I have read a number of sources that say to run the PPPoE server on an interface without an IP address. I assume that this is a safety mechanism so the interface isn’t broadcasting IP information and the client can’t stiff the interface for broadcasts. The only problem for me is that I don’t understand how I would access the CPE device if the interface it was linked to had no IP address. I have mikrotik and ubnt CPE devices.
Google and a forum search doesn’t return anything, but I might be asking the question wrong. Any help would be appreciated.
The “PPPoE interface” (not the physical interface) DOES have an IP address (the “local” address in the pppoe profile). So, as long as the CPE is connected with PPPoE, you can access it. If the CPE was NOT connected with the PPPoE, then of course you could not access it… and that is the whole safety point.
But how do I access the CPE device if PPPoE is an issue? It has no IP so I am stuck with layer 2. I can live with that on mt but I also have ubnt hardware. Without a local IP I can’t connect to it to troubleshoot.
Correct, if it is not connecting to PPPoE, then you can not access it.
However, if it is not connecting to PPPoE and you have it set to always connect to PPPoE, then there is probably something wrong that you would not be able to access it remotely anyway even if you had IP access (unplugged, locked up, storm damage, etc…)
Alternatively, you could set an IP on the physical interface of your main router, and just enable/disable the physical IP when you really need it.