AP stuck in adopting phase after a container restart

bartounet · 10 March 2022 09:20

Hello.
I have a little problem.

with each update of the docker unifi container, I lose my statistics
and I have to force the adoption of my unifi equipments so that it goes back to the controller with set-inform http://IP:8080/inform

I didn’t have this behavior before switching to docker

bartounet · 10 March 2022 21:06

This issue seems to be my problem

github.com/linuxserver-archive/docker-unifi

AP stuck in adopting phase after a container restart

opened 06:28PM - 22 Jan 18 UTC

closed 07:43PM - 23 Jan 18 UTC

EugenMayer

I am running several linuxserver and a gazzilion other docker images, so docker …should not be the actual issue. Logs seem fine, just starting up as expected. Initially i adopted my AP and first a little message about stun not available popped up but then disappeared again. Without restarting the actual docker-container now, everything works just fine. Unless i restart it, then the API stays in the adopting phase, never adopting .. as if the Controller has not proper key or what so ever. After some time the API is shown as disconnected, when i click on it, it starts adopting again. I exposed all ports: ![image](https://user-images.githubusercontent.com/136934/35236776-c8ed55da-ffa8-11e7-882b-f9cdd6de2f43.png) And i am using a host-volume mount for the `/config` partition: ![image](https://user-images.githubusercontent.com/136934/35236791-d99cc726-ffa8-11e7-8301-4a4a4acd2fba.png) I am running the latest image version Which is used, as i can see on the host. Using the controller on my mac worked just fine, so i expect a state issue in the docker-image... But well obviously i am not the only one using this image and others would have complained too. Looked at https://help.ubnt.com/hc/en-us/articles/204950304-UniFi-UAP-stuck-in-adopting-state but did not find anything particular helpful. Since now --host network is used is ask myself how sensitive the controller is the the broadcast not working ( if it is used ) ? what i have done to diagnose: 1. ping from AP to my synology NAS which is running the docker container: that one works, interestingly i cannot telnet to any port, not even 8443 2. `curl -L -k https://10.10.10.10:8443` worked from the AP 3. my networks router is an `opnsense` appliance, the AP / nas have both static leases and did not change their API during the "pairing"

But i have no option hostname controller on my controller

rob · 8 April 2022 23:53

If you’re losing your statistics its because you’re not saving them somewhere on your host and each time the container restarts, they go bye bye.

What did you use for this:

volumes:
      - <path to data>:/config

This will probably help the statistics issue.

The issue with the AP doesn’t seem to be an issue in my setup either. I have my docker container exposed on its internal 172.x.x.x:PORT. The host of that computer is on a 10.x.x.x network. I point my set-inform to the 10.x.x.x:8080/inform with no issues. What are you using?