Hello,
I have successfully tested home assistant as a systemd service on centos, straight from the box and as a docker container and all good, lovely bit of software.
Now I am wondering what if the best method? Does one method have better benefits than another?
Whats your thoughts? and Why if you chose a particular installation method way why did you choose that way?
Why raspberry pi? Because it will be consuming much less power when you compare it a workstation or a laptop. Also, it has some digital and analog outputs if you want to use them.
Regarding to installation methods (independent of hardware) there is a page which shows pros/cons of each one.
I agree with @finity and run the docker install also for the same reasons. I have an old Mac mini that is running Ubuntu and a Shinobi NVR already, so none of the other install methods would work for me unless I went and created a virtual machine somehow. Docker is much more light weight and a VM would bog the system down. The biggest downside is you can’t use add-ons with the container version, but you usually can install them separately as their own docker containers. That is a bit more work but if you already know docker or are willing to learn it I think it’s the way to go. The other install methods manage the OS for you and have a supervisor that manages and installs the add-ons, so that type of install is probably easier for someone starting out. The Home Assistant OS/Supervisor install creates more of an “appliance” device that just runs home assistant, but then you can’t really do anything else on the device outside of running Home Assistant and supported and approved add-ons. The container install gives you the most flexibility and you really could run Home Assistant on any piece of hardware you want running any version of linux that way, without trying to navigate and worry about the mess described here related to a supervised install - architecture/0014-home-assistant-supervised.md at dca73b25652e903a0ac2fe1099d43a7b138ed05b · home-assistant/architecture · GitHub or figure out how to deal with an inefficient VM.