Install & maintain openHAB on Raspi in Docker together with Mosquitto, Duplicati, influxDB, Samba etc - Step by step instruction

[Cplant]

I started with openHAB with the fabulous openHABian, but am now since ~1,5 years running openHAB on a Raspi 3 B+ in a Docker environment. I like this setup, because it can be set up easily (e.g. in case of hardware failure) and I can install and try out new software with limited effort.

Below is my step-by-step instruction which I wrote for myself with the help of a friend, to a) set up the system initially or b) restore my system from my daily Duplicati-Backups in case of a failure. Last time I tried it out, getting from a “blank SD card” to a “restored installation based on last night’s last Duplicati backup” took me around 45-60 minutes.

Disclaimer 1: I’m not a pro, but I got this running / it works. In case I missed something / something can be improved, let me know.
Disclaimer 2: As you can read in this thread, running all of the below on Docker requires more than 1 GB of RAM. It also runs with just 1 GB of RAM, but Raspbian then starts to swap, which reduces speed and lifetime of the SD card.

The following softwares are included in my instruction in addition to openHAB:

• Portainer (for easier Docker Maintenance)
• influxDB (as a data base for openHAB, instead of the default RRD4j)
• Duplicati (for daily backups of the entire system, in my case to Google Drive)
• Frontail (for log analysis of openHAB)
• Mosquitto (for MQTT-devices, e.g. all sorts of Tasmota devices)
• Samba (for easier access to the relevant files of the containers)
• wmbusmeters (to receive data sent by my water and heating meters via wireless M-Bus)

Step 1: Install Raspberry Pi OS

• Install the Raspberry Pi Imager on your PC (Linux or Windows) and launch it
• Chose “Raspberry Pi OS Lite (32-bit)” as the OS
• Under “Advanced Options” activate SSH…
• … set a username (in my case: pi) and password…
• and configure the wireless LAN (SSID & Password)
• Click “Write” to write the OS to the SD card
• When you’re done, plug your SD card in to the raspi and boot the raspi (no monitor, mouse or keyboard need to be connected)
• Afterwards, you should be able to reach the Raspi via a SSH software (in Linux: Terminal) by entering ssh pi@192.168.0.xxx followed by your password. If you don’t know the IP address which got assigned to the Raspi, check your router.

Step 2: Prepare Raspbian OS and install Docker

• Increase the Raspian OS swapfile. Without this, things were somewhat unstable for me under load. You can check this yourself when, after you’re done with everything, you run htop and will see that Raspbian starts to swap.
  Enter sudo nano /etc/dphys-swapfile and increasing the swap file to 1240 (CONF_SWAPSIZE=1240). Exit with Ctrl + X.
• Update the system by entering sudo apt-get update && sudo apt-get upgrade.
• Define the desired IP-address by entering sudo nano /etc/dhcpcd.conf. In the file include / uncomment

interface wlan0
static ip_address=192.168.0.xxx/24
static routers=192.168.0.1
static domain_name_servers=192.168.0.1

If you want to use the ethernet interface instead of wifi, write eth0 instead of wlan0. Exit with Ctrl + X.

• Perform a reboot via sudo reboot
• Install Docker by entering curl -sSL https://get.docker.com | sh
• Add the pi-user to the Docker Group by entering sudo usermod -aG docker pi
• Perform a reboot via sudo reboot
• Install Portainer by entering sudo docker pull portainer/portainer-ce:linux-arm
• Run and configure Portainer by entering sudo docker run -d -p 9000:9000 --name=portainer --restart=always -v /var/run/docker.sock:/var/run/docker.sock -v portainer_data:/data portainer/portainer-ce:linux-arm
• Access Portainer by opening a Browser and entering http://192.168.0.xxx:9000. Define your Portainer Password.
• In case you wonder: Portainer is a GUI for Docker, which makes it easier to work with Docker containers.

Step 3: Deploy the docker containers

• Access Portainer by opening a Browser and entering http://192.168.0.xxx:9000.
• Go to “home” → “local” → “Stacks”.
• Use the “Add Stack”-button to add the following Stack and deploy the respective containers defined in that stack.
• You can either paste the stack-definition code (below) into the Web Editor or (what I would recommend) save the code into local files and use the “Upload” function. Give the stack a meaningful name (I chose openhab as a stack name).

