Hi all,
I have a strange problem with my current setup. It is made of a Raspberry Pi (4) running deCONZ and Phoscon, and a ConBee 2 stick.
There are approx. 90 devices registered (mostly IKEA, Lumi Switches, nothing strange). Everything worked well till yesterday evening all switches stopped working, and this morning also the lights etc. stopped working.
Looking into deCONZ I see all devices, but no device has a link to any other. Bringing them closer to the ConBee2 does not change the situation.
Resetting the ConBee2 via deCONZ (selecting + F5) does not change much, in 1 of 5 cases its finds a connection to two or three random switches, but the network does not get bigger.
Any hints how to further track down this issue?
I find my Conbee 2 also in this situation quite regular. Restoring/restarting and such do not help in this caase. I did find a workaround however that has been working for a few months now.
In the decONZ map, press Leave network. Than open network settins (the lides icon) and press save. Choose join network again and you will see things start again.
My network chokes about once a month, hope it will be resolved soon
Hope this will help you,
Wim
Thank you for the idea. Now the ConBee2 recognizes at least one IKEA range extender, three switches and two lights. So slightly better. All these devices are actually quite close to the ConBee, but there are closer devices it does not see.
A bit more information - sorry, it’s my first larger problem with deCONZ and the ConBee2. So in case you would like to see logs or anything, please just ask.
In the logs I can see activities of the switches that are connected to the ConBee, but the lights are not reacting, even if they have a link according to deconz. In Phoscon, all lights are not shown as active. But there, all switches are shown active even if they are not connected according to deconz.
Do you have an error message in logs ?
Have you tried to unplug/replug the conbee ? (to restart the gateway and clean table, a reboot doesn’t work)
Sorry, no error messages in the logs. If I turn logging to Error (1/2), nothing is coming up.
But I found a few things in the system. In the syslog/journal, I find the following message from deCONZ repeating while it is running:
Nov 27 18:42:11 Homeserver deCONZ[563]: This plugin does not support propagateSizeHints()
And when I do a
sudo systemctl stop deconz
the service stops with an “exit-code”. No idea if this is normal.
Nov 27 18:52:37 Homeserver systemd[1]: Stopping deCONZ: ZigBee gateway -- REST API...
Nov 27 18:52:40 Homeserver systemd[1]: **deconz.service: Main process exited, code=exited, status=143/n/a**
Nov 27 18:52:40 Homeserver systemd[1]: **deconz.service: Failed with result 'exit-code'.**
Nov 27 18:52:40 Homeserver systemd[1]: Stopped deCONZ: ZigBee gateway -- REST API.
I tried to unplug, wait a minute and plug the ConBee2 in again. It is recognized by the system:
Nov 27 18:42:09 Homeserver kernel: [ 4243.461087] usb 1-1.3: new full-speed USB device number 13 using xhci_hcd
Nov 27 18:42:09 Homeserver kernel: [ 4243.600172] usb 1-1.3: New USB device found, idVendor=1cf1, idProduct=0030, bcdDevice= 1.00
Nov 27 18:42:09 Homeserver kernel: [ 4243.600182] usb 1-1.3: New USB device strings: Mfr=1, Product=2, SerialNumber=3
Nov 27 18:42:09 Homeserver kernel: [ 4243.600190] usb 1-1.3: Product: ConBee II
Nov 27 18:42:09 Homeserver kernel: [ 4243.600197] usb 1-1.3: Manufacturer: dresden elektronik ingenieurtechnik GmbH
Nov 27 18:42:09 Homeserver kernel: [ 4243.600204] usb 1-1.3: SerialNumber: DE2400543
Nov 27 18:42:09 Homeserver kernel: [ 4243.606874] cdc_acm 1-1.3:1.0: ttyACM0: USB ACM device
Nov 27 18:42:09 Homeserver mtp-probe: checking bus 1, device 13: "/sys/devices/platform/scb/fd500000.pcie/pci0000:00/0000:00:00.0/0000:01:00.0/usb1/1-1/1-1.3"
Nov 27 18:42:09 Homeserver mtp-probe: bus: 1, device: 13 was not an MTP device
Nov 27 18:42:09 Homeserver mtp-probe: checking bus 1, device 13: "/sys/devices/platform/scb/fd500000.pcie/pci0000:00/0000:00:00.0/0000:01:00.0/usb1/1-1/1-1.3"
Nov 27 18:42:09 Homeserver mtp-probe: bus: 1, device: 13 was not an MTP device
Correction. Your posting let me start deCONZ a little bit different… turning on logging while running the graphical user interface looked a bit “too late” to me. So I did it this way.
