This might be a big one and for the future, but in my eyes also an important one.
If my Home Assistant goes down, I’m pretty screwed. Even with backups on my network storage, to get/replace hardware and restore configuration, it needs time, energy and know how. If I can’t to that right away, it sucks. In certain situations alot. After years, so many things do run on HA for me, and even with backup strategies (every light can still be controlled without HA), many things will stop working.
So I can’t stop thinking about, what if I could just pre-configure a Raspi or a VM and it would take over if the main host stops working?
E.g. a offline Raspi: it knows the location of my backups and the main host ip, I plug it in and it just fires up, restores config, send me a notification and possible migration steps (e.g. move my zigbee stick to the new host), and within a minute it’s back up and running.
E.g. a more complex online Raspi: Raspi runs in slave mode, gets updated config, automations, states etc automatically synced. If main HA instance gets unavailable and IP is also unoccupied, just take over, if IP is still blocked by the main HA OS (without a running HA) shut itself down and notify slave. Then send me
a notification and possible migration steps.
Down the road, at a point where there is nothing like directly plugged in hardware (or technical solutions to work around that) which would require a migration step, it could even take over just short reboots. No downtime when updating? Hell yeah.
I can only imagine this would be hell of a challenge. To get everything right and really reliable. With all the different configs, underlying OS and hardware.
But if accomplished, what the heck, how cool would that be?!