version: '3.8'

services:
  openhab:
    image: openhab/openhab:latest
    container_name: openhab
    depends_on:
      - influxdb
      - mosquitto
    restart: always
    privileged: true
    volumes:
      - "/etc/localtime:/etc/localtime:ro"
      - "/etc/timezone:/etc/timezone:ro"
      - data_openhab_conf:/openhab/conf
      - data_openhab_addons:/openhab/addons
      - data_openhab_userdata:/openhab/userdata
    environment:
      TZ: ${TZ}
      OPENHAB_HTTP_PORT: ${OPENHAB_HTTP_PORT}
      OPENHAB_HTTPS_PORT: ${OPENHAB_HTTPS_PORT}
      CRYPTO_POLICY: "unlimited"
      EXTRA_JAVA_OPTS: "-Duser.timezone=${TZ}"
      USER_ID: 999
      GROUP_ID: 994
      # Test to run openHAB with German number format
      # Can be tested by entering "locale" after "docker exec -t -i openhab /bin/bash"
      LANG: de_DE.UTF-8
      LANGUAGE: de_DE.UTF-8
      LC_ALL: de_DE.UTF-8
    devices:
      - /dev/ttyACM0:/dev/ttyACM0
    network_mode: host

  mosquitto:
    image: eclipse-mosquitto:latest
    container_name: mosquitto
    restart: always
    volumes:
      - data_mosquitto_config:/mosquitto/config
      - data_mosquitto_data:/mosquitto/data
      - data_mosquitto_log:/mosquitto/log
    ports:
      - ${MOSQUITTO_MQTT_PORT}:1883
      - ${MOSQUITTO_MQTTS_PORT}:8883
    environment:
      TZ: ${TZ}
    networks:
      app-nw:
        aliases:
          - mqtt

  influxdb:
    image: influxdb:1.8
    container_name: influxdb
    restart: always
    volumes:
      - data_influxdb:/var/lib/influxdb
      - data_influxdb_config:/etc/influxdb
    ports:
      - ${INFLUXDB_DB_PORT}:8086      
    environment:
      - INFLUXDB_DATA_ENGINE=tsm1
      - INFLUXDB_REPORTING_DISABLED=false
      - INFLUXDB_ADMIN_USER=${INFLUXDB_ADMIN_USER}
      - INFLUXDB_ADMIN_PASSWORD=${INFLUXDB_ADMIN_PASSWORD}
      - INFLUXDB_USER=${INFLUXDB_USER}
      - INFLUXDB_USER_PASSWORD=${INFLUXDB_USER_PASSWORD}
      - INFLUXDB_READ_USER=${INFLUXDB_READ_USER}
      - INFLUXDB_READ_USER_PASSWORD=${INFLUXDB_READ_USER_PASSWORD}
      - INFLUXDB_DB=${INFLUXDB_OPENHAB_DB}
      - TZ=${TZ}
    networks:
      app-nw:
      backend-nw:
        aliases:
          - db

  wmbusmeters:
    image: weetmuts/wmbusmeters
    container_name: wmbusmeters
    privileged: true
    restart: always
    volumes:
      - data_wmbusmeters_data:/wmbusmeters_data
      - /etc/localtime:/etc/localtime:ro
      - /dev/:/dev/
    networks:
      - app-nw
      - backend-nw

  duplicati:
    image: duplicati/duplicati
    container_name: duplicati
    environment:
      - PUID=0
      - PGID=0
      - TZ=Europe/Berlin
      - CLI_ARGS= #optional
    volumes:
      - /var/lib/docker/volumes/duplicati_config/_data:/config
      - /var/lib/docker/volumes/duplicati_backups/_data:/backups
      - /var/lib/docker/volumes/duplicati_source/_data:/source
      - /var/lib/docker/volumes/openhab_data_openhab_conf:/source/openhab_conf
      - /var/lib/docker/volumes/openhab_data_openhab_addons:/source/openhab_addons
      - /var/lib/docker/volumes/openhab_data_openhab_userdata:/source/openhab_userdata
      - /var/lib/docker/volumes/openhab_data_mosquitto_config:/source/mosquitto_config
      - /var/lib/docker/volumes/openhab_data_mosquitto_data:/source/mosquitto_data
      - /var/lib/docker/volumes/openhab_data_mosquitto_log:/source/mosquitto_log
      - /var/lib/docker/volumes/openhab_data_influxdb:/source/influxdb
      - /var/lib/docker/volumes/openhab_data_influxdb_config:/source/influxdb_config
      - /var/lib/docker/volumes/openhab_data_wmbusmeters_data:/source/wmbusmeters_data
    ports:
      - 8200:8200
    restart: unless-stopped

  frontail:
    image: welteki/frontail-openhab:latest
    container_name: frontail
    restart: always
    ports:
      - "9001:9001"
    volumes:
      - /var/lib/docker/volumes/openhab_data_openhab_userdata/_data:/openhab/userdata:ro \

