I really like the idea of parallelizing my "build all" function with the "pairing" function since it takes so long from the Home Assistant add-on and I have several machines on my network which could take up some of the slack. However, I'm struggling with how best to deploy a set of esphome-device-builder setups.
Ideally, this would be a docker container which can be redeployed when needing to update but my searching has only uncovered instructions like:
Use the existing ESPHome container
Override the entrypoint to have it install and use esphome-device-builder
I wonder if there's a full docker deployment of this new fancy esphome-device-builder all ready to go so that this two-step thing can be skipped. Maybe that's where the devs are going, eventually?
Best to do it the way above and follow the official releases, rather than from unofficial ones.
An alternative way is to build your own docker image above the official one, adding the needed steps.
Shouldn't be too much of a hassle if you are already managing your own docker environment.
Excellent! Thank you, I hadn't found that post, myself.
Is there some way to use the esphome-device-builder to include the localhost (running Home Assistant) in the round robin of assigning out firmware build tasks? My HA machine isn't a Pi so it can do some work but I want to spread the love across my network...
The current implementation of pairing seems to provide two options:
ALL local builds - no remote
ALL remote builds - no local
It would be nice to have another option allowing all paired hosts AND the local host to be in the pool of compilers.