Zigbee2mqtt stopped working overnight (Herdsman errors) - help needed

Configuration Zigbee
, ,
1

Hi all,
Need some help as Zigbee2mqtt stopped working overnight and won’t start anymore.
Some context:
I’m moving from conbee ii on a RPI to Sonoff Dongle E (flashed with: ncp-uart-hw-v7.4.4.0-zbdonglee-115200.gbl) on a Proxmox VM with HA core (2025.4.3).
The new VM with Sonoff is running with a new Zigbee network with a different channel for some weeks without issues.
Yesterday I started moving devices to the ‘Sonoff network’, from 4 devices to about 25.
So far so good. Everything works
During the night it started with these errors (sensor values appear in HA up to this time and before this time only info related messages):

[2025-04-21 01:01:37]  [31merror [39m: 	zh:ember:ezsp: ERROR Transaction failure; status=ASH_ERROR_TIMEOUTS. Last Frame: [FRAME: ID=52:"SEND_UNICAST" Seq=148 Len=7].
[2025-04-21 01:01:37]  [31merror [39m: 	zh:ember: Adapter fatal error: ASH_ERROR_TIMEOUTS
[2025-04-21 01:01:37]  [32minfo [39m: 	zh:ember:uart:ash: ASH COUNTERS since last clear:
[2025-04-21 01:01:37]  [32minfo [39m: 	zh:ember:uart:ash:   Total frames: RX=1216, TX=1483
[2025-04-21 01:01:37]  [32minfo [39m: 	zh:ember:uart:ash:   Cancelled   : RX=0, TX=0
[2025-04-21 01:01:37]  [32minfo [39m: 	zh:ember:uart:ash:   DATA frames : RX=1209, TX=267
[2025-04-21 01:01:37]  [32minfo [39m: 	zh:ember:uart:ash:   DATA bytes  : RX=30399, TX=7462
[2025-04-21 01:01:37]  [32minfo [39m: 	zh:ember:uart:ash:   Retry frames: RX=7, TX=0
[2025-04-21 01:01:37]  [32minfo [39m: 	zh:ember:uart:ash:   ACK frames  : RX=0, TX=1216
[2025-04-21 01:01:37]  [32minfo [39m: 	zh:ember:uart:ash:   NAK frames  : RX=0, TX=0
[2025-04-21 01:01:37]  [32minfo [39m: 	zh:ember:uart:ash:   nRdy frames : RX=0, TX=0
[2025-04-21 01:01:37]  [32minfo [39m: 	zh:ember:uart:ash:   CRC errors      : RX=0
[2025-04-21 01:01:37]  [32minfo [39m: 	zh:ember:uart:ash:   Comm errors     : RX=0
[2025-04-21 01:01:37]  [32minfo [39m: 	zh:ember:uart:ash:   Length < minimum: RX=0
[2025-04-21 01:01:37]  [32minfo [39m: 	zh:ember:uart:ash:   Length > maximum: RX=0
[2025-04-21 01:01:37]  [32minfo [39m: 	zh:ember:uart:ash:   Bad controls    : RX=0
[2025-04-21 01:01:37]  [32minfo [39m: 	zh:ember:uart:ash:   Bad lengths     : RX=0
[2025-04-21 01:01:37]  [32minfo [39m: 	zh:ember:uart:ash:   Bad ACK numbers : RX=0
[2025-04-21 01:01:37]  [32minfo [39m: 	zh:ember:uart:ash:   Out of buffers  : RX=0
[2025-04-21 01:01:37]  [32minfo [39m: 	zh:ember:uart:ash:   Retry dupes     : RX=7
[2025-04-21 01:01:37]  [32minfo [39m: 	zh:ember:uart:ash:   Out of sequence : RX=0
[2025-04-21 01:01:37]  [32minfo [39m: 	zh:ember:uart:ash:   ACK timeouts    : RX=0
[2025-04-21 01:01:37]  [31merror [39m: 	zh:ember:ezsp: ERROR Transaction failure; status=NOT_CONNECTED. Last Frame: [FRAME: ID=52:"SEND_UNICAST" Seq=148 Len=7].
[2025-04-21 01:02:07]  [32minfo [39m: 	zh:ember:uart:ash: Port closed.
[2025-04-21 01:02:07]  [32minfo [39m: 	zh:ember:uart:ash: ======== ASH stopped ========
[2025-04-21 01:02:07]  [32minfo [39m: 	zh:ember:ezsp: ======== EZSP stopped ========
[2025-04-21 01:02:07]  [32minfo [39m: 	zh:ember: ======== Ember Adapter Stopped ========
[2025-04-21 01:02:07]  [31merror [39m: 	z2m: Adapter disconnected, stopping
[2025-04-21 01:02:07]  [32minfo [39m: 	z2m:mqtt: MQTT publish: topic 'zigbee2mqtt/bridge/state', payload '{"state":"offline"}'
[2025-04-21 01:02:07]  [32minfo [39m: 	z2m: Disconnecting from MQTT server
[2025-04-21 01:02:07]  [32minfo [39m: 	z2m: Stopping zigbee-herdsman...
[2025-04-21 01:02:07]  [32minfo [39m: 	z2m: Stopped zigbee-herdsman
[2025-04-21 01:02:07]  [32minfo [39m: 	z2m: Stopped Zigbee2MQTT