  samba:
    image: dperson/samba
    container_name: samba
    restart: always
    ports:
      - "137:137/udp"
      - "138:138/udp"
      - "139:139/tcp"
      - "445:445/tcp"
    healthcheck:  
      disable: true  
    environment:
    - TZ='Europe/Berlin'
    - PUID=9001
    - PGID=9001  
    - WORKGROUP=workgroup  
    - PERMISSIONS=true
    - RECYCLE=false
    - USER1=pi;raspberry;
    - SHARE1=openhab_conf;/shares/openhab_conf;yes;no;yes;pi
    - SHARE2=openhab_addons;/shares/openhab_addons;yes;no;yes;pi
    - SHARE3=openhab_userdata;/shares/openhab_userdata;yes;no;yes;pi
    - SHARE4=mosquitto_config;/shares/mosquitto_config;yes;no;yes;pi
    - SHARE5=mosquitto_log;/shares/mosquitto_log;yes;no;yes;pi
    - SHARE6=wmbusmeters_data;/shares/wmbusmeters_data;yes;no;yes;pi
    - SHARE7=influxdb;/shares/influxdb;yes;no;yes;pi
    - SHARE8=docker;/shares/docker;yes;no;yes;pi
    - SHARE9=log;/shares/log;yes;no;yes;pi
    volumes:
      - data_openhab_conf:/shares/openhab_conf
      - data_openhab_addons:/shares/openhab_addons
      - data_openhab_userdata:/shares/openhab_userdata
      - data_mosquitto_config:/shares/mosquitto_config
      - data_mosquitto_log:/shares/mosquitto_log
      - data_wmbusmeters_data:/shares/wmbusmeters_data
      - data_influxdb:/shares/influxdb
      - /docker:/shares/docker
      - /var/log:/shares/log

  chronograf:
    image: chronograf:latest
    container_name: chronograf
    restart: always
    ports:
      - '8888:8888'
    volumes:
      - /var/lib/docker/volumes/openhab_data_influxdb/_data:/influxdb
    environment:
      - INFLUXDB_URL=http://db:8086
      - INFLUXDB_USERNAME=$INFLUXDB_USER
      - INFLUXDB_PASSWORD=$INFLUXDB_USER_PASSWORD
    networks:
      app-nw:
      backend-nw:

volumes:
  data_openhab_conf:
  data_openhab_addons:
  data_openhab_userdata:
  data_mosquitto_config:
  data_mosquitto_data:
  data_mosquitto_log:
  data_influxdb:
  data_influxdb_config:
  data_wmbusmeters_data:

networks:
  app-nw:
    internal: false
    driver: bridge
    driver_opts:
      com.docker.network.bridge.name: app-${PROJECT_NAME}
  backend-nw:
    internal: true
    driver: bridge
    driver_opts:
      com.docker.network.bridge.name: be-${PROJECT_NAME}

Include the following enviroment variables (define your own passwords first) by saving the following code into a local .env-file on your computre and select this file by clicking on “Load variables from .env file”. Make sure that you define the respective passwords for you (<password>).

# Config File for openHAB Application

# Docker Compose Project Name
# max length 11 characters
PROJECT_NAME=openhab3

# InfluxDB Configuration
INFLUXDB_ADMIN_USER=admin
INFLUXDB_ADMIN_PASSWORD=<password>
INFLUXDB_USER=openhab
INFLUXDB_USER_PASSWORD=<password>
INFLUXDB_READ_USER=openhab_ro
INFLUXDB_READ_USER_PASSWORD=<password>
INFLUXDB_OPENHAB_DB=openhab_db

# Port Configuration openHab
OPENHAB_HTTP_PORT=8080
OPENHAB_HTTPS_PORT=8443

# Port Configuration InfluxDB
INFLUXDB_DB_PORT=8086

# Port Configuration Grafana
GRAFANA_HTTP_PORT=3000

# Port Configuration Chronograf
CHRONOGRAF_HTTP_PORT=8888

# Port Configuration Mosquitto
MOSQUITTO_MQTT_PORT=1883
MOSQUITTO_MQTTS_PORT=8883

# Timezone
TZ=Europe/Berlin

Deploy the respective containers by clicking “Deploy the stack”. This will take some time (10 - 15 minutes) an should end with the message “Stack deployed”.

Step 4: 1st launch of openHAB

• Perform a reboot via sudo reboot
• Launch openHAB for the first time by accessing it in a browser via 192.168.0.xxx:8080
• If everything worked correctly, you should be able to access and configure openHAB.

