Unifi Adoption Loop after Update

I finally updated the switches of one of the installations I manage to the latest Unifi firmware, and I’ve got about 8 devices that are in an adoption loop. They currently keep their existing configuration, so they’re technically up, but the controller is showing them in a constant loop. I don’t want to do a hard reset until the business closes tomorrow, and I’m hoping my weekend isn’t shot trying to figure it out.

This network has 20 devices, and I’m only showing a screen clip of the last few, as the others are currently up and not toggling.

Any ideas on where to look and how to troubleshoot would be greatly appreciated.

Unifi documentation says to factory reset and update the firmware. When I have run into this in the past with Unifi APs I have found a factory reset (of the AP) to be a quick fix. Assuming your devices are now on the latest firmware, what Unifi Network version are you on?

And now I get to track down another issue. It turns out that I had a multi-cast flow that is only enabled when equipment is on (and that’s only for four hours each week, and I happen to not look at the unifi controller during that time). That equipment was on last night, as we (me and the other group working with the audio equipment) were both working after hours to minimize customer impact.

That equipment is putting out four multi-cast dante streams, for a total 24Mb of traffic. Not a huge deal – it’s a gig network. But IGMP was only working for the SG300’s that the equipment is directly plugged into. There are two other devices that “woke up” and started sending their information to that network, but they’re on a two USW-PRO 24p switches.

Long story short, the Unifi switches are erroneously passing 12Mb of data to the pfSense, and although no traffic flow is showing up (according to the pfSense traffic monitor) on the interface that has the unifi controller, that 12Mb of traffic is causing 6-8 devices to lose connection and adoption to the unifi controller. If I cut those devices off, everything is happy. Put them back on, and unifi controller gets upset again.

This is confusing and frustrating on so many levels. Now I’ve got to figure out why the traffic is going through pfSense in the first place. I need to figure out why it’s impacting something on a seemingly separate interface, and I need to find out how 12Mb of traffic is enough to cause issues for unifi. It’s not supposed to ever see anything but the switch management traffic, but even if I had it all on one network, I wouldn’t expect a controller to have issues with 12Mb of traffic.

Sometimes this job isn’t fun…