(I have watchdog off for now to notice if there would be issues).
Every time when I restart zigbee2mqtt I get these errors:

[2025-04-21 07:23:21] info: 	zh:ember: Using default stack config.
[2025-04-21 07:23:21] info: 	zh:ember: ======== Ember Adapter Starting ========
[2025-04-21 07:23:21] info: 	zh:ember:ezsp: ======== EZSP starting ========
[2025-04-21 07:23:21] info: 	zh:ember:uart:ash: ======== ASH Adapter reset ========
[2025-04-21 07:23:21] info: 	zh:ember:uart:ash: RTS/CTS config is off, enabling software flow control.
[2025-04-21 07:23:21] info: 	zh:ember:uart:ash: Serial port opened
[2025-04-21 07:23:21] info: 	zh:ember:uart:ash: ======== ASH starting ========
[2025-04-21 07:23:24] info: 	zh:ember:uart:ash: ======== ASH Adapter reset ========
[2025-04-21 07:23:24] info: 	zh:ember:uart:ash: ======== ASH starting ========
[2025-04-21 07:23:26] info: 	zh:ember:uart:ash: ======== ASH Adapter reset ========
[2025-04-21 07:23:26] info: 	zh:ember:uart:ash: ======== ASH starting ========
[2025-04-21 07:23:29] info: 	zh:ember:uart:ash: ======== ASH Adapter reset ========
[2025-04-21 07:23:29] info: 	zh:ember:uart:ash: ======== ASH starting ========
[2025-04-21 07:23:31] info: 	zh:ember:uart:ash: ======== ASH Adapter reset ========
[2025-04-21 07:23:31] info: 	zh:ember:uart:ash: ======== ASH starting ========
[2025-04-21 07:23:34] error: 	z2m: Error while starting zigbee-herdsman
[2025-04-21 07:23:34] error: 	z2m: Failed to start zigbee-herdsman
[2025-04-21 07:23:34] error: 	z2m: Check https://www.zigbee2mqtt.io/guide/installation/20_zigbee2mqtt-fails-to-start_crashes-runtime.html for possible solutions
[2025-04-21 07:23:34] error: 	z2m: Exiting...
[2025-04-21 07:23:34] error: 	z2m: Error: Failed to start EZSP layer with status=HOST_FATAL_ERROR.
    at EmberAdapter.initEzsp (/app/node_modules/.pnpm/[email protected]/node_modules/zigbee-herdsman/src/adapter/ember/adapter/emberAdapter.ts:672:19)
    at EmberAdapter.start (/app/node_modules/.pnpm/[email protected]/node_modules/zigbee-herdsman/src/adapter/ember/adapter/emberAdapter.ts:1538:24)
    at Controller.start (/app/node_modules/.pnpm/[email protected]/node_modules/zigbee-herdsman/src/controller/controller.ts:136:29)
    at Zigbee.start (/app/lib/zigbee.ts:69:27)
    at Controller.start (/app/lib/controller.ts:104:13)
  • What does the HOST_FATAL_ERROR mean? Is that related to the mosquitto host (core-mosquitto)? or communication to/from the Sonoff adapter, something else?

