During my last holiday the cleaning lady killed my HASS by mistake (broke the power supply). This was the second time that during a time of longer absence I had issues with my automation setup. Normally all runs super stable but as Murphy would agree, obviously these things happen if you are not in.
Although in the meantime if have migrated to a NUC and put this in a much safer place, this made me think about a redundant setup to be used in case of total failure during my absence. The idea is simple:
Have a fully operational PI connected -but powered down- ready on the network. It seems the zwave stick can be backed up (here) and restored on a fresh stick. Also that stick (with the current zwave loaded via the restore) would already be in the PI.
If in case my main HASS has total failure, I power up the backup one (incl. ‘cloned’ zwave stick’ ). Powering up can be either by a zwave switch connected to my Vera (which for several reasons is also still up and running) or via POE (which I can switch in/off via my Unifi setup). I always have snapshots stored on my windows server so even those I could restore remotely to the ‘backup PI’ if needed. In theory I can even automate this.
Am I over simplifying this or do you think this is indeed a valid way to get this done?