My goal is to eliminate single points of failure. I currently have Home Assistant and Zwave-JS-UI running on virtual machines in a high-availability cluster, connected to my Z-Wave controller (Zooz ZST10) over IP using ser2net.
I have a second ZST10 that I would like to use as a redundant or hot spare radio in the event the primary ZST10 experiences an issue. Has anyone done something like this? I’ve searched this and other forums and can’t find much that is relevant.
Thanks!
Because the Z-wave controller holds all the configuration in NVM storage achieving this degree of availability isn’t easily achievable. You would have to sync configurations of the controllers regularly, and you couldn’t have them both online at the same time as I believe I’ve seen statements that Home Assistant doesn’t yet support primary (SIS) and secondary (SIC/SUC) controllers. Ideally that’s the architecture that Z-wave is designed to support for redundancy. Not sure why this isn’t supported, if it’s a Z-wave-JS driver issue, or what.
You best supported path at this point is to keep a backup copy of the NVM store and a spare identical controller and restore the NVM and bring the spare controller online after a failure. Or move to a system that supports redundant controllers, like HomeSeer. Even then there will be some period of failure until the failover is detected and failover occurs.
Thanks, I was afraid that might be the answer.
I want explore having a cold spare, that could be powered up and grab the latest NVM store before going live. Do you happen to know if there is any way to script updating/restoring the NVM on the controller through Zwave-JS-UI?
Configuration-based backups: https://zwave-js.github.io/zwave-js-ui/#/usage/setup?id=backup
Scripting can only be done via a) driver function (via cron, UI or MQTT API), b) MQTT API, or c) server websocket API.