Important first step before doing anything else: Configure influxDB

1. Install the influxDB-Binding (Settings → Bindings → influxDB Persistence)
2. Activate influxDB (Settings → Persistence → inxluxDB Persistence Layer)
3. Configure influxDB (Settings → influxDB Persistence Layer): Database URL: http://localhost:8086; Database version: influxDB 1; username and password: as defined above).

If influxDB version 2 is to be chosen, also influxDB 2 has to be deployed (see docker-compose.yml file: image: influxdb:1.8)

Step 5: Configure Duplicati

This step is not needed but highly recommended to run daily backups. Without this step, if your SD card breaks, everything is gone. Duplicati allows you to backup everything (e.g. every night) to e.g. Google Drive).

• Launch Duplicati by accessing it in a browser via 192.168.0.xxx:8200.
• Create a new backup (in my case scheduled daily) which saves all essential folders of your installation (combined in the/source/ folder, see pre-configuration above) to a destination of your choice (in my case Google Drive). So in a worst case scenario, you can fall back to last night’s state of your entire installation (in case of daily backups).
• If you want to backup your data to a thumb drive (which I would definitely not recommend because a) if it’s always plugged in, it can break alongside the Raspi in case of an electrical failure or b) if it is not always plugged in, then there’s no point in having Duplicati run for regular updates) make the thumb drive accessible to the Duplicati container similarly than the Wireless M-BUS USB Stick is made available to the openHAB container, by including the following section into the Duplicati stack, with the respective address of the thumb drive (perform ls in /dev/ and search for your thumb drive). The part on the left is the physical address as seen to the Raspbian, the part on the right is the one that will be visible to Duplicati in its container.

    devices:
      - /dev/ttyACM0:/dev/ttyACM0

• Export your configuration (Configuration → Export) and save it locally, so that you still have your Duplicati configuration in case your SD card fails.

Step 6 (optional): Configure Mosquitto

This step is only needed if you want to hook up devices to openHAB via MQTT (e.g. Tasmota devices or wmbusmeters). The configuration is text-only, but if you’ve followed the docker-compose.yml you should see a share mosquitto_config if you open the samba share via smb://192.168.0.xxx.

1) Prepare mosquitto.conf

1. Open the mosquitto.conf file by accessing the samba share smb://192.168.0.150/mosquitto_config/ (either via Terminal or via your normal PC).
2. Uncomment (or include if not already in there by default) the following lines

# =================================================================
# Listeners
# =================================================================
listener 1883
listener 8883
# =================================================================
# Persistence
# =================================================================
persistence true
persistence_location /mosquitto/data
# =================================================================
# Logging
# =================================================================
log_dest stderr
# =================================================================
# Security
# =================================================================
allow_anonymous false
# -----------------------------------------------------------------
# Default authentication and topic access control
# -----------------------------------------------------------------
password_file /mosquitto/config/mosquitto.passwd

Save the file.

2) Define username and password

1. Create a file mosquitto.passwd in the same drive smb://192.168.0.148/mosquitto_config/
2. Your file should contain your username and your password in the following format:
   
3. Log into your mosquitto-container via Terminal → docker exec -it mosquitto /bin/sh
4. Convert the plain-txt-password into a hashed version via mosquitto_passwd -U mosquitto.passwd
   With the example above, the file now looks like this:
   
   

Afterwards the openHAB MQTT-binding can be installed (Settings → Bindings → MQTT Binding). Afterwards, in openHAB an MQTT Broker-Thing can be created. The following settings are proven to work (most of them should be default, though).

UID: mqtt:broker:7519737e0f
label: MQTT Broker
thingTypeUID: mqtt:broker
configuration:
  lwtQos: 0
  publickeypin: true
  clientID: openhab
  keepAlive: 60
  hostnameValidated: true
  secure: false
  birthRetain: true
  certificatepin: true
  shutdownRetain: false
  password: <yourpassword>
  protocol: TCP
  qos: 0
  reconnectTime: 60000
  port: 1883
  mqttVersion: V3
  host: localhost
  lwtRetain: true
  enableDiscovery: true
  username: openhab

Done.

Step 7 (only relevant for restoring from a Duplicati backup, e.g. in case of hardware failure)

• Perform Steps 1 - 3 on a fresh SD card.
• Stop all containers except Duplicati
• Launch Duplicati by accessing it in a browser via 192.168.0.xxx:8200.
• Click “Add backup”
• Select “Import from file”. Select the Duplicati configuration file (.json) which you have stored locally. Uncheck the box “Import Metadata” and check the box “Save immediately”. Confirm with “Import”.
• Select “Repair” in the overlay (your Config file which you saved some time back does not match the backup anymore → this is normal). Recreating the database might take some minutes.
• Click “Restore” → then the last option (the name of your backup). Select the time and the folders of the backup which you want to restore. Check the boxes “Original Location”, “Overwrite” and “Restore read/write permissions” and press “Restore”. This might take some minutes.
• Afterwards (you should see no error messages), restart your containers in Portainer.
• If everything worked correctly, you should be able to access openHAB again from the state saved in your backup.

