"No deConz bridge discovered" ..... can't add my Conbee connected devices (HASSIO 0.81.1

With updating from 0.81 to 0.81.1dev (HASSIO) my deconz configuration disappeared. Now when I click on the phoscon app “Authenticate app” and before 60 seconza I try to do the doceonz integration, I receive the message in subject, hence I can’t see my devices connected to the Conbee stiock.

Any clue?

Can you enable debug and print corresponding logs

Of course, what shall I put in the log option, as below?

logger:
  default: error
  logs:
    homeassistant.components.deconz: critical

Check component documentation

2018-10-30 18:50:16 DEBUG (MainThread) [pydeconz.utils] Sending {} to https://dresden-light.appspot.com/discover

2018-10-30 18:50:20 ERROR (MainThread) [homeassistant.helpers.service] Error rendering data template: UndefinedError: ‘None’ has no attribute ‘state’

2018-10-30 18:50:21 ERROR (MainThread) [pydeconz.utils] Error getting deCONZ data from https://dresden-light.appspot.com/discover.

2018-10-30 18:50:21 INFO (MainThread) [pydeconz.utils] No discoverable bridge available.

if I click above I receive correctly

[{“macaddress”: “00212EFFFFxxxxxx”, “name”: “zigbee”, “internalipaddress”: “192.168.1.238”, “publicipaddress”: “xx.xx.xx.xx”, “internalport”: 80, “id”: “00212EFFFFxxxxxx”}]

Then it is probably something in your environment. You can try to enable the discover component to trigger a different discovery of deconz.

I will definitely need to add manual input of address…

You wrote the component?

Anyway it was working fine in 0.80.3, although it would forget entity_id’s after a reboot (had to “authenticate app” again and configure deconz in HASS again, and everything would restore immediately).

After updating to 0.81 I don’t think I did anything to my environment

Other people have mentioned that behavior. Do you happen to have many components and a slow pi?

The hypothesis is that a timeout gets hit and the component doesn’t retry connecting. Im working on retry functionality

Many components yes, I run HASSIO on a NUC, but I do see strange/slow behaviour sometimes

Is there anything I can do to try to make it work?

I already have discovery: in my configuration (had it before also)

The fix with tests are done, I will merge it to Hass dev prior to the first beta this week.

There is nothing else that can be done today. You can try re-register deconz or restart the system again and hope it connects. I actually got this behaviour myself this morning with the 81.2 upgrade. It worked after a restart.

Well I manage to understand what it was. It seems I had a DNS iissue in hassio_supervisor (don’t tell me how I got it, I still do not understand how).

Basically hassio_supervisor was not resolving DNS names, funny that only few components (including deconz) were not able to resolve it but some would … very strange. Now it works

Thats weird. How did you solve it?

I applied so many settings that I am not sure whioch one solved the issue. Some settings are related to my environment, anyway some good information I got from this post

Hey, the current beta fixed this issue to me, thanks! Now I dont have to reconnect to get back my deconz devices, so after a restart I can see all of them.

Awesome to get it verified! Thanks for the feedback!