The title says it all.
Simple - because there’s nothing easy about this. At all. Implementing HA (high availability) in a way that works reliably and seamlessly, is very complex. How are you going to handle hardware like Zigbee or ZWave sticks ?
3 Likes
Hi. I run my zigbee stick in a remote Raspberry Pi that communicates to the HA server via mqtt, for this very reason. That part was easy to implement. Being able to easily point HA instances to common databases and config files would need to be part of any eventual solution.
1 Like
I would use this. I have a completely redundant system so in the event that I kick one down a flight of stairs, I can have another one up in about 15 mins but I’d much rather it be immediately available…or without me. All my backups are on NAS in case of HA failure. I have a duplicate set of sticks.
What happens if that Zigbee Pi fails ?
Does that have a failover system?
1 Like
The zigbee specification does have the ability to have multiple coordinators with one being able to take control if the other goes down. It’s never been implemented to my knowledge. I have asked about this before.
Z-Wave may have something as well. Again never been implemented likely because it’s not that simple.
The zigbee gadgets would cease working. To me, a few sensors going down is way more preferable to the whole system going down. In my setup, zigbee sensors are just 2% of my iot sensors.
Here’s the post about the zigbee specification.
I have also talked about this with puddly before on a zha issue GitHub post.
1 Like