Maintenance

Update openHAB within Docker
Since the image is specified via image: openhab/openhab:latest, clicking on “Update the Stack” in the stack-definition and confirming “Re-pull image and redeploy” in the popup should do the trick and pull the latest version of openHAB. Make sure you do a backup before.

Alternatively, if to be done via Terminal

1. Terminal → ssh pi@192.168.0.xxx
2. docker stop openhab to stock openHAB
3. docker rm openhab to remove the current openHAB container
4. docker pull openhab/openhab:latest to pull the latest update of openHAB
5. In Docker re-deploy the openHAB stack

Update Portainer
To update Portainer (in contrast to Docker itself, Portainer is not updated automaticially via the regular sudo apt-get update && sudo apt-get upgrade) do the following:

1. Terminal → ssh pi@192.168.0.xxx
2. docker stop portainer to stop the Portainer container
3. docker rm portainer to remove the Portainer image
4. docker pull portainer/portainer-ce to pull the latest Portainer image
5. docker run -d -p 8000:8000 -p 9000:9000 --name=portainer --restart=always -v /var/run/docker.sock:/var/run/docker.sock -v portainer_data:/data portainer/portainer-ce to run the new version

Additional notes:

• I highly recommend to try the entire cycle (incl. the restore) once in a while on a second SD card, to see whether a) it runs through without problems and b) whether the backup & restore works correctly.
• As mentioned, in my case, this takes ~45 - 60 minutes (not too bad)
• As a test, I used this protocol to install everything on an x86/debian system. So far everything seems to work smoothly

Potential problems after upgrading from 3.x to 4.x

• If you encounter the following problem

2023-09-12 22:55:33.583 [WARN ] [org.apache.felix.fileinstall        ] - /usr/share/openhab/addons does not exist, please create it.
2023-09-12 22:55:33.590 [ERROR] [org.apache.felix.fileinstall        ] - Cannot create folder /var/lib/openhab/tmp/bundles. Is the folder write-protected?
2023-09-12 22:55:33.591 [ERROR] [org.apache.felix.configadmin        ] - [org.osgi.service.cm.ManagedServiceFactory, id=50, bundle=18/mvn:org.apache.felix/org.apache.felix.fileinstall/3.7.4]: Unexpected problem updating configuration org.apache.felix.fileinstall.4a5be290-c3c5-4491-a66c-0c0d3dbd032a
java.lang.RuntimeException: Cannot create folder: /var/lib/openhab/tmp/bundles
	at org.apache.felix.fileinstall.internal.DirectoryWatcher.prepareDir(DirectoryWatcher.java:647) ~[?:?]
	at org.apache.felix.fileinstall.internal.DirectoryWatcher.prepareTempDir(DirectoryWatcher.java:627) ~[?:?]
	at org.apache.felix.fileinstall.internal.DirectoryWatcher.<init>(DirectoryWatcher.java:179) ~[?:?]
	at org.apache.felix.fileinstall.internal.FileInstall.updated(FileInstall.java:250) ~[?:?]
	at org.apache.felix.fileinstall.internal.FileInstall$ConfigAdminSupport$Tracker.updated(FileInstall.java:380) ~[?:?]
	at org.apache.felix.cm.impl.helper.ManagedServiceFactoryTracker.updated(ManagedServiceFactoryTracker.java:159) ~[?:?]
	at org.apache.felix.cm.impl.helper.ManagedServiceFactoryTracker.provideConfiguration(ManagedServiceFactoryTracker.java:93) ~[?:?]
	at org.apache.felix.cm.impl.ConfigurationManager$ManagedServiceFactoryUpdate.provide(ConfigurationManager.java:1266) ~[?:?]
	at org.apache.felix.cm.impl.ConfigurationManager$ManagedServiceFactoryUpdate.run(ConfigurationManager.java:1210) ~[?:?]
	at org.apache.felix.cm.impl.UpdateThread.run0(UpdateThread.java:122) ~[?:?]
	at org.apache.felix.cm.impl.UpdateThread.run(UpdateThread.java:84) ~[?:?]
	at java.lang.Thread.run(Thread.java:833) ~[?:?]

refer to this post [openHAB Docker] "/usr/share/openhab/addons does not exist" after upgrade to OH 3.3.0 - #7 by wborn for a fix.

