IP addresses and repeaters

[elorimer]

I have a wee suggestion about how IPs for devices on the far side of a repeater bridge get reported.

I have an RT-N16 with latest Merlin. I also have two RT-N10+ routers, with the lastest Asus firmware operating in repeater mode, which I use as client bridges; several wired devices plug into the ethernet ports. (The RT-N10+ routers can't operate just in client bridge mode.)

The RT-N10+ routers have static IP addresses, and I've reserved those static IP addresses to the wireless MAC address for each. The devices plugged in to these routers will pull an IP address from the RT-N16. This is reported in the client list as the wired MAC address for the respective RT-N10+, so I have three different IPs reporting the same MAC address.

I can, however, reserve an IP address for the wired MAC of the device and give it a name (thanks!!). (I can't do it by selecting the right device in the pull down menu, but I can do it by manually entering the MAC address.) Then the device will pull the reserved address, and in the client table and in the per-ip traffic monitor it will distinguish the device by name and IP address.

Since the RT-N16 is clearly keeping track of the MAC address of the device, I'm wondering if the client table should report the MAC address of the separate device, and not the MAC address of the repeater. Then it would be easier to reserve IP addresses and read the client table for unreserved IPs.

 

[cogumel0]

I am not too sure how this is handled in Asus firmware (or even Asuswrt-Merlin for that matter) but from what you are saying it sounds like the router creates a pseudo-bridge rather than a real bridge.

This explains pseudo-bridges quite well:

http://lartc.org/howto/lartc.bridging.proxy-arp.html

If that is the case, this is quite normal. AFAIK there's no real disadvantage to this but if you want a real bridge you'll have to use WDS I believe (which unfortunately is limited to WEP or nothing in both Asus firmware and Asuswrt-Merlin).

This being said, my understanding is that broadcom chipsets implement WET mode which should allow us to create a proper bridge, even if it appears to still have some caveats (see http://wiki.openwrt.org/doc/recipes/bridgedclient and http://wiki.openwrt.org/doc/howto/clientmode#why.it.works.on.brcm-2.4), though I'm not so sure how up-to-date that is.

In fact, I just bought a RT-N66U to act as a real non-pseudo bridge for my RT-AC66U but am on the same boat as you: anything behind the bridge reports as having the MAC of the bridge itself.