I am experiencing more and more internet connection issues. At these times, I can't ping 8.8.8.8 (via cable) for some mins, then it recovers automatically.
The #unificontroller tells me, the "Internet Health" is super green all the time.
So I am wondering if this status is just crappy or if the controller could do a better job here. No helpful logs there.
I think it is a provider issue, but what if not? Or just an internal #unifi #network issue?
Afaik #synology NAS are able to run #docker #container s. Right now I have some PIs laying around that host several things that could hopefully be migrated to such a NAS container.
I found already instructions on how to get for example #unificontroller #ubiquiti #unifi or the #homeassistant (HA) running.
Does somebody has noticed negatives by going this way?
Especially, the HA could be an issue, since it has this #zigbee antenna in one USB port.
Would be happy to hear your experiences on this.
I’ve finally found other references/threads that seem extremely close to my experience with UniFi in-wall access points and DNS.
https://www.reddit.com/r/Ubiquiti/comments/mcyk5y/dns_failing_until_ap_is_rebooted/
I suspect a secondary problem is that the UniFi in-wall units overheat. The dropout problem has been very bad this summer. It also seems to correspond to some of the warmest days. The UAP-U6IW dropped off for more than 24 hours over the weekend and was crazy hot when I finally got around to pulling it off the wall. I need to find a better way to monitor this, as I don’t see this info in the controller.
The logs on the affected APs still indicate a DNS resolution problem for the Inform URL. So, I overrode the Inform URL with the IP address of the controller. This also did not eliminate the problem.
I found that I had fewer dropouts of my UniFi in-wall APs, but this did not eliminate the problem.
I recently replaced Pfsense with Opnsense for two main reasons. I had wanted to try out and compare Opnsense. I was also hoping the different integration of DHCP leases with the local Unbound DNS would help my maddening UniFi AP dropout problem, since that seemed to hinge on DNS resolution of the UniFi Inform URL.
Since the two #Ubiquiti #UniFi #AccessPoints disconnected again, I tried more stuff.
- unplugged the APs
- tested and verified cables and switch settings
- changed another device on our home network so that 192.168.1.20 (#UAP default IP) was no longer in use
- bound proxy in front of #UniFiController container to an IP address and route
- ran “set-inform” twice with known good controller address, as apparently two inform addresses are stored by each AP.
Upgraded #opnsense, #UnifiController, #ProxmoxBackupServer, #ProxmoxVE,...
Also, I have problems connecting to the #UniFiController from a browser and the #UniFiApp so there remains a distinct possibility that this setup is not working smoothly yet. Not sure if I got the entire App setup in #TrueNASScale correct, especially with the traffic that is or is not sent through #Traefik.
I set up #UniFiController as an App in #TrueNASSCALE. I don’t know if it the controller is a problem, or if it’s just alerting me to the fact that two of my APs are constantly disconnecting/reconnecting, or a bit of both. Maybe I need to downgrade the firmware on the APs to see if that helps.
Even when opt'ed out of #ubiquiti #datacollection in the #unifiController , the #ap's are frequently trying to access trace.svc.ui.com. The opt-out ain't an opt-out, I guess the editing of the config is still necessary after years: https://community.ui.com/questions/Where-do-I-put-the-config-properties-file-to-opt-out-analytics-100percent/3192f835-8ec0-4d1d-a09c-5c75ee8e15dc. Or just use #pihole and block their stuff.
Also mein Ziel ist es alle meine #Dienste auf #MicroK8s zu migrieren...
im Moment habe ich schon ein paar interne Sachen(#LogitechMediaServer, #UnifiController) damit laufen...
Wenn ich das mit den Zertifikaten mal in den Griff bekomme soll meine #Nextcloud und #ejabberd auch umziehen...