[monigrip (rob)]

This is great, much appreciated for the share.

[boris03 (Boris)]

Thanks a lot for that documentation!
Not easy to ask questions for that huge post
Let me try.
I would like to use that for a system based on debian not running on a Pi.
I also don’t need Duplicati nor Wmbusmeters.
I am not really familar with portainer and docker but both installed and running.
Below you can find some questions and comments to your description.
Would be great if you could check and answer.

Your Step 2

• Add the pi-user to the Docker Group by entering sudo usermod -aG docker pi
  Is that necessary, when Yes I assume I should use here my debian user instead right?

Your Step 3
When adding these stack definitions directly to the stacks section does it also pull the images?

Stack Duplicati
when not required I just ignore that, right?

Openhab
Clear I need to adjust User and Group_ID

Is this command below valid for all possible USB devices or just one specific device?
devices:

• /dev/ttyACM0:/dev/ttyACM0

Will that command below really automatically update OH, InfluxDB and Mosquitto?
It looks like this would need to be installed before, right?
when not required can I just delete these line with watchtower?
labels:
com.centurylinklabs.watchtower.enable=true

Stack wmbusmeters:
Cause I don’t need that I just would ignore this

volumes:

  data_openhab_userdata:
  data_influxdb:
  data_influxdb_config:
  data_grafana:                     -> Where is this coming from I think this isn't included?
  data_mosquitto_data:
  data_mosquitto_log:
  data_mosquitto_config:
  data_openhab_addons:
  data_openhab_conf:
  wmbusmeters_data:                    -> Assuming Not required for me when not used?
    driver: local                       > Assuming Not required for me?
    name: wmbusmeters_data             -> Assuming Not required for me?

This networks section will I use that as-is or do I need to adjust something?

networks:
  app-nw:
    internal: false
    driver: bridge
    driver_opts:
      com.docker.network.bridge.name: app-${PROJECT_NAME}
  backend-nw:
    internal: true
    driver: bridge
    driver_opts:
      com.docker.network.bridge.name: be-${PROJECT_NAME}

For the env-file you wrote:
Before deploying the openHAB stack make sure to include the following environment variables (define your own passwords first) by saving the following code into a local .env-file on your computre and select this file by clicking on “Load variables from .env file”:
→ What do you mean with define your own passwords?
You mean add passwords here and these will be just set by loading the env-file, correct?

Port Configuration in the env file:
Port Configuration Grafana → Required? Not clear cause not in use?
GRAFANA_HTTP_PORT=3000

Port Configuration Chronograf → Required? Not clear cause not in use?
CHRONOGRAF_HTTP_PORT=8888

With this configuration below Samba will run in a container right?
Would it not be easier to run Samba directly on Debian and just share the 3 Docker
Openhab folders?
Or is there an advantage for running Samba in a container?

Samba

version: '3'
 
services:
  samba:
    image: dperson/samba
    container_name: samba
    restart: always
    ports:
      - "137:137/udp"
      - "138:138/udp"
      - "139:139/tcp"
      - "445:445/tcp"
    healthcheck:  
      disable: true  
    environment:
    - TZ='Europe/Berlin'
    - PUID=9001
    - PGID=9001  
    - WORKGROUP=workgroup  
    - PERMISSIONS=true
    - RECYCLE=false
change   - USER1=pi;raspberry;   -> pi to be replaced by my debian user, what is raspberry here?
    - SHARE1=openhab_conf;/shares/openhab_conf;yes;no;yes;
    - SHARE2=openhab_userdata;/shares/openhab_userdata;yes;no;yes;
    - SHARE3=mosquitto_config;/shares/mosquitto_config;yes;no;yes;
    - SHARE4=mosquitto_log;/shares/mosquitto_log;yes;no;yes;
delete    - SHARE5=wmbusmeters_data;/shares/wmbusmeters_data;yes;no;yes; 
    - SHARE6=openhab_addons;/shares/openhab_addons;yes;no;yes;
    - SHARE7=docker;/shares/docker;yes;no;yes;
    - SHARE8=log;/shares/log;yes;no;yes;
delete    - SHARE9=duplicati_config;/shares/duplicati_config;yes;no;yes;
    - SHARE10=influxdb;/shares/influxdb;yes;no;yes;
    volumes:
      - openhab_data_openhab_conf:/shares/openhab_conf
      - openhab_data_mosquitto_log:/shares/mosquitto_log
      - openhab_data_mosquitto_config:/shares/mosquitto_config
      - openhab_data_openhab_userdata:/shares/openhab_userdata
      - openhab_data_openhab_addons:/shares/openhab_addons
delete      - wmbusmeters_data:/shares/wmbusmeters_data
      - /docker:/shares/docker
      - /var/log:/shares/log
delete      - duplicati_config:/shares/duplicati_config
      - openhab_data_influxdb:/shares/influxdb
      - samba_lib:/var/lib/samba
      - samba_run:/run/samba
      - samba_cache:/var/cache/samba
      - samba_etc:/etc
      - samba_log:/var/log/samba

