"Reinterviewing" all nodes after HA reboots

[Dunlapjc]

Forgive me if I'm using terminology wrong in here, and feel free to correct me.

1. My Z-Wave network is working, even though Z-wave js UI dashboard looks like the screenshot below
2. But after restoring my HA instance from a backup (backup/restore included restoring ZWave JS UI) my network now comes up with spinning circles on every node (except for the controller)
3. I can reinterview nodes (powered nodes are easier, obviously) and then I get the standard green check box
4. I can get the whole grid back to green checkboxes, but one update and a restart later and I'm back to a grid of spinning circles

I can't for the life of me figure out what's going on, what the impact may be, or how to correct this.
Any help, or ideas of what's going on?

After reboot:

After re-interviewing:

Thanks for your insight and your help.
-Jonathan

[robertsLando]

Please make a driver log, loglevel debug and attach it here as a file (drag & drop into the text field). Be sure to restart your application so we can see what's blocking the interview. Also be sure you are running latest 9.30.0 version

[Dunlapjc]

Daniel,
Was not expecting you to answer! I figured someone in the community has experienced this before and would chime in.
Running latest version(s) of HA:
Core. 2025.1.4
Supervisor. 2024.12.3
Operating System 14.2
Frontend. 20250109.2
Z-Wave JS UI: 3.19.0

Here are two versions of the log file done after rebooting:
A debug version and a "silly" level version.
zwave-js-ui-store -Debug 28-01-25.zip
zwave-js-ui-store -Silly 28-01-25.zip

If that's not the helpful file, please flag me and I'll correct.

Thanks so much!
-Jonathan

[robertsLando]

@AlCalzone Any clue?

[AlCalzone]

Could be that this is the same issue as zwave-js/zwave-js#7559, which is fixed in driver version 14.3.8 that I just released.

[Dunlapjc]

Will do and will report back!
Thanks for the attention and thought on this.
(I had thought about rebuilding from scratch, but some nodes are outside and need a ladder and it's -8°C here in New Hampshire..., and it would take the better part of a day)
-JD

[Dunlapjc]

Hi group,
Not seeing any updates pushed through / published yet that have hit my box.
Is there a way to manually force an update or pull it down?
(or is it just a waiting game until it propagates).
Thanks,
JD

[AlCalzone]

I suppose you're using the HA Addon? That might take a few days, unless you want to spin up a separate instance of Z-Wave JS UI for testing.

[Dunlapjc]

@AlCalzone, You pegged it correct. Will probably try waiting, as loading up a separate instance will likely have me mucking things up more....

[Dunlapjc]

Hello, wise and helpful brain trust,
The HA Addon finally pushed the update package to bring zwave-js-ui to 9.30.1, but sadly it didn't resolve the issue of all the nodes going blank. Any other thoughts or insight here would be greatly appreciated. I feel like I already took too much of the core group's time.

Current Versions:
zwave-js-ui: 9.30.1
zwave-js: 14.3.8
Core: 2025.2.1
Supervisor. 2025.02.0
Operating System. 14.2
Frontend. 20250205.0
Controller: Zooz ZST39 running Firmware 1.50 (based on SDK 7.22.1)

A little more investigation/insight:

• does NOT seem to happen if I only I restart home assistant. -Interviewed devices come back online with Green Check
• DOES happen when I reboot the entire x86 system -the entire mesh (except for the controller) needs to be re-interviewed
• Network data still shows up in the logs from devices in the mesh with devices that have NOT been reinterviewed
• But nothing is addressable and no automations work until I re-interview the devices
• My Z-wave mesh is a little abnormal in that it was a Z-wave JS that I later migrated to a new Zooz ZST39 that during the migration took Node ID 120 and it didn't hurt anything so I never bothered to research or put the work into fixing or finding out why it didn't migrate to Node ID 1

Log File:
z-ui_2025-02-08.log

Thanks, I truly appreciate the insight and wisdom,
Jonathan

[Dunlapjc]

Any thoughts on this topic?

At a loss whether I should start backing things up, uninstalling and reinstalling, or just rebuild from scratch (the least preferred option).

Thanks,
Jonathan

[AlCalzone]

I really have no clue. Sounds like the docker volume or part thereof gets lost on restart, but no idea why.

[Dunlapjc]

A couple of more clues to see if it triggers any ideas or input...
Long story short, I haven't solved this one yet...

1. From within Z-Wave JS UI, I did a nodes.json backup file.
2. From within Z-Wave JS UI, I backed up the config file with all my keys
3. I uninstalled Z-Wave JS UI from the Home Assistant Addin menu/interface
4. When prompted I chose to delete all files/folders/data
5. Rebooted (note my Zwave integration remained, along with all my node names, but nothing worked)
6. I reinstalled Z-Wave JS UI and tried to load the backup files (settings.json and nodes.json).
7. Neither backup restored anything, but manually opening the jsons every detail is there
8. I manually put the configuration data back in, my keys and controller selection, and I was back to having all the nodes show up (no names and spinning interview state)
9. Reinterviewed all the nodes. Z-Wave JS UI is usable (except for battery operated nodes, which are a PITA to interview)

Still no clue as to what actually is causing the issue, but maybe since the backups didn't restore correctly maybe that indicates something file structure wise that might not have deleted cleanly?