zigbee2mqtt
zigbee2mqtt copied to clipboard
Zigbee network crashing because of HA update
What happened?
Everything worked stable without problem. I added an RFXcom adaptor in Oracle Virtual box around the 8th or 9th of March Probably I also did some of the HA updates and other plugin updates and I also did a Oracle Virtual Box update to the latest version. To the best of my knowledge after rebooting my virtual machine the zigbee network stopped working, although I'm not 100% sure since my focus was on getting the RFXcom adapter to work.
Diggin into the zigbee problem I noticed that it runs very shortly after which it crashes. Zigbee2MQTT continues to run. Devices are listed but not responding, for example switching a device on/off gives me this:
z2m: Publish 'set' 'state' to 'Power Kantoor' failed: 'Error: ZCL command 0x54ef441000523187/1 genOnOff.off({}, {"timeout":10000,"disableResponse":false,"disableRecovery":false,"disableDefaultResponse":false,"direction":0,"reservedBits":0,"writeUndiv":false}) failed (SRSP - AF - dataRequest after 6000ms)'
What did you expect to happen?
My zigbee network was rock solid, hope to get it back.
How to reproduce it (minimal and precise)
When I reboot my virtual machine zigbee starts working, it also reports stats (power, temp, etc), however after 5 to 10 minutes nothing is reported anymore. After that I see that in my log the devices stop to respond. See see my log.
Zigbee2MQTT version
2.1.3-1
Adapter firmware version
20210708
Adapter
Sonoff ZIgbee 3.0 CP210x
Setup
Hassio 15 / 25.3.3 on Virtual Box 7.1.6
Debug log
Have you tried unplugging the sonoff adapter for a while? Looks like it's sending all kinds of corrupted frames (unknown commands). If it's not the adapter's internal state that has a problem, then I'd say you're looking at some kind of interference... at the radio, at the connection, at the passthrough... Does it still do it if you remove that new adapter you added?
Thanks @Nerivec for your help.
I tried all your suggestions and more, but nothing worked—until I decided to load an old virtual image in my virtual machine. Without any hardware changes, everything is now working again, and my log file shows no errors.
Based on this:
- There is no hardware issue.
- The problem is not related to an Oracle VirtualBox update.
Potential causes I’ve ruled out:
- Recent plugin updates (e.g., TP-Link/PowerCalc, Landroid Cloud). After restoring my old backup, doing these updates, everything, including Zigbee, continued to function correctly.
- Virtual image size. Initially, it was 32GB, and I resized it to 64GB. However, increasing the disk size in my old backup had no impact—Zigbee still worked.
My conclusion: A Home Assistant update likely caused the Zigbee issue.
I was running Hassio 15 / 25.3.3 with which I experienced the zigbee problems. After reverting to my backup with Core 2025.2.5 and Supervisor 2025.03.3, everything functions normally again.
Does anybody know if this can happen? Can I maybe upgrade the core without the supervisor? To find out which one causes the problem? Or manually upgrade with smaller steps, so not directly to the latest version?
I am experiencing the same issues. Is there a way to downgrade superviser?
Did you mean superviser 2025.02.3 worked?
My entire network is going down, I notice some delay in the response of some automations yesterday, then I decided to update zigBee2Mqtt and Home Assistant and today the entire network went down it was woking flawlessly for the past 10 months.
Anyone brave enough and tried with 2025.3.4? Im afraid my WAF will decrease again 😅
Update: Unfortunately my network crashed again in the middle of the night. I was on 2025.2.5 so maybe it is the OS update instead of the core update?
I made a copy of my Virtual box image and doing some test;
Working setup: core 2025.2.5 supervisor 2025.03.3 OS 14.2 front-end 20250221.0
Updated: core 2025.3.4 (updated) supervisor 2025.03.3 OS 14.2 front-end 20250306.0 (seems part of the core update)
So far no problems, I first noticed directly crashes of my whole Zigbee network as stated in the first post. I will let this run tonight. If this works I will attempt to update to OS to 15.0 tomorrow get confirmation that the OS update causes the problem. If confirmed, I will wait until 15.x or whatever comes next to see if the problem has been resolved.
update: 12hours later: Zigbee is still running stable after the core update.
Just updated to 15.1. Now need to wait to see if the Zigbee network continues to run stable. Will update here.
After 13 hours it looks to be running without problem. 15.0 crashed after a few minutes.
Conclusion: 15.1 fixes the problem for me which 15.0 introduced.