I’ve been researching the options, reading articles and asking questions in these forums. Purpose of this thread is to document the different steps and decisions I’ve gone through.
First, my setup. I run home assistant on a Synology NAS running DSM 6.x using Docker. I like using docker as it gives me flexibility in running multiple containers, different instance of HA. I find this useful when developing packages and automations to do it on a test instance rather than monkeying around on the prod system.
Overtime I’ve made most of my config modular in packages and have different package templates for different devices - such that if I deploy a new temp / humidity sensor - I can automate getting the sensor up with high / low alarm limits, alerts, etc. I use shell scripts and sed to automate the generation. It’s nice, since if I change a rule or automation it’s then easy to redeploy. But I digress….
I know that migrations are rarely easy. For example, humans migrated to North America across the bering land bridge. As a lifelong dev, I’ve been through many migrations….
So, one of my main goals is to do the migration in steps, on my own time, identify / mitigate the inevitable issues and then have a seamless prod cutover when I’m ready. Rather than doing an epic 18 hour upgrade and end up tossing the NAS down the stairs .
Since I’m on docker, running the zwavejs2mqtt docker container is the way to go.
Based on some excellent advice from @finity I found out that if you Stop the Zwave (deprecated) Network in HA that releases the stick to be accessible by zwavejs2mqtt. Likewise If you shutdown the zwavejs2mqtt container that releases the stick and you can restart the zwave network in HA. So with this approach I can take zwave outages on the prod system to work on the zwave-js and then switch back when I’m done with the current task or need a break.
Speaking of breaks I need one right now. Next post will be on how to get zwavejs2mqtt setup on Synology docker.