        - /opt/openhab/conf:/openhab/conf:Z
        - /opt/openhab/userdata:/openhab/userdata:Z
      keep_volumes: yes
      hostname: openhab.localnet
      memory: 512m
      pull: true
      restart_policy: unless-stopped
      env:
        EXTRA_JAVA_OPTS="-Duser.timezone=Europe/Berlin"

Accessing the console

You can connect to a console of an already running openHAB container with following command:

docker exec -it openhab /openhab/runtime/bin/client

The default password for the login is: habopen

Startup modes

Server mode

The container starts openHAB in server mode when no TTY is provided, example:

docker run --detach --name openhab openhab/openhab:3.1.0

When the container runs in server mode you can also add a console logger so it prints logging to stdout so you can check the logging of a container named "openhab" with:

docker logs openhab

To use a console logger with openHAB 2.x, edit userdata/etc/org.ops4j.pax.logging.cfg and then:

• Update the appenderRefs line to: log4j2.rootLogger.appenderRefs = out, osgi, console
• Add the following line: log4j2.rootLogger.appenderRef.console.ref = STDOUT

To use a console logger with openHAB 3.x, edit userdata/etc/log4j2.xml and add the following appender to the "Root logger" and "openhab.event" logger configurations: <AppenderRef ref="STDOUT"/>

Regular mode

When a TTY is provided openHAB is started with an interactive console, e.g.:

docker run -it openhab/openhab:3.1.0

Debug mode

The debug mode is started with the command:

docker run -it openhab/openhab:3.1.0 ./start_debug.sh

Environment variables

• EXTRA_JAVA_OPTS=""
• EXTRA_SHELL_OPTS=""
• LC_ALL=en_US.UTF-8
• LANG=en_US.UTF-8
• LANGUAGE=en_US.UTF-8
• OPENHAB_HTTP_PORT=8080
• OPENHAB_HTTPS_PORT=8443
• USER_ID=9001
• GROUP_ID=9001
• CRYPTO_POLICY=limited

User and group identifiers

Group id will default to the same value as the user id. By default the openHAB user in the container is running with:

• uid=9001(openhab) gid=9001(openhab) groups=9001(openhab)

Make sure that either

• You create the same user with the same uid and gid on your Docker host system

groupadd -g 9001 openhab
useradd -u 9001 -g openhab -r -s /sbin/nologin openhab
usermod -a -G openhab myownuser

• Or run the Docker container with your own user AND passing the userid to openHAB through env

docker run \
(...)
-e USER_ID=<myownuserid> \
-e GROUP_ID=<myowngroupid> \
(...)

You can obtain your user and group ID by executing the id --user and id --group commands.

Java cryptographic strength policy

Due to local laws and export restrictions the containers use Java with a limited cryptographic strength policy. Some openHAB functionality may depend on unlimited strength which can be enabled by configuring the environment variable CRYPTO_POLICY=unlimited

Before enabling this make sure this is allowed by local laws and you agree with the applicable license and terms:

• Debian: Zulu (Cryptography Extension Kit)
• Alpine: OpenJDK (Cryptographic Cautions)

The following addons are known to depend on the unlimited cryptographic strength policy:

• Eclipse IoT Market
• KM200 binding
• Loxone binding
• MQTT binding

Parameters

• -it - starts openHAB with an interactive console (since openHAB 2.0.0)
• -p 8080 - the HTTP port of the web interface
• -p 8443 - the HTTPS port of the web interface
• -p 8101 - the SSH port of the Console (since openHAB 2.0.0)
• -p 5007 - the LSP port for validating rules (since openHAB 2.2.0)
• -v /openhab/addons - custom openHAB addons
• -v /openhab/conf - openHAB configs
• -v /openhab/userdata - openHAB userdata directory
• --device=/dev/ttyUSB0 - attach your devices like RFXCOM or Z-Wave Sticks to the container

Passing devices with symlinks

On Linux, if you pass a device with a symlink or any non standard name (e.g. /dev/ttyZWave), some addons require the device name to follow the Linux serial port naming rules (e.g. "ttyACM0", "ttyUSB0" or "ttyUSB-9999") or will otherwise fail to discover the device.

This can be achieved by mapping the devices to a compliant name like this:

docker run \
(...)
--device=/dev/ttyZWave:/dev/ttyACM0
--device=/dev/ttyZigbee:/dev/ttyACM1
(...)

