deCONZ 6.15 very unstable in HA 2022.8 - watchdog does not restart it.

[cortado]

Describe the issue you are experiencing

Stops at least once every day, and Watchdog does not restart it (some other plugins also had that issue in HA 2022.8, but was patched. so missing patch?)

deConz have been rock solid since i started with Conbee II and HA. Not anymore! Witch is very troublesome for all security sensors and lights.

Any solution? Logs does not tell me much, and I am not a developer. Therefor run everything as vanilla as possible. So no log entries from deCONZ that gives any hint, but a lot of errors from websocket, and Supervisor also gets timeouts – so there seems to be many problems with this version of HA

Downgrade to 2022.7?

What type of installation are you running?

Home Assistant OS

Which operating system are you running on?

Home Assistant Operating System

Which add-on are you reporting an issue with?

deCONZ

What is the version of the add-on?

6.15.0

Steps to reproduce the issue

...

Anything in the Supervisor logs that might be useful for us?

No response

Anything in the add-on logs that might be useful for us?

No response

Additional information

System:
RPib+ (32)
Home Assistant 2022.8.6
Supervisor 2022.08.3
Operating System 8.4
Grensesnitt 20220802.0 – latest

Conbee II
2.17.01 / 1.7.2022
Firmware
26720700

[davidrustingha]

I'm having the same problems. After a host reboot it works for a couple of hours, and then first only inside deconz it works, and later deconz becomes unresponsive too.

[aerox845]

for me its not working at all anymore somehow it fully stopped working after updating - to HA 2022.8 -
COM: /dev/serial/by-id/usb-dresden_elektronik_ingenieurtechnik_GmbH_ConBee_II_DE2257841-if00 / serialno: DE2XXXXX, ConBee II
13:10:59:260 start reconnect to network
13:11:02:128 device state timeout (handled)

and pretty much tried alls the fixes - even updated the stick FW

[mdegat01]

Can you share your system health report? Just copy and paste it from the link.

Also can you go to your list of integrations and do "download diagnostics" from the deconz integration and share that?

It would also help if you shared the logs from the deconz addon. And the ones from supervisor you saw about timeouts.

[aerox845]

system health:

System Information

version	core-2022.8.6
installation_type	Home Assistant OS
dev	false
hassio	true
docker	true
user	root
virtualenv	false
python_version	3.10.5
os_name	Linux
os_version	5.15.55
arch	x86_64
timezone	Europe/Berlin
config_dir	/config

Home Assistant Cloud

logged_in	false
can_reach_cert_server	ok
can_reach_cloud_auth	ok
can_reach_cloud	ok

Home Assistant Supervisor

host_os	Home Assistant OS 8.4
update_channel	stable
supervisor_version	supervisor-2022.08.3
agent_version	1.2.1
docker_version	20.10.14
disk_total	87.9 GB
disk_used	17.4 GB
healthy	true
supported	true
board	ova
supervisor_api	ok
version_api	ok
installed_addons	ESPHome (2021.12.3), TasmoAdmin (0.20.0), Node-RED (13.3.2), File editor (5.3.3), Grafana (7.6.0), Samba share (10.0.0), Studio Code Server (5.3.0), deCONZ (6.15.0)

Dashboards

dashboards	1
resources	0
views	3
mode	storage

Recorder

oldest_recorder_run	13. August 2022 um 06:48
current_recorder_run	22. August 2022 um 14:06
estimated_db_size	234.41 MiB
database_engine	sqlite
database_version	3.38.5

config_entry-deconz-78954d013b0d11ebbe94d30b645c482e.json.txt

10:10:08:963 Skip idle timer callback, too early: elapsed 856 msec
10:10:10:104 Announced to internet https://phoscon.de/discover
10:10:11:383 device state timeout (handled)
10:10:11:464 void zmMaster::handleStateIdle(zmMaster::MasterEvent) not connected goto OFF state
10:10:12:963 DEV Tick.Init: booted after 8000 seconds
[10:10:15] INFO: Successfully send discovery information to Home Assistant.
10:10:17:975 New websocket 172.30.32.1:40438 (state: 3)
10:10:19:972 GW update firmware found: /usr/share/deCONZ/firmware/deCONZ_ConBeeII_0x26720700.bin.GCF
10:10:25:471 COM: /dev/serial/by-id/usb-dresden_elektronik_ingenieurtechnik_GmbH_ConBee_II_DE2257841-if00 / serialno: DE2257841, ConBee II
10:10:26:963 start reconnect to network
10:10:28:807 device state timeout (handled)
10:10:28:887 void zmMaster::handleStateIdle(zmMaster::MasterEvent) not connected goto OFF state
10:10:31:963 failed to reconnect to network try=1
10:10:36:963 failed to reconnect to network try=2
10:10:41:963 failed to reconnect to network try=3
10:10:43:471 COM: /dev/serial/by-id/usb-dresden_elektronik_ingenieurtechnik_GmbH_ConBee_II_DE2257841-if00 / serialno: DE2257841, ConBee II
10:10:44:472 COM: /dev/serial/by-id/usb-dresden_elektronik_ingenieurtechnik_GmbH_ConBee_II_DE2257841-if00 / serialno: DE2257841, ConBee II
10:10:46:963 failed to reconnect to network try=4
10:10:48:471 COM: /dev/serial/by-id/usb-dresden_elektronik_ingenieurtechnik_GmbH_ConBee_II_DE2257841-if00 / serialno: DE2257841, ConBee II

