Home Assistant installation in Mac mini 2014: docker vs direct install


#1

I was running hass.io in a raspberry 2, but it has already bricked me 2 micro sd after running for around 3 months each. I have a Mac mini that is running Plex 24/7 so I might as well run Home Assistant in it.

What are the pros and cons of installing directly to the Mac (pip3 install homeassistant) vs installing through docker vs any other method???

Other considerations: I am using an Aeotec Z-Stick (USB) to access Z-wave device. Can the configuration be restored simply by copying the files to the config directory?


#2

I use docker for a few services on one server, but I run HA directly on a pi B+.

The advantage of docker is that it creates a self contained environment that is easy to upgrade and remove from your system, and won’t be corrupted by other changes made to your system.

The disadvantage is that it is a learning curve to using docker, and integrating hardware devices can be tricky.

If you are willing to spend a little time learning about docker, it is certainly worthwhile and I would recommend using that method, provided you can get the usb device to work. I have no experience with z-wave so can’t help there.

If you do go the direct install route, I would recommend creating a separate user for HA, with restricted access for security. And also creating a virtual environment, to avoid changes to the system python libraries affecting your HA install.

In either case, you can still copy your configuration yaml files across, but I don’t know about z-wave configuration.


#3

Yes you can pass a usb stick through to docker. What OS are you running on the mac mini?

On Ubuntu you can follow the hass docs for configuring a z-wave stick to find out where it is on /dev/tty*.

Then for docker you pass that through with a device bind. so for docker-compose file it looks like:

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

Where /dev/ttyACM0 is where the z-wave stick shows up. you then add that do your config under the z-wave component just like you should already have.

And yes copying all of the config directory over to where you mount your volumes on the host and bind to the hass docker container will just work as far as z-wave is concerned. I think the main one needed is the z-wave.xml file. I just moved from a venv install on an Rpi3 to docker on a NUC.