More information about serial ports and symlinks can be found here.

Executing shell scripts before openHAB is started

It is sometimes useful to run shell scripts after the "userdata" directory is created, but before Karaf itself is launched. One such case is creating SSH host keys, and allowing access to the system from the outside via SSH. Exemplary scripts can be found in the contrib/cont-init.d directory

To use this, create a directory called

/etc/cont-init.d

and add a volume mount to your startup:

  ...
  -v /etc/cont-init.d:/etc/cont-init.d \
  ...

and put your scripts into that directory. This can be done by either using a volume mount (see the examples above) or creating your own images which inherit from the official ones.

Upgrading

Upgrading OH requires changes to the user mapped in userdata folder. The container will perform these steps automatically when it detects that the userdata/etc/version.properties is different from the version in dist/userdata/etc/version.properties in the Docker image.

The steps performed are:

• Create a userdata/backup folder if one does not exist.
• Create a full backup of userdata as a dated tar file saved to userdata/backup. The userdata/backup folder is excluded from this backup.
• Show update notes and warnings.
• Execute update pre/post commands.
• Copy userdata system files from dist/userdata/etc to userdata/etc.
• Update KAR files in addons.
• Delete the contents of userdata/cache and userdata/tmp.

The steps performed are the same as those performed by running the upgrade script that comes with OH, except the backup is performed differently and the latest openHAB runtime is not downloaded. All messages shown during the update are also logged to userdata/logs/update.log.

Common problems

Error: KARAF_ETC is not valid

This error message indicates that the data in the volumes is not properly initialized. The error is usually the result of a permissions issue or already putting files into volumes without the volumes having been initialized by the container first. When mounting directories into the container, check that these directories exist, are completely empty and are owned by the same USER_ID and GROUP_ID as configured in the container ENV variables.

Missing some preinstalled package

Docker containers are kept as small as possible intentionally to decrease download times and the number of potential vulnerabilities for everyone. If you want additional packages installed, create your own container based on this container. Another option is to install the package by executing a shell script before openHAB is started.

No logging after "Launching the openHAB runtime..."

By default this will always be the last logged message. A console logger can be configured for more detailed logging.

OpenMediaVault

The default filesystem mount flags of OpenMediaVault contain the noexec flag which interferes with the serial library used by openHAB. To be able to use serial devices with openHAB, make sure the userdata volume mounted by the container is not backed by a filesystem having the noexec flag. See the OMV documentation on how to remove the noexec flag from an existing filesystem.

Portainer

The default values of ENV variables are always stored by Portainer (see portainer/portainer#2952). This causes issues such as endless restart loops when upgrading the container with Portainer. To resolve this issue when upgrading openHAB, first remove all default (non-overridden) ENV variables before starting the new container.

SELinux

When using the container on a Linux distribution with SELinux enabled (CentOS/Fedora/RHEL), add the :z or :Z option to volumes to give the container write permissions. For more information on this see the Docker documentation.

Building the images

Checkout the GitHub repository, change to a directory containing a Dockerfile (e.g. /debian) and then run these commands to build and run a Docker image for your current platform:

$ docker build --build-arg JAVA_VERSION=11 --build-arg OPENHAB_VERSION=3.1.0 --tag openhab/openhab .
$ docker run openhab/openhab

To be able to build the same image for other platforms (e.g. arm/v7, arm64 on amd64) Docker CE 19.03 with BuildKit support can be used.

First enable BuildKit support, configure QEMU binary formats and a builder using:

$ echo '{"experimental":true}' | sudo tee /etc/docker/daemon.json
$ export DOCKER_CLI_EXPERIMENTAL=enabled
$ docker run --rm --privileged docker/binfmt:a7996909642ee92942dcd6cff44b9b95f08dad64
$ sudo systemctl restart docker
$ docker buildx create --name builder --use

Change to a directory containing a Dockerfile (e.g. /debian) and then use the following command to build an ARMv7 image:

$ docker buildx build --build-arg JAVA_VERSION=11 --build-arg OPENHAB_VERSION=3.1.0 --platform linux/arm/v7 --tag openhab/openhab --load .

The build script in the root of the repository helps to simplify building the openHAB images with BuildKit. It can be used to build the images of multiple openHAB versions and correctly tag and push them to a Docker registry. Execute ./build -h for usage instructions and examples.

Contributing

Contribution guidelines