$ /usr/bin/deCONZ --dbg-error=2 -platform minimal
libpng warning: iCCP: known incorrect sRGB profile
This plugin does not support propagateSizeHints()
This plugin does not support propagateSizeHints()
This plugin does not support raise()
This plugin does not support propagateSizeHints()
This plugin does not support propagateSizeHints()
19:03:07:296 void deCONZ::ZclDataBase::load(const QString&),3500: assertion 'val <= nt->max' failed
19:03:07:296 void deCONZ::ZclDataBase::load(const QString&),3500: assertion 'val <= nt->max' failed
19:03:07:297 void deCONZ::ZclDataBase::load(const QString&),3500: assertion 'val <= nt->max' failed
19:03:07:297 void deCONZ::ZclDataBase::load(const QString&),3500: assertion 'val <= nt->max' failed
19:03:07:297 void deCONZ::ZclDataBase::load(const QString&),3500: assertion 'val <= nt->max' failed
19:03:07:297 void deCONZ::ZclDataBase::load(const QString&),3500: assertion 'val <= nt->max' failed
19:03:07:318 void deCONZ::ZclDataBase::load(const QString&),3499: assertion 'val >= nt->min' failed
19:03:07:705 DeviceDescription::Item DDF_ParseItem(DDF_ParseContext*, const QJsonObject&),3242: assertion 'rid.type == DataTypeString || rid.type == DataTypeTime || rid.type == DataTypeTimePattern' failed
19:03:08:031 OTAU: image path does not exist: /home/pi/otau
19:03:08:501 SQL exec failed: SELECT attr FROM sensors WHERE uniqueid = '8c:f6:81:ff:fe:55:ae:30-01-1000' AND deletedState = 'normal', error: no such column: attr (1)
19:03:08:501 SQL exec failed: SELECT attr FROM sensors WHERE uniqueid = '8c:f6:81:ff:fe:55:ae:30-01-1000' AND deletedState = 'normal', error: no such column: attr (1)
19:03:08:501 SQL exec failed: SELECT attr FROM sensors WHERE uniqueid = '8c:f6:81:ff:fe:55:ae:30-01-1000' AND deletedState = 'normal', error: no such column: attr (1)
19:03:09:060 UPNP socket not bound, state: 0
This plugin does not support propagateSizeHints()
This plugin does not support raise()
This plugin does not support propagateSizeHints()
19:03:11:092 could not delete /var/tmp/deconz-firmware-version
19:03:13:490 discovery reply doesn't contain valid version info```There is 2 deconz service
- “deconz” for the headless application
- “deconz_gui” for the version with GUI
You can have ONLY ONE running in same time, so you need to close one before using the second one.
On command line you are using “-platform minimal” so it’s fine, it’s the headless service.
On I don’t see blocking error on your logs, try adding --dbg-info=2
You have tried to just unplug/replug the gateway ?
Thank you for pointing out the differences between the services - and yes, I’m only using one of them at the same time. Actually, for debugging, I’m mostly using it like shown from the command line.
Today I tried a fresh install in a docker container to sort out the binaries in my raspberry, this also did not improve the situation. And yes, I tried several time to unplug/replug the Conbee2.
You asked for --dbg-info=2 … creates a very long log, as all scheduled jobs are “counting down”. I cannot upload the file (not allowed extension zip or txt), but I have searched for the work “block” in it: nothing.
The output looks quite reasonable.
