Bridged Tomato network - problems with some devices on bridged wired segment

Question

This is my network:

Dotted lines indicate wireless connections, solid lines indicate Ethernet connections.

I used two ASUS routers and Tomato to bridge my network. So I have have two wired segments and a wireless network provided by the main router connected to my cable modem.

Everything seemed to be fine when I set up the bridge last week.

The main .1 router is set up for DHCP and virtually everything is DHCP with reservations by MAC address. The bridge has .2 assigned to it's LAN and it is set in wireless bridge mode. The .2 router is set in bridged mode and connects wirelessly to .1, providing connectivity for all the devices on its Ethernet ports.

But what I'm seeing today is that Pogoplug 2 (.195) cannot seem to get to the Internet.

It cannot ping 192.168.1.1.

It can ping .2 (it's bridge), .199 (wireless Mac), .197 (wired), .196 (wired) etc.

I am connected from .199 to .195 via SSH.

If I use Tomato's tools on the bridge, it can ping .1, 8.8.8.8 etc, so it seems to be fine.

What am I missing here? Seems like something simple.

Pogoplug2 ifconfig:

[root@alarm ~]# ifconfig
eth0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        inet 192.168.1.195  netmask 255.255.255.0  broadcast 192.168.1.255
        inet6 fe80::225:31ff:fe05:38bc  prefixlen 64  scopeid 0x20<link>
        ether 00:25:31:05:38:bc  txqueuelen 1000  (Ethernet)
        RX packets 169  bytes 20660 (20.1 KiB)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 89  bytes 11323 (11.0 KiB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0
        device interrupt 11  

lo: flags=73<UP,LOOPBACK,RUNNING>  mtu 16436
        inet 127.0.0.1  netmask 255.0.0.0
        inet6 ::1  prefixlen 128  scopeid 0x10<host>
        loop  txqueuelen 0  (Local Loopback)
        RX packets 4  bytes 260 (260.0 B)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 4  bytes 260 (260.0 B)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

A PC on the bridged part is fine - pings 192.168.1.1

OSX, Ubuntu, Windows devices on the bridged segment seem to be fine.

Pogoplug, Raspberry Pi have problems. Problem happens on either Tomato and DD-WRT firmwares. I have had a friend confirm what looks like the same problem with a FreeBSD device (but not a Windows client) on a bridged segment like this with Tomato on Cisco-Linksys routers.

Is this configuration or client defects or sensitivities or what?

Answer 1

A friend who happens to be a network administrator and Cisco/router person set up a system which demonstrates the problem and then sent me a solution:

You would think it would have been discussed out there more. Who knows? I really think it's a combination of bugs. When you setup the 2nd router as a wireless bridge, the 1st router should only see the MAC address of the 2nd router's WLAN. It should proxy-arp all clients on its side. The logic of this is explained decently in sections 16.3 and 16.3.1 on this site: http://lartc.org/howto/lartc.bridging.proxy-arp.html

At any rate, I have a workaround for you. On your 192.168.1.1 router, go to Administration -> Scheduler. Enable "Custom 1" and put "arp -d 192.168.1.195" (your PogoPlug IP) in the Command Box without the quotes. Set it to execute every 3 minutes of every day. Shouldn't be necessary, but it's been working for my FreeBSD box connectivity for over 45 minutes.

Answer 2

I've had this problem too, but with a specific PC. If I set the IP address of the problem machine statically, it worked. There wasn't anything visibly wrong.

However, I must point out that I switched about a year ago to DDWRT and haven't had the problem since. You could try that.