After updating from Openhab 3 to 4, I receive the following errors and deconz is switching between online and offline:
2023-08-16 15:33:03.157 [INFO ] [ab.event.ThingStatusInfoChangedEvent] - Thing 'deconz:deconz:48a214a6' changed from UNINITIALIZED (NOT_YET_READY) to INITIALIZING
2023-08-16 15:33:03.374 [INFO ] [ab.event.ThingStatusInfoChangedEvent] - Thing 'deconz:deconz:48a214a6' changed from INITIALIZING to UNKNOWN
2023-08-16 15:34:12.209 [INFO ] [ab.event.ThingStatusInfoChangedEvent] - Thing 'deconz:deconz:48a214a6' changed from UNKNOWN to ONLINE
2023-08-16 15:36:12.210 [INFO ] [ab.event.ThingStatusInfoChangedEvent] - Thing 'deconz:deconz:48a214a6' changed from ONLINE to OFFLINE (COMMUNICATION_ERROR): Watchdog timed out after 120s. Websocket seems to be dead.
2023-08-16 15:36:22.253 [INFO ] [ab.event.ThingStatusInfoChangedEvent] - Thing 'deconz:deconz:48a214a6' changed from OFFLINE (COMMUNICATION_ERROR): Watchdog timed out after 120s. Websocket seems to be dead. to ONLINE
2023-08-16 15:38:22.254 [INFO ] [ab.event.ThingStatusInfoChangedEvent] - Thing 'deconz:deconz:48a214a6' changed from ONLINE to OFFLINE (COMMUNICATION_ERROR): Watchdog timed out after 120s. Websocket seems to be dead.
2023-08-16 15:38:32.290 [INFO ] [ab.event.ThingStatusInfoChangedEvent] - Thing 'deconz:deconz:48a214a6' changed from OFFLINE (COMMUNICATION_ERROR): Watchdog timed out after 120s. Websocket seems to be dead. to ONLINE
I have tried to make a new install of openhab 4.01 and still get the same error.
I have tried to switch the ConBee II stick and still see the same error:
After a downgrade to openhab 3 everything works again.
Any ideas on how to fix the problem?
omr@shs2:~$ sudo systemctl status deconz
● deconz.service - deCONZ: ZigBee gateway -- REST API
Loaded: loaded (/lib/systemd/system/deconz.service; enabled; vendor preset: enabled)
Drop-In: /etc/systemd/system/deconz.service.d
└─override.conf
Active: active (running) since Tue 2023-08-22 20:15:29 CEST; 1 day 17h ago
Main PID: 1079 (deCONZ)
Tasks: 5 (limit: 4447)
Memory: 49.7M
CPU: 5h 14min 4.855s
CGroup: /system.slice/deconz.service
└─1079 /usr/bin/deCONZ -platform offscreen --http-port=80
Aug 22 20:15:29 shs2 systemd[1]: Started deCONZ: ZigBee gateway -- REST API.
Aug 22 20:15:30 shs2 deCONZ[1079]: QStandardPaths: XDG_RUNTIME_DIR not set, defaulting to '/tmp/runtime-omr'
Aug 22 20:15:31 shs2 deCONZ[1079]: libpng warning: iCCP: known incorrect sRGB profile
Aug 22 20:15:31 shs2 deCONZ[1079]: libpng warning: iCCP: known incorrect sRGB profile
Aug 22 20:15:36 shs2 deCONZ[1079]: This plugin does not support propagateSizeHints()
Aug 22 20:15:36 shs2 deCONZ[1079]: This plugin does not support propagateSizeHints()
Aug 22 20:15:39 shs2 deCONZ[1079]: This plugin does not support propagateSizeHints()
I have tried both port 80 and 8081 and they are not used by other servers.
When setup both port 80 and 8081 can be browsed.
list | grep -i deconz looks OK
The deconz service is happy
But when it checked the version number of deconz, I realize that the Firmware is not connected.
Do you know why?
Hi all
Unfortunately, the problem could not be solved for me with this suggested solution.
The problem occurred only after the update to openhab 4.0.1 or now 4.0.2.
But I also have a different initial situation:
The firmware in the Deconz board was always connected - the version is up to date.
HTTP connection via port 80
Websocket is not used
Gateway: Pi Zero w and RaspBee II
Error message: Watchdog timed out after 120s. Websocket seems to be dead.
Does anyone have an clue what could be the reason for this?
When you browse into deconz on port 80 with a WEB browser, does everything work OK from the WEB interface? Are you able to get sensor data and turn lights on/off?
Is deconz the only hint of trouble you see in log? (Thinking of upgrade to Java 17, which is a requirement)
Yes, I can read out sensor data via the web interface on port 80, everything looks good. I have not connected switches there.
I also use a second gateway setup with a Pi 3 and Cobnee II, with this gateway I have no problems.
Correct, deconz is the only thing in the log showed continuous error messages.