18:13:17:858 /home/pi/.local/share/dresden-elektronik/deCONZ/config.ini exists and is writeable
18:13:17:858 /home/pi/.local/share/dresden-elektronik/deCONZ/zll.db exists and is writeable
18:13:18:033 COM: /dev/ttyACM0 : ConBee II : dresden elektronik ingenieurtechnik GmbH (0x1CF1/0x0030)
18:13:18:069 COM: /dev/ttyS0 : : (0x0000/0x0000)
18:13:18:069 dev /dev/ttyS0 (/dev/ttyS0)
18:13:18:070 ZCLDB init file /home/pi/.local/share/dresden-elektronik/deCONZ/zcldb.txt
18:13:18:158 DDF enabled for Gold status
18:13:18:174 parent process -bash
18:13:18:174 gw run mode: normal
18:13:18:174 GW sd-card image version file does not exist: /home/pi/.local/share/dresden-elektronik/deCONZ/gw-version
18:13:18:174 sd-card cid: 035344534336344780ffffffff014900
Followed by a lot of SQL/sqlite as it reads the config from the DB. Then it shows the stick:
18:13:21:915 Device firmware version 0x26780700 ConBee II
18:13:21:962 unlocked max nodes: 512
18:13:21:975 Device protocol version: 0x010E
18:13:22:107 [Master] read param with arg 0x19
18:13:22:195 Current channel 26
18:13:22:279 CTRL ANT_CTRL 0x03
18:13:22:298 CTRL ZDP_RESPONSE handler 0x0001
18:13:22:384 CTRL read param resp id: 0x13, status: 0x04
This is followed by 53 index resource entries, all similar to the following
18:13:23:034 index resource items for rules, handle: 53 (Rule TOGGLE_OFF)
18:13:23:034 /sensors/80:4b:50:ff:fe:f5:dd:1b-01-1000/state/buttonevent op: eq
18:13:23:034 /sensors/80:4b:50:ff:fe:f5:dd:1b-01-1000/state/lastupdated op: dx
18:13:23:034 /groups/23/state/any_on op: eq
18:13:23:034 state/lastupdated (trigger)
Then followed by SQL again, before you see mostly schedules counting down…
I searched for errors in the file, and found the following:
18:15:06:120 max transmit errors for node 0xBC33ACFFFE7C2FAB, last seen by neighbors 107 s
...
18:15:07:636 0xBC33ACFFFE7C2FAB seems to be a zombie recv errors 6
...
18:15:17:182 API error 101, /, link button not pressed
The last error repeats every 15 seconds.
As the Conbee3 is currently quite cheap on Amazon, I ordered one to see if the Conbee is maybe the problem.
This error concern more third app trying to use deconz in same time, you can have this error if an app trying to get a new API key.
There few chnace the conbee is broked.
No more informations around “max transmit errors for node” ? Not possible to see if there is a network issue or a connection issue.
Devices are grayed on the GUI, have you trie to re-include one to test ?
No more informations around “max transmit errors for node” ?
Not really: Here are the other lines around this error:
18:15:06:120 max transmit errors for node 0xBC33ACFFFE7C2FAB, last seen by neighbors 107 s
18:15:06:749 schedule 1 diff 15294, Fr. Nov. 28 22:30:00 2025
18:15:07:636 0xBC33ACFFFE7C2FAB seems to be a zombie recv errors 6
BC33ACFFFE7C2FAB is an IKEA repeater. The strange thing about the repeater is, that it is one of the very few components that are shown with a link in the gui.
Little update concerning “reconnecting”. The new ConBee arrived. While with the old a reconnected light worked for maybe half an hour, with the new it worked now for 2… Keep you posted.
Update: I reconnected about 15 of my 90 devices, and the new ConBee 3 works stable now for 6hrs. Looks like I need to reconnect everything…
Can try to play with zigbee network backup, make a backup on the not working network and retore it on the new conbee, but for me this procedure will not work, if effectively you have a zigbee network issue.
I agree with you, it is something within the Zigbee network - and I already have some idea who could be the “disturbing factor”.
For the time being, I will keep the ConBee3 with a “minimal” set of components so that light in the flat is working. (So to say: ConBee3 is my “production” network. I have the impression that the antenna in ConBee3 is better than in ConBee2, at least I reach parts from the flat directly that needed to go via two lights with ConBee2).)
I have now created a second network with the ConBee2 and are slowly adding components to see when problems arise. I have the impression it has something to do with one of the components I have not connected to the ConBee3, yet, but that had a connection to the ConBee2 network.
So first lesson learned: Allways have to networks - one for testing, one for production.