angelnu/ccu2 repository overview

⁠docker-ccu2

Script to create a docker container with the CCU2 firmware on the raspberry

This script will download the CCU2 firmware for the Homematic, re-package it as docker image and start it on your raspi. Other ARM-based boards might also work (see Depencies section).

If you want to skip the build and just download the ready to use docker container please visit this⁠.

⁠Dependencies

• ARM HW. Following combinations tested:
  • raspberry (tested on raspberry 3) and raspberrian jessy
  • Orange Pi Plus 2 and Armbian (Ubuntu 16.04 and Kernel 4.9) - Self built
• no programs using ports 80 and 2001

⁠How to build

1. ssh into your ARM device
2. sudo -i
3. git clone this repository
4. (Optional) cp -a settings.template settings and edit it
5. ./build.sh
6. (optional) Delete the cloned repo: the docker image and its data are stored in /var/lib/docker

#How to deploy

1. ssh into your ARM device
2. sudo -i
3. git clone this repository
4. (Optional) cp -a settings.template settings and edit it
5. ./deploy.sh
6. (optional) Delete the cloned repo: the docker image and its data are stored in /var/lib/docker

After the above steps you can connect to . The CCU2 docker image will be started automatically on every boot of the raspberry: the container is started in auto-restart mode.

⁠How to update to a new CCU2 firmware

Just update the CCU2_VERSION field in settings and execute ./build.sh && ./deploy.sh. Your CCU2 settings will be preserved.

⁠How to configure local antena

If you have added a homematic radio module⁠ to your raspberry, then you can use it with your CCU2 docker image. If not, you would need to use the external LAN gateway. You can also use a combination of both.

⁠Intructions

1. (raspberry 3) You need to avoid that the bluetoth module uses the HW UART. You can either disable it or let it use the miniUART. I use the second option but since I do not use Bluetoth on the Raspi I do not know if this breaks it. More info here⁠.
   • Add dtoverlay=pi3-miniuart-btto /boot/config.txt
2. (raspberry) Make sure that the serial console is not using the UART
   • Replace console=ttyAMA0,115200 with console=tty1 in /boot/cmdline.txt
   • More info here⁠
3. Run again ./deploy.sh

⁠How to import settings from an existing CCU2

1. Enable ssh in your CCU2. Instructions (in German) here⁠
2. ssh into your rapberry
3. sudo -i
4. ./undeploy.sh
5. Copy the old rfd.conf: cp /var/lib/docker/volumes/ccu2_data/_data/etc/config/rfd.conf{,org}
6. rsync -av \[your CCU2 IP\]/usr/local/* /var/lib/docker/volumes/ccu2_data/_data/
7. Diff the orginal rfd.conf with the one you copied from the CCU2: diff -u /var/lib/docker/volumes/ccu2_data/_data/etc/config/rfd.conf{,org}
8. Make sure you keep the original lines for:   * #Improved Coprocessor Initialization = true - commented out   * AccessFile = /dev/null - notice the blank at the start of the line   * ResetFile = /dev/ccu2-ic200 - notice the blank at the start of the line
9. ./deploy.sh

⁠Swarm

You can also deploy this docker image to a docker swarm. For this you need to:

• set up a docker swarm in multiple Raspberries. They all need to have local antennas at least you use a LAN gateway
• have a shared folder mounted at the same location in all the members of the cluster. Examples:
  • Mount a NAS folder. This is simple but then the NAS is the single point of failure
  • Cluster FS such as glusterfs. TBD: upload instructions

1. Change settings parameters
   • Set DOCKER_CCU2_DATA to the absolute path of your shared folder. Example: _/media/glusterfs/ccu2_
   • Set DOCKER_MODE to swarm
   • Set DOCKER_OPTIONS to --constraint node.labels.architecture==arm
2. ./deploy.sh