[warning : bad day today, now extremely heavily drugged up with new load of pain medication, so v fuzzy and. 😵 ]
I just an intriguing thought. Seeing as I happen to have a spare router, I have a spare Firebrick, an FB2500, I could perhaps put it to some good use. What if I plug a small switch into one of the modems? Then I run several further ethernet cables:
i. First cable goes from the new small switch back to the main Firebrick, to the WAN port that the modem is supposed to be connected to. That restores normal service from the modem.
ii. Second cable goes from the small switch to the secondary Firebrick.
iii. The secondary Firebrick now has a second cable to the main LAN large switch.
The secondary Firebrick picks up stuff from the modem and re-routes it, but also possibly uses its optional NAT rewriting abilities to change IPv4 addresses. It certainly can rewrite source addresses and ports but the worry is whether it can do the same for destination addresses. I could do with a little guidance here.
Between the secondary Firebrick and its modem let us use an IPv4 subnet of <whatever?>, perhaps an RFC1918 address range. We could perhaps just use the range that the modem defaults to for its IPv4 admin http interface. If I can get the Firebrick to receive incoming packets from the main LAN and route them to the modem, NAT'ing on the way, then a box on the main LAN could talk to the modem and monitor it.
Main LAN is a subnet that has a static public, globally-routable (non-rfc1918) ‘straight’ IPv4 /26, and for IPv6 it has a static, globally-routable, public /64.
I get the feeling that I’m being incredibly stupid here and all I'm really asking for is what a domestic CPE does anyway, with the modem being the ‘home LAN’ and my main LAN being the ‘internet’. Is that correct?
(Insanity check: I dont see the point of putting the modem into a /29 or smaller that lies inside the main LAN's /26 and getting the Firebrick to re-route from one port to another as I think ?I'm just going to confuse the Firebrick because of interfaces that have overlapping address ranges. I presume that would be madness?)
If I get this set up correctly, I could use the web admin i/f of the modem to see some numbers. I could perhaps telnet into it.
Wasn't there an extensive thread about exactly the latter done by one of our superb researcher kitizens some years back?
The Dlink DSL-320B-Z1 in question doesn't have Broadcom-type firmware in it, but I dimly remember that you can get the usual extensive range of stats out of the thing, albeit using commands that are perhaps somewhat differ [?] than those from the standard well-known Broadcom litany. But if memory serves our earlier researcher did a lot of work already in documenting them.
I don't have an existing library or software tool that I can use productively on the iPad that can speak telnet (nor even straight TCP). I have a Python development system for and interpreter for the iPad, but I have zero clue about Python and wouldn't know where to begin. I might be able to possibly find something from somewhere though. Shudders.
I could perhaps more likely write something for the Rasberry Pi if I ever get it going again, but I've never done any Linux C programming. I'm a big fan of the D language (a more advanced C++), I have two excellent D compilers, GDC (GCC family), LDC (LLVM family) and both produce superb code quality. Perhaps I can dig out some friendly networking library there.
But a box to talk to a modem and monitor it would be fun. I would need a UI of some sort.