Restarted several times (restart zigbee2mqtt, restart HA, restart VM), the error stays. I did not change anything while sleeping!
The configuration.yaml:

version: 4
mqtt:
  base_topic: zigbee2mqtt
  server: mqtt://core-mosquitto:1883
  user: addons
  password: xxxxx
  keepalive: 60
  reject_unauthorized: true
  version: 4
serial:
  port: /dev/ttyUSB0
  adapter: ember
  baudrate: 115200
  rtscts: false
advanced:
  log_level: info
  channel: 25
  network_key:
    - [16 lines with numbers]
  pan_id: xxxxx
  ext_pan_id:
    - [8 lines with numbers]
frontend:
  enabled: true
  port: 8099
homeassistant:
  enabled: true
devices:
[devices after this line]

Mosquitto is running fine and receives messages from several other non-zigbee mqtt devices.
I tried the config also with:

  port: >-
    /dev/serial/by-id/usb-Itead_Sonoff_Zigbee_3.0_USB_Dongle_Plus_V2_c84c4fc1d112ef11bca275b8bf9df066-if00-port0
user: mosquitto
password: [the password]

No difference.

In the zigbee2mqtt configuration tab of the addon the mqtt and serial sections where empty.
I added to test this morning:

mqtt:
  base_topic: zigbee2mqtt
  server: mqtt://core-mosquitto:1883
  user: mosquitto
  password: [password]
  • I understand that Herdsman errors are related to mqtt communication but do not see what changed/went wrong.
  • Is something scheduled at 01:00 hours for zigbee2mqtt?
  • I really do not understand what I am doing wrong. Can someone point me into the right direction? Thanks!
2

Do you have more then one zigbee network? Like running zha and z2m.
If you do use just one.

3

Thanks for your reaction. Some context to clarify.

Currently I have (temporarily) 2 Zigbee networks.
The old one on channel 15 on a dedicated rpi with conbee II, still about 5 device to migrate to the new network. This network is connected to a different HA instance.
The new network is the Sonoff E with Ember firmware using channel 25. The Sonoff is connected to a Proxmox VM with HA. HA sees the adapter (and it has been working for a few weeks just fine, but only with a few devices). Did the migration yesterday. Worked fine for about half a day until that moment in the night (see log above).
It is a bit complicated as I am halfway of two migrations:
The two networks are completely independent.

  • The conbee II is connected to another HA instance on rpi.
  • Besides changing the conbee to Sonoff I move my HA installation from the rpi to Proxmox VM.
  • Restored backup from rpi to VM, removed conbee II from the VM and installed Zigbee2mqqt.
  • Restored zigbee2mqqt backup from the trial weeks.
4

Solved the issue myself. Did the following steps which did not work:

  • Made a backup of the zigbee2mqtt folder.
  • Removed zigbee2mqtt addon
  • Removed the zigbee2mqtt folder.
  • Installed it again.
  • Copied back the backup folder.
  • Start again. Same error.
  • Did the onboarding again
  • Copied back database.db, state.json, and copied the device and ID info from the original configuration.yaml. Error stayed.

Then, just to be sure I removed the extension cable from the Sonoff adapter and reconnected.

  • Zigbee2mqtt started again without errors!
  • No data from sensors and could not toggle switches.
  • Had to re-pair all devices (did the ones connected to mains first). Everything works again.

Conclusion: The USB cable ‘decided’ during to not have a proper connection with the USB adapter. Maybe it was not seated well. I will never know!
Solved for now!

5

FYI - The errors you saw contained a link to the Z2M troubleshooting page 2 lines above the host error line.

Apart from mentioning all possible causes of host error, one of the items on that page specifically lists checking the connectors on the usb cable