networks:
  default:
volumes:
    openhab_data_openhab_conf:
        external: true
        name: openhab_data_openhab_conf
    openhab_data_mosquitto_log:
        external: true
        name: openhab_data_mosquitto_log
    openhab_data_mosquitto_config:
        external: true
        name: openhab_data_mosquitto_config
    openhab_data_openhab_userdata:
        external: true
        name: openhab_data_openhab_userdata
    openhab_data_openhab_addons:
        external: true
        name: openhab_data_openhab_addons
delete    wmbusmeters_data:
delete        external: true
delete        name: wmbusmeters_data
delete    duplicati_config:
delete        external: true
delete        name: duplicati_config
    openhab_data_influxdb:
        external: true
        name: openhab_data_influxdb
    samba_lib:
        name: samba_lib
    samba_run:
        name: samba_run
    samba_cache:
        name: samba_cache
    samba_etc:
        name: samba_etc
    samba_log:
        name: samba_log

[Cplant]

@boris03

Is that necessary, when Yes I assume I should use here my debian user instead right?

Not a pro myself (this part I took over from tutorials on the web, but I’d say yes.

@boris03

When adding these stack definitions directly to the stacks section does it also pull the images?

Yes

@boris03

when not required I just ignore that, right?

Correct.

@boris03

Is this command below valid for all possible USB devices or just one specific device?
devices:

• /dev/ttyACM0:/dev/ttyACM0

As always: it depends. If you want to use a USB device within docker, check with ls within /dev/ the address of your device.

@boris03

Will that command below really automatically update OH, InfluxDB and Mosquitto?
It looks like this would need to be installed before, right?
when not required can I just delete these line with watchtower?

What you write makes sense. And honest answer: I don’t know for sure (that part I got from another tutorial on the net).

@boris03

data_grafana: -> Where is this coming from I think this isn't included?

My bad. I also have Grafana installed but not included in my step-by-step.

@boris03

wmbusmeters_data:                    -> Assuming Not required for me when not used?
    driver: local                       > Assuming Not required for me?
    name: wmbusmeters_data             -> Assuming Not required for me?

True.

@boris03

This networks section will I use that as-is or do I need to adjust something?

Would also recommend to take as-is.

@boris03

What do you mean with define your own passwords?
You mean add passwords here and these will be just set by loading the env-file, correct?

Yes. Replace with the password of your choice.

@boris03

Port Configuration in the env file:
Port Configuration Grafana → Required? Not clear cause not in use?
GRAFANA_HTTP_PORT=3000

Port Configuration Chronograf → Required? Not clear cause not in use?
CHRONOGRAF_HTTP_PORT=8888

Correct. Not required. Though also doesn’t hurt, since these variables will not be grabbed by the stack.

@boris03

With this configuration below Samba will run in a container right?
Would it not be easier to run Samba directly on Debian and just share the 3 Docker
Openhab folders?
Or is there an advantage for running Samba in a container?

You might be right. If you have it running already, it should be able to do the job just right. I’m using samba as part of Docker because I’d like to start with a clean slate (fresh install of Raspi OS) and get everything running just via deploying the stacks, without doing anything else to the linux installation.

[Piffer (Tim)]

@Cplant

@boris03

Will that command below really automatically update OH, InfluxDB and Mosquitto?
It looks like this would need to be installed before, right?
when not required can I just delete these line with watchtower?

What you write makes sense. And honest answer: I don’t know for sure (that part I got from another tutorial on the net).

You have to have watchtower to be running. You can configure watchtower to use this label to know if it should update that container or not.

I don’t recommened unattended updates via watchtower, especially for openHAB as there might come changes, that break something and you should be aware before.

I personally use watchtower to let me know, if there are updates for my containers available.

[Cplant]

@Piffer

I personally use watchtower to let me know, if there are updates for my containers available.

Hmm, I never really thought about that part in my stack-code. Is watchtower a separate container which you can run and informs you via mail etc on available updates?

Or, the other way round, if you have the watchtower-code in the stack of one container, but watchtower is not running on your machine, the code basically does nothing?

