Furthermore I didn't say they were invincible. I just said they were invisible to the UDM-Pro's UI. Unless you have a blanket ban on outgoing LAN traffic, which would be absurd, there's no way to block access for a particular client or a particular destination address for that client.
In the case I gave, a Chinese robot vacuum with no on-device interface, please tell me how to find the IP of this robot, then block outgoing traffic from it, without SSH'ing into the UDM and running scripts. That's right, you can't, because the UDM-Pro doesn't support it.
> Unless you have a blanket ban on outgoing LAN traffic, which would be absurd, there's no way to block access for a particular client or a particular destination address for that client.
To the contrary; this is exactly what you should be doing. Isolated subnet for these untrusted devices. Block by default. (Whitelist only)
I used the word invisible to describe it missing in the ui. I used the word invincible to describe your lack of “management” (ie; blocking) of the device.
What I am trying to suggest, however, is that the UDM is likely not the root cause of these issues. I certainly don’t mean to suggest they are the best. The lack of compatibility of features between their product lines is a nightmare.
The default logging may not capture the individual child clients, depending on your configuration (eg double nat), sure... but those child clients are still entirely at the mercy of your configuration otherwise. Saying that the clients are completely invisible/invincible, and that the fault is the Ubiquiti product, is not true.