Mine is something like this:
In raspberry I stopped deconz service and started deconz-gui service to access this.
Mine is something like this:
In raspberry I stopped deconz service and started deconz-gui service to access this.
It seems, that you do not have a mesh. Are you shure, that all devices are available in Phoscon?
And can you control the devices via HA? I can not imagine.
Can it be, that the distance between the switches or sensors and the stick is too large?
By the way: the deconz GUI is the phoscon app which enables you to manage and configure the devices
Phoscon does not provide the mesh-view
Thanks, I’ll pick up a few the next time I’m dragged in there kicking and screaming!
Just to let you know, I’m in the same boat. installed deconz as an hass.io addon. After the upgrade some of the devices are unavailable, or to be more precise, they’re flapping. I’ll also try and investigate further.
there is 104.3 out with a discovery fix:
Nope, nobody listening on 2222 :/.
Think this might be the moment for a clean install.
usually how soon the fix will reach hass docker?
Yes all devices are available and I can control them by HA.
When trying to restart Home Assistant (Hassio), it always fails restarting because of the following:
2020-01-22 08:25:32 ERROR (MainThread) [homeassistant.components.deconz] Error connecting to deCONZ gateway at 172.30.33.0
But a reboot of the computer always works. Any idea how to avoid this error?
Edit: Seems like it’s deCONZ that stops the whole Hassio from restarting…
Hello Robban,
I can confirm at this point that it’s related to 104 version. I have rolled back to 103.6 and I have had no connection issues. Something was changed that resulted on this.
Can someone help me on how to debug this?
Thank you all in advance.
Ok, then there might be an issue with your TigerVNC or the tool you are using.
Sorry, I can not help further
How long do you wait?
When you stop the normal deconz and start the GUI it can take some minutes before the GUI updates the links. It always starts with no links shown. That is normal. It does not mean that the linking is not there. It just takes time for the graphical tool to pick up next status update from each device showing the link quality
thank you for your and Pirol_62 answers. I waited few (1 or 2 minutes). Maybe I’ll check again waiting more time. Is there any way to start client without stopping deconz service (used by HA)
@Cosby86
Hi, I see the problem.
You’re stopping the service. No service - no mesh.
There is no need to stop the service.
Simply install the VNC on your desktop and connect it to your deConc instance like shown above
https://tigervnc.org/ Download here
I’m in raspbian, I have deconz and deconz-gui services. If I keep deconz service up and running, with vnc i access the RASPBIAN DESKTOP where i find deconz app
if i dont stop deconz and start deconz-gui I cannot log into the app
Issues with Raspbee
I have
I did setup everything and paired some xiaomi aqara door sensors and temperature sensors. However, all of the sudden something crashed when pairing a temperature sensor. From thereon everthing has gone wrong, and nothing seems to work. I tried a lot of different reboots, resets etc but could not make it work.
Now I reflashed the Hassio-installed. Still not working.
Entering the Deconz web UI through the official Hassio addon I see alot of 503 errors in the dev console. They were there before the reflashing also. I suspect this has something to do with my problems.
*Deconz log after starting addon
starting version 232
[19:01:15] INFO: Waiting for device...
[19:01:16] INFO: Starting VNC server...
[19:01:20] INFO: Starting the deCONZ gateway...
libEGL warning: DRI2: failed to open swrast (search paths /usr/lib/arm-linux-gnueabihf/dri:${ORIGIN}/dri:/usr/lib/dri)
libEGL warning: DRI2: failed to open swrast (search paths /usr/lib/arm-linux-gnueabihf/dri:${ORIGIN}/dri:/usr/lib/dri)
libpng warning: iCCP: known incorrect sRGB profile
[19:01:21] INFO: Starting Nginx...
[19:01:21] INFO: Running Hass.io discovery task...
2020/01/22 19:01:21 [notice] 372#372: using the "epoll" event method
2020/01/22 19:01:21 [notice] 372#372: nginx/1.10.3
2020/01/22 19:01:21 [notice] 372#372: OS: Linux 4.19.93-v7
2020/01/22 19:01:21 [notice] 372#372: getrlimit(RLIMIT_NOFILE): 1048576:1048576
2020/01/22 19:01:21 [notice] 372#372: start worker processes
2020/01/22 19:01:21 [notice] 372#372: start worker process 380
[19:01:21] INFO: Running the deCONZ OTA updater...
[19:01:21] INFO: Running the IKEA OTA updater...
[19:01:21] INFO: Running the LEDVANCE/OSRAM OTA updater...
[19:01:21] INFO: deCONZ is set up and running!
[19:01:33] INFO: Successfully send discovery information to Home Assistant.
So I have tried reinstall and reflash hassio but nothing works. What can I do to further troubleshoot?
Cosby86 is correct. Do not mix up running Deconz as a hass.io addon and running it on bare metal.
What the hass.io addon actually does is to run the deconz-gui in a docker with a simple desktop environment attached to a vnc server.
When you run it on bare metal you can do the same. You can let the Raspbian auto login to the desktop and let the desktop environment start deconz-gui. This however takes more resources than running the simple deconz daemon without starting any desktop.
But yoh have to run one or the other. You cannot run both at the same time
When deconz-gui runs all the normal restAPI stuff and Phoscon are available. But with the deconz-gui it starts also the geeky desktop deconz app from where you can see the mesh and run manual OTA upgrades.
Your entire house and the connection from Home Assistant work exactly the same with the gui version. The only downside is that you have a 30 second interuption when you change between the two modes and that the GUI need time to refresh the mesh view. The Conbee stick remains active so the real mesh is not really affected by the change.
With a Raspberry Pi, I would always run with the non GUI under normal operation. It is really rare that you need to look at the mesh. Once things run who cares?
Ok, learned.
I’m on hassio. Thats not my game
So something happened - without doing nothing. I was able to pair a couple of sensors.
All of the sudden it stopped working again. I cannot pair anymore. Looking in devtools, I get this:
Does not seem right?
And the Hassio --> System --> System log says:
20-01-22 20:49:10 WARNING (MainThread) [hassio.api.ingress] Ingress for RDKGz5k5lAnoe5m8IRjI4I9zNB7ChNOr60wfnScMT9I not available
20-01-22 20:49:10 WARNING (MainThread) [hassio.api.ingress] Ingress for RDKGz5k5lAnoe5m8IRjI4I9zNB7ChNOr60wfnScMT9I not available
20-01-22 20:49:10 WARNING (MainThread) [hassio.api.ingress] Ingress for RDKGz5k5lAnoe5m8IRjI4I9zNB7ChNOr60wfnScMT9I not available
20-01-22 20:49:10 WARNING (MainThread) [hassio.api.ingress] Ingress for RDKGz5k5lAnoe5m8IRjI4I9zNB7ChNOr60wfnScMT9I not available
20-01-22 20:49:13 INFO (MainThread) [hassio.tasks] All core tasks are scheduled
20-01-22 20:49:13 INFO (MainThread) [hassio.core] Hass.io is up and running
20-01-22 20:50:53 ERROR (MainThread) [hassio.api.ingress] Ingress error: None
20-01-22 20:53:13 ERROR (MainThread) [hassio.api.ingress] Ingress error: [Errno None] Can not write request body for http://172.30.33.1:8099/api/BF77FEF3C6/config
20-01-22 20:53:23 ERROR (MainThread) [hassio.api.ingress] Ingress error: Cannot write to closing transport
```