Conbee III lost devices one week after installation

Advice
1

Hi all,

first of all, let me say I’m a great fan of Deconz - it has been running half of my house reliably for 5 years now.

I am also aware that there are several threads with issues similar to the one I’m encountering, but the details are always a bit different, so I’m describing my specifics and hope to get some advice on the next steps.

What happend:

  • I’ve been running a network with a few dozen devices (lights, sensors, switches) controlled by a RaspBee (honestly, not sure which version - bought in 2019) on a Raspberry Pi 3, running Deconz in Docker, controlled through the Home Assistant Deconz integration
  • A few weeks ago, I decided I needed to update the docker image - was running an ancient one
  • I realized I needed to reinstall Raspberry Pi OS first, which was also ancient
  • After reinstallation, things didn’t work properly. I thought it would be a good idea to upgrade to a ConBee III instead of the RaspBee. (I later found out the issues were very likely due to using the beta version of Deconz pulled with the latest tag instead of latest-stable)
  • I updated the firmware on the ConBee after receiving it
  • Switching to the ConBee III (Zigbee database etc. all retained) I had to manually reinclude every device - this was a lot of work, but in the end I had a perfectly working home for one week
  • I went away for the weekend, came back, and nothing was working. All greyed out in Phoscon and unavailable in Home Assistant.

Attempts to recover that did not work:

  • Restarting the container
  • Rebooting the host
  • Unplugging and replugging the ConBee III
  • Restarting Home Assistant
  • Waking battery devices, power cycling mains devices

Deconz shows:

ConBee III
Gateway Version 2.31.2
Firmware 26550900

My next step would be to try this:

Can someone confirm that this is the right thing to try now, or is there anything else I can do?

Thank you! :folded_hands:

2

No, you are fine for me.
You know you can make a zigbee backup using Phoscon ? usefull to migrate from a gateway to another. This prevent to have to re-include all devices.

All have worked for a week, so I think you was fine too.

Do you have the GUI ? if yes just take a look in deconz/help/debug view with flag “info”+“info_l2”+“error”+“error_l2”, if all is broken, not possible you haven’t information here.

All grayed mean generealy disconnected, can be a problem on the gateway, sofware (can check zigbee setting,), but hardware too (power supply).

If you have changed the host and the gateway, not sure you will see a zigbee network backup on the last link you have show.

3

Thanks a lot! It took me a while but I managed to connect to the GUI via VNC.

I can see all my devices there in the network view, but absolutely no links betwen them :open_mouth:

Nothing telling in the logs.

Where can I go from here?

4

In the hidden “Zigbee configuration” page, the most recent entry has deconz version 2.31.2, the one I’m running now. I am tempted to click “Load” (because I don’t know what else to try), but I don’t know what will happen or if I can break something…

5

It will re-use an old zigbee configuration.
Can be a previous network key for exemple, you can see in the “table” the difference between backup.

On the GUI logs are in deconz/help/debug view, no errors inside ?

Can check the zigbee network setting Network lost and configuration restore does not help · dresden-elektronik/deconz-rest-plugin Wiki · GitHub

You have no links, but they are not responsive too ?

6

No, unfortunately not.

I tried that, it did not bring back the links.

Depends on what you mean by responsive - if I right-click a device box and select “read active neighbors”, the “LED” blinks red a few times.

7

No, make a try on a router, for exemple a bulb, go on cluster 0x0006 and try a on/off.

Can you share some logs ? Because it’s strange you haven’t error and a full not working network.

8

Thanks for the support!

If I open the + dropdown on a node and click on “0006 On/Off”, nothing happens (neither in the GUI nor with the bulb).

Here is the current docker log (it got way busier since I enabled VNC and am using the GUI in parallel…): gist:f63e4686066451f4e896ef59bc172356 · GitHub

9

16:12:31:700 API error 901, /config/utc, Error setting date and time

This message is not blocking but strange, the docker is connected to internet ? (the OS ask server to get time).

What happen at 15:54:24:526 ? a restart ?

--appdata=/opt/deCONZ --dbg-info=1 --dbg-aps=0 --dbg-zcl=0 --dbg-ddf=0 --dbg-dev=0 --dbg-zdp=0 --dbg-ota=0 --dbg-error=0

You have something to overwrite thoses setting ? I think yes, In my mind need to see --dbg-info=2 --dbg-error=2 but you have some of them in your logs, so I think there are enabled from other place.

16:33:46:415 Daylight now: dusk, status: 210, daylight: 0, dark: 1

Lot of spam here too, the check is based on time difference, without time …

02:25:14:979 discovery reply doesn't contain valid version info

Need an internet connexion to have discovery.

I realy think you docker have no connexion to “outside” it’s something normal ?

About your network configuration (available on your link), you have it visible on your log too.
So we can see

  • You have switched channel from 15 to 11
  • New Macadress and Network key , mean you restarted all from scratch
  • You have switched the setting staticNwkAddress from false to true.

Honneslty if all was working during a week, for me all is fine.

16:25:57:414 send permit join, duration: 65

You have tried a permit join ?

I m not sure there is at least 1 device make a request to the gateway …
Have you tried to re-include a device to test, a router (easier) ?

And nothing beetween

17:25:15:844 discovery reply doesn't contain valid version info
...................
13:39:46:417 Daylight now: solarNoon, status: 170, daylight: 1, dark: 0

???