[Piffer (Tim)]

Yes and yes. You can find the documentation here. Give it a try

[boris03 (Boris)]

Thanks, this helps me a lot!
Great to see and beginning to understand how Docker is working.

Would be even much better when we could get a kind of an official script
(or is it called YAML?) what could create an Openhabian on Docker?
Is that possible?

Maybe I just open a new post for that and see if others may have already something created for their own, like you started already with such an initiative here.

[Cplant]

@boris03

Would be even much better when we could get a kind of an official script
(or is it called YAML?) what could create an Openhabian on Docker?
Is that possible?

I think the idea makes sense. Would add another option “Docker” (advanced users) in addition to “openHABian” (beginners) and “linux installation” (experts). For the Docker-option there are several starts of this on the net, e.g. also here.

@boris03

Maybe I just open a new post for that and see if others may have already something created for their own, like you started already with such an initiative here.

Sounds good!

[stonke (Kristoph)]

Guys I stuck on stack no3. Have you seen such an error before? Stack has been written to file as per instruction.

[11194 (Вася Пупкин)]

@Cplant

You might be right. If you have it running already, it should be able to do the job just right. I’m using samba as part of Docker because I’d like to start with a clean slate (fresh install of Raspi OS) and get everything running just via deploying the stacks, without doing anything else to the linux installation.

Then samba part and users/folders creation should be done with Ansible.

[stonke (Kristoph)]

at the notification Tab have found that all of error notifications are finished with “no matching manifest for linux/arm/v7 in the manifest list entries”
what this is about? I did not modify any of steps.
My hardware is original rpi 3b+

[Cplant]

It is possible that somehow an error sneaked in my instructions. Will double-check this, but might take some time.

[monigrip (rob)]

Hi I deleted the wmbusmeters part, which fixed this, i didn’t need it, hope this helps.

[stonke (Kristoph)]

@Cplant

  wmbusmeters_data:
    driver: local
    name: wmbusmeters_data

Good point, did the same and stack deployed forward
unfortunately wmbusmeters is main reason why considered Docker so have to wait for Cplant help
hope zigbee2mqtt would be running with all of these stacks too.

[boris03 (Boris)]

You just mentioned zigbee2mqtt…
Is anyone having a working portainer custom template for zigbee2mqtt?

[Cplant]

@stonke

unfortunately wmbusmeters is main reason why considered Docker so have to wait for Cplant help

Thanks for your comments. You found something that would’ve probably come in my next firedrill as well).

Turned out we weren’t the only ones stumbling upon this (see here). Not absolutely sure why the platform now has to be defined, but adding platform: linux/armv7 is a quick fix.

Will also update the step-by-step (incl. the docker-compose.yml) above, but have to find some time to re-do the entire thing and test it once more. Hence the quick fix.

[monigrip (rob)]

I added zigbee2mqtt to the stack, was pretty straight forward.

  zigbee2mqtt_assistant:
    container_name: zigbee2mqtt_assistant
    image: carldebilly/zigbee2mqttassistant
    restart: unless-stopped
    ports:
    - 8880:80
    environment:
    - VIRTUAL_HOST=~^zigbee2mqtt_assistant\..*\.xip\.io
    - Z2MA_SETTINGS__MQTTSERVER=mosquitto
    - VIRTUAL_PORT=8880
    networks:
      app-nw:
        aliases:
          - mqtt

  zigbee2mqtt:
    container_name: zigbee2mqtt
    image: koenkk/zigbee2mqtt:1.26.0
    environment:
    - TZ=Etc/UTC
    ports:
    - 8085:8085
    volumes:
    - data_zigbee2mqtt:/app/data
    devices:
    - /dev/ttyUSB0:/dev/ttyACM0 # should work even if no adapter
   #- /dev/ttyACM0:/dev/ttyACM0 # should work if CC2531 connected
   #- /dev/ttyUSB0:/dev/ttyACM0 # Electrolama zig-a-zig-ah! (zzh!) maybe other as well
    restart: unless-stopped
    networks:
      app-nw:
        aliases:
          - mqtt

[Cplant]

I cleaned up the stack-file (included all relevant stacks into one) and edited the instruction for clarity.

In case of feedback, let me know

[stonke (Kristoph)]

@Cplant

http://localhost:8086;

The updated instructions seem to work, 1stack with 9containers installed successfully! I have easily access to portainer, duplicati or openhab via published ports.
But how do I get into wmbusmeters data and configuration? I don’t see a published port for this plugin, so I understand that the service is text-only, BUT after logging into the terminal via port :22, I see only 9 folders.
pi@raspberrypi:~ $ ls
Bookshelf Documents Music Public Videos
Desktop Downloads Pictures Templates