i cant find the logs from deconz addon hmm

[cortado]

System Information

version	core-2022.8.6
installation_type	Home Assistant OS
dev	false
hassio	true
docker	true
user	root
virtualenv	false
python_version	3.10.5
os_name	Linux
os_version	5.15.32-v7
arch	armv7l
timezone	Europe/Oslo
config_dir	/config

Home Assistant Cloud

logged_in	false
can_reach_cert_server	ok
can_reach_cloud_auth	ok
can_reach_cloud	ok

Home Assistant Supervisor

host_os	Home Assistant OS 8.4
update_channel	stable
supervisor_version	supervisor-2022.08.3
agent_version	1.2.1
docker_version	20.10.14
disk_total	57.8 GB
disk_used	11.1 GB
healthy	true
supported	true
board	rpi3
supervisor_api	ok
version_api	ok
installed_addons	File editor (5.3.3), Node-RED (13.3.2), Mosquitto broker (6.1.2), Samba share (10.0.0), Duck DNS (1.15.0), SSH & Web Terminal (12.0.2), deCONZ (6.15.0), chrony (2.4.0), phpMyAdmin (0.8.1), Log Viewer (0.14.0), MariaDB (2.5.1), Z-Wave JS (0.1.65), Filebrowser (2.22.4), Z-Wave JS to MQTT (0.45.0), Cloudflared (2.0.13), Zigbee2MQTT (1.27.0-1)

Dashboards

dashboards	1
resources	1
views	10
mode	storage

Recorder

oldest_recorder_run	19. august 2022, 19:06
current_recorder_run	22. august 2022, 15:59
estimated_db_size	483.28 MiB
database_engine	mysql
database_version	10.6.8

deCONZ diagnostics (not all of the file it was too long..)

"integration_manifest": {
"domain": "deconz",
"name": "deCONZ",
"config_flow": true,
"documentation": "https://www.home-assistant.io/integrations/deconz",
"requirements": [
"pydeconz==102"
],
"ssdp": [
{
"manufacturer": "Royal Philips Electronics",
"manufacturerURL": "http:https://www.dresden-elektronik.de"
}
],
"codeowners": [
"@Kane610"
],
"quality_scale": "platinum",
"iot_class": "local_push",
"loggers": [
"pydeconz"
],
"is_built_in": true
},
"data": {
"config": {
"entry_id": "aacfed5820894a7d8c48a44b71b9dfc3",
"version": 1,
"domain": "deconz",
"title": "00212E048ACA",
"data": {
"api_key": "REDACTED",
"host": "core-deconz",
"port": 40850
},
"options": {
"allow_clip_sensor": true,
"allow_deconz_groups": false,
"master": true,
"allow_new_devices": true
},
"pref_disable_new_entities": false,
"pref_disable_polling": false,
"source": "hassio",
"unique_id": "REDACTED",
"disabled_by": null
},
"deconz_config": {
"UTC": "2022-08-22T14:00:46",
"apiversion": "1.16.0",
"backup": {
"errorcode": 0,
"status": "idle"
},
"bridgeid": "REDACTED",
"datastoreversion": "93",
"devicename": "ConBee II",
"dhcp": true,
"disablePermitJoinAutoOff": false,
"factorynew": false,
"fwversion": "0x26720700",
"gateway": "172.30.33.1",
"internetservices": {
"internet": "connected",
"remoteaccess": "disconnected",
"swupdate": "connected",
"time": "connected"
},
"ipaddress": "172.30.33.4",
"lightlastseeninterval": 60,
"linkbutton": false,
"localtime": "2022-08-22T16:00:46",
"mac": "REDACTED",
"modelid": "deCONZ",
"name": "Phoscon-GW",
"netmask": "255.255.254.0",
"networkopenduration": 180,
"panid": "REDACTED",
"portalconnection": "disconnected",
"portalservices": false,
"portalstate": {
"communication": "disconnected",
"incoming": false,
"outgoing": false,
"signedon": false
},
"proxyaddress": "none",
"proxyport": 0,
"replacesbridgeid": null,
"rfconnected": true,
"starterkitid": "",
"swupdate": {
"checkforupdate": false,
"devicetypes": {
"bridge": false,
"lights": [],
"sensors": []
},
"notify": false,
"text": "",
"updatestate": 0,
"url": ""
},
"swupdate2": {
"autoinstall": {
"on": false,
"updatetime": ""
},
"bridge": {
"lastinstall": "2022-07-01T13:20:49",
"state": "noupdates"
},
"checkforupdate": false,
"lastchange": "",
"state": "noupdates"
},
"swversion": "2.17.1",
"timeformat": "24h",
"timezone": "Etc/UTC",
"uuid": "1ba851d5-96a4-4b46-88e2-82dc508a94b0",
"websocketnotifyall": true,
"websocketport": 8081,
"whitelist": {
"00B97F5CF7": {
"create date": "2021-01-24T14:00:59",
"last use date": "2021-01-24T14:05:27",
"name": "Phoscon#B375x728"
},