You are using the USB extension ? even the gateway alone make more logs than that.

And this is not good for me

16:01:49:930 node 0xB43522FFFEEE3F44 leave wait state

And we never see them again.

10

Not sure where to go from here.

I tried a permit join and to reconnect one of the light bulbs, no effect - using the exact same procedure I used last week to include them.

I also checked that I can access the internet (curl) from within the docker container.

11

Some more stuff from the log.

Btw, yes, I am using a USB extension cable.

19:35:24:333 add task 387 type 0 to 0x842E14FFFE586FB4 cluster 0x0003 req.id 114
19:35:24:868 add task 390 type 0 to 0x842E14FFFE586FB4 cluster 0x0003 req.id 117
19:35:24:869 emit event/tick: 0x842E14FFFE586FB4
19:35:24:870 delay sending request 117 dt 0 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:24:879 add task 392 type 0 to 0xD0CF5EFFFEC80794 cluster 0x0003 req.id 119
19:35:24:881 delay sending request 117 dt 0 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:24:957 delay sending request 117 dt 1 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:25:057 delay sending request 117 dt 1 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:25:157 delay sending request 117 dt 1 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:25:257 delay sending request 117 dt 1 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:25:357 delay sending request 117 dt 1 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:25:456 delay sending request 117 dt 1 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:25:557 delay sending request 117 dt 1 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:25:657 delay sending request 117 dt 1 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:25:756 delay sending request 117 dt 1 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:25:857 delay sending request 117 dt 1 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:25:957 delay sending request 117 dt 2 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:26:057 delay sending request 117 dt 2 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:26:156 delay sending request 117 dt 2 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:26:256 delay sending request 117 dt 2 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:26:357 delay sending request 117 dt 2 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:26:456 delay sending request 117 dt 2 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:26:557 delay sending request 117 dt 2 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:26:657 delay sending request 117 dt 2 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:26:757 delay sending request 117 dt 2 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:26:860 delay sending request 117 dt 2 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:26:966 delay sending request 117 dt 3 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:27:057 delay sending request 117 dt 3 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:27:156 delay sending request 117 dt 3 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:27:256 delay sending request 117 dt 3 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:27:356 delay sending request 117 dt 3 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:27:457 delay sending request 117 dt 3 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:27:557 delay sending request 117 dt 3 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:27:656 delay sending request 117 dt 3 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:27:757 delay sending request 117 dt 3 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:27:856 delay sending request 117 dt 3 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:27:957 delay sending request 117 dt 4 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:28:056 delay sending request 117 dt 4 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:28:157 delay sending request 117 dt 4 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:28:256 delay sending request 117 dt 4 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:28:356 delay sending request 117 dt 4 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:28:456 delay sending request 117 dt 4 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:28:557 delay sending request 117 dt 4 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:28:656 delay sending request 117 dt 4 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:28:756 delay sending request 117 dt 4 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:28:857 delay sending request 117 dt 4 ms to 0x842E14FFFE586FB4, ep: 0x01 cluster: 0x0003 onAir: 1
19:35:29:151 Erase task req-id: 114, type: 0 zcl seqno: 134 send time 5, profileId: 0x0104, clusterId: 0x0003
19:35:29:701 Erase task req-id: 119, type: 0 zcl seqno: 136 send time 5, profileId: 0x0104, clusterId: 0x0003
19:35:29:898 add task 397 type 0 to 0xD0CF5EFFFEC80794 cluster 0x0003 req.id 128
19:35:29:900 emit event/tick: 0xD0CF5EFFFEC80794
19:35:32:957 Daylight now: nightStart, status: 230, daylight: 0, dark: 1
19:35:33:871 Erase task req-id: 117, type: 0 zcl seqno: 135 send time 4, profileId: 0x0104, clusterId: 0x0003
19:35:34:721 Erase task req-id: 128, type: 0 zcl seqno: 137 send time 5, profileId: 0x0104, clusterId: 0x0003
19:35:36:168 node 0x842E14FFFE586FB4 leave wait state
19:35:39:256 node 0xD0CF5EFFFEC80794 leave wait state
19:35:42:957 Daylight now: nightStart, status: 230, daylight: 0, dark: 1
19:35:49:957 send permit join, duration: 65
19:35:52:956 Daylight now: nightStart, status: 230, daylight: 0, dark: 1
12

Can you tell me if there’s a way to reset everything and start from scratch? The other human beings in this house are getting angry at me. At this point, manually reconnecting every device seems the lesser evil.

13

I tried a permit join and to reconnect one of the light bulbs, no effect - using the exact same procedure I used last week to include them

So the inclusion have worked, the device was visible on the API, but can’t be used ?

On phoscon you can see a reset procedure on the gateway part, but idk what it does.
Something you can try is installing deconz on another machine (a PC) migrate the database and try it.
With hat you will see if the problem is from the gateway or the OS.

14

No, the inclusion did not even work.

I tried all options like resetting the stick or restoring an old Z-Wave config.

In the end, I decided to plug my old RaspBee back in and reinclude everything from scratch. Now my home is running again. I’ll return the ConBee - probably a hardware failure.

Thanks again for your help!

15

IDK, it have worked during a week, how the conbee can break sudently without action ?