I’ve also started experiencing this issue since about ~3 days ago. Didn’t change anything, didn’t update (still on 2022.11.5), HA is running on a dedicated raspberry PI.
What was mentioned earlier in the thread seems to be true, my RPi is showing 94%-100% CPU usage and 0.7/1GB memory utilization
I had a similar experience with HA suddenly acting very weird, which a reboot would magically fix - for a while. I upgraded my 1GB RPi to an 8GB RPi 3 weeks ago, restored from my last backup, and everything has been working perfectly since then.
It seems that HA sometimes “loses its mind” when memory gets scarce. My old RPi was running at about 80% memory utilization normally. My new RPi runs at about 15% memory and 3% processor usage. So HA weird behavior may somehow be related to high memory usage.
Here as well, since yesterday it has occurred twice. Never experienced this the last 8 months I’ve been using HA. I was unable to ping any zwave device, a reboot solved the issue but 12 hours later, it happened again. So, after another reboot, waiting for the next event and a bit of time to check things/logs.
The funny thing is, is that all my sensors and smart meters where reporting, so the zwave network works. But I was unable to control any switch or ping any device. So, very much a software issue I guess.
The only things that have changed the last few days are the 2023.8.4 update (from 2023.8.3) and I tried to upgrade the firmware on Aeotec multisensors 6 + Indoor Siren. The OTA firmware upgrades failed by the way.
My setup is a dedicated Intel NUC i3, 16GB, HAOS 2023.8.4, Z-stick Gen7. I’ll check the memory and CPU usage when it happens again, any other tips are much appreciated.
Having the same issue here. I thought at first it might be some of the common issues with the extension cable, I didn’t have one prior so I tried that and it didn’t help. I wasn’t having any issues when I first set this up even without the extension. I also tried a powered USB hub as on other forums some have mentioned it has helped. Neither of those work. I don’t see any errors anywhere indicating an issue with z-wave and in fact, It seems that everything is well except when this happens I can’t control or ping any of the devices. I have a total of 59 z-wave devices so I would say it’s a pretty good sized mesh. I did have some chatty devices reporting voltage pretty much every second (even while this issue is happening). I disconnected them from power so that they wouldn’t report back and the issue continues. My processor runs at 30% and I am using 1.7 of a total of 4GB of memory. I am using a Lenovo M93 running on top of a proxmox hypervisor. I have another USB dongle for Bluetooth and it is working well so I do not believe it is USB related. In case it helps, I am using an Aeotec Z-stick 7 USB Stick and I recently updated the firmware to the latest available as of today (V7.19.3). I have had this happening to me for maybe about 2 months now and with different firmware as well. I haven’t had much time to play with it but it is driving me crazy because my entire house stops working pretty much since the majority of the devices are z-wave.
Hello all !
Same problem here since last upgrade of Zwave.
I must switch off my VM on my synology and switch on to have zwave available.
is it possible to rollback to the previous version of zwave component only ?
This is happening to me too, RPi4 with latest updates, ZStick 7 with latest firmware. I have mostly Fibaro Wall Plugs, they keep reporting consumption but stops responding to on/off commands after a while.
For me it seams like restarting Z-Wave JS resolves the issue, but it comes back the next day.
It looks like I’ve got my issues solved. I’m now 3 days up and running, the controller statistics shows almost half a million received commands of which only a few (200) dropped. And all devices still respond immediately when pinged or through automations.
My setup includes about 80 Z-Wave devices of which 3 Qubino Smart meters which generate most traffic. One of them was included with S0 security. I also had tried a few over-the-air firmware updates with a couple of Aeotec Multisensors and an indoor siren. The OTA updates all failed (got stuck after 5-15%, I retried a few times). I assumed that an Abort command would cancel all that, which I did for each of these devices.
After a restart of the Z-Wave JS UI add-on, the system would work fine for a few hours. Statistics showed a lot of traffic and a lot of dropped commands. After about 80.000 received commands or a few hours, devices would not respond to a ping command or any command really. Statistics also showed a few thousand of dropped commands. I also found out that the Z-Wave JS UI add-on was just updated.
So, what did I do? I increased the threshold for the smart meters from 1% to 5%, reducing network traffic. I excluded the S0 security smart meter and included it without S0 security, this also reduces network traffic. And the devices I tried to update were excluded, reset and included again. I suspected that something still was going on with those devices.
The issues all started after the automatic update of Z-Wave JS UI to 1.15.8, but coincidently I tried a few firmware updates at about the same time. The update attempts could be the cause of my issues but I strongly suspect the automatic update to 1.15.8 is the culprit. My Z-Wave setup with A LOT of traffic was running fine for the last 8 months, right up until that add-on update. So now, with reduced traffic, it’s running fine. I don’t think I’ll ever try an OTA update again nor will I ever include any device with secure communications.
Nope, same issues again! After the 2023.9.1 update today, all my Z-Wave devices work properly, reporting all kinds of data and thus the network is intact. However, I am unable to control any of them. No responds to a ping or automations. A restart of the Z-Wave JS UI solved it.
What is wrong with Z-Wave JS UI?!
Very very frustrating to loose confidence in my system.
Rollback your Z-Wave JS UI to 1.15.8 and see if it fixes it (from System / Backups, just select the add-on backup from when it upgraded). You’ll need to then go into the add-on settings for Z-Wave JS UI and de-select auto-update (if you have it enabled).
I know this is not helpful, but I’ve also had my entire network crippled after the z-wave JS update I did before the HA 2023.9.0 release (I can’t remember the z-wave JS versions)
After a LOT of diagnosis, I’ve discovered my ZEN20’s were the cause of it. My biggest problem was no error or feed back in z-wave js to help point to the problem. It was a long process of elimination, in which I got lucky.
Rebooting HA, Restarting Z-Wave JS UI brings the network back on-line, but a few hours later it would die again.
Now that I shut the main switch off on my ZEN20’s all has been good.
@Swallowtail thanks for the suggestion. But in the mean time 2 updates were released, 1.15.10 and 1.15.11. If I interpret the changelog correctly, 1.15.11 should have fixed the issue…?
I’m a bit hesitant to rollback as I’m away from home, for about 10 days still and I’ve got everything under control now…
Well, sadly, even with 1.15.11 it keeps happening to us. We upgraded a few times over the past weeks, but a few versions back everything was fine. Now I seem to be getting meter reports and motion/temperature/illuminance sensor reports, but I cannot ping any device or control anything. Seems like the Z-stick 7 can receive messages ok, but is unable to send - though it seems to have started from a specific version of Z-wave JS UI onwards. It has improved somewhat since 1.15.11, but after some 18 hours of operation, everything seems stuck again - pretty daunting since the entire house is running Z-wave devices.
Would appreciate if anyone can guide me on what to do here.
Clearly, this is not the same for me. I have just restarted Z-Wave JS UI for the second time today, after only 10 hours. Restarting every several hours is not a solution, obviously. I am on Home Assistant Core 2023.9.1, with a total of 93 ZWave devices on the network, all linked to that Z-stick 7. The machine it runs on has ample memory and space and processing resources available to it.
Indeed - I have plenty more devices which I have not migrated (I guess another 30 or so) from Smartthings as I was hit by this problem, which is quite debilitating. Indeed - right now I feel like I stepped into the midst of a beta test…
I guess the only solution I have is to wait for a fix of ZWave JS UI. It was okay till ZWave JS UI was on V 1.15.8, and now it’s far from okay. Unfortunately, I do not have full backups right before that change, and even with the backups we have, I am not entirely comfortable restoring them. Might resort to VM snapshots in future.