[simonfox770]

Also experiencing this. The DeCONZ add-on loads and then enters into a loop of reconnecting to the network.
deCONZlog.txt

[simonfox770]

Tonight I did the following:
Rolled back DeCONZ add-on from 6.15 to 6.13 (Running Core 2022.8.6): Same problem.
Rolled back Core to 2022.6.7 (Still using rolled back DeCONZ 6.13): Same problem.

This setup was definitely working at the times the backups were created. The only thing I can't roll back is Home Assistant OS, making me think the problem lies in there somewhere.

[mdegat01]

@cortado it looks like you have both Deconz and Z2M installed. Are you trying to have them both use the same conbee ii stick? If so that won't work. You cannot share the same zigbee stick between two zigbee services

[mdegat01]

@aerox845 hrm nothing jumping out at me. By any chance do you also have ZHA added to HA and trying to use the same conbee ii stick as Deconz? If so that won't work for the same reasons I just said for cortado above.

[aerox845]

@mdegat01 ich got it fixed myself and i guess its the same issue for nearly everyone here:

i am using Unraid as Server and got Homeassistant running in a VM - i updated everything the last weekend which was a mistake in doing a big "upgrade all" and got confused what caused it to stop working.

but for everyone that gets into a "connect network, reconnect loop" the Linux Kernel the new Unraid OS uses is faulty with passing thru USB-Devices to the VM. even in the newes Unraid beta this isnt fixed! there are workarounds but i just downgraded back to Unraid 6.9.2 and everything instantly started working like a charm again.

Linux kernel: version 5.10.28 <- for Unraid 6.9.2
vs unraid 6.10.2:
Linux kernel: version 5.15.43-Unraid (CVE-2022-21499) <- VM Passthru stopped working

[cortado]

@cortado it looks like you have both Deconz and Z2M installed. Are you trying to have them both use the same conbee ii stick? If so that won't work. You cannot share the same zigbee stick between two zigbee services

No, just have used it for OTA/setup initially. It is off.

[cortado]

I did however discover one thing: I had an automation set up in the Phoscon app, connecting a motion sensor to a light directly (other automations are in Node Red) – when I deleted that, it stopped crashing every night. Could there be a bug? Also: rest of HA become almost as stable as before with 2022.8.7 and updated Supervisor. After these 3 changes I have not had deCONZ crash.

[mdegat01]

I had an automation set up in the Phoscon app, connecting a motion sensor to a light directly (other automations are in Node Red) – when I deleted that, it stopped crashing every night

It's definitely possible. To be honest I'm not very familiar with the actual Phoscon app, I only know addons and HA. If it is an issue with the Phoscon app itself you should submit a bug here.

rest of HA become almost as stable as before with 2022.8.7 and updated Supervisor.

This is also definitely possible. There were a few significant issues with addons resolved by 2022.08.5 of supervisor. Particularly immediately following a host reboot, addons were sometimes unable to get their config. So if the timing of this issue was matching up with a host reboot then there's a decent chance the supervisor update took care of it.

[simonfox770]

Managed to fix the problem by connecting the ConBeeII to my PC and using the command line tool to upgrade the firmware.

[github-actions]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.