Hey. Just got my first couple of BalenaSound devices running. I’m attempting to run it on very underpowered hardware (Raspberry Pi Zero W). Volumio worked very well on the same hardware so I had high hopes for BalenaSound. Unfortunately, it seems that the CPU is being maxed out (mostly by the Node.js process) and the multiroom-client, multiroom-server, and sound-supervisor services have trouble staying in the running state.
I was able to connect to it via Bluetooth and play some music, but it was stuttering quite badly.
I’m hoping that I can configure my fleet to just not bother to start up the Spotify or AirPlay services since I’ll never use them. I found this other post from a couple of years ago, but I think I might need more of a walkthrough since the only docker-compose.yml file I can find lists only bluetooth.
Thanks!
Hi Steve, yes, you are correct…the Pi Zero is really starting to struggle with recent releases of balenaSound. Unfortunately, it just doesn’t have a lot of compute power. It is great for some projects, but, our implementation of containers, streaming, and audio are becoming too much. 
I haven’t tested the project if it will work properly without building those Airplay and Spotify services, so I don’t know if that will actually end up crashing the whole thing…but you can cut them out and see what happens, sure! Do a local git clone of the balenaSound GitHub repo, then delete the Airplay and Spotify services beginning here: balena-sound/docker-compose.yml at master · balenalabs/balena-sound · GitHub
Then, with the balenaCLI installed, do a balena push YourFleetNameYouCreated. It will take a bit to build in the cloud, then can take up to an hour for the Pi Zero to download and install the containers. After that, give it a try!
Awesome, I’ll give this a try.
Thanks!