Fallback Bridging
First off, thanks to the two sites below, i finally learned what this beast was about. Thanks guys!
I got around to play with fallback bridging yesterday. I want to summarize its important points here.
To understand it in the first place, it helps to give some information on why its needed.
In using routed protocols, such as IP, we have the possibility of going across VLAN boundaries by using ip routing between vlan interfaces. Using non-routed protocols such as IPX, Appletalk and other legacy protocols, we dont have this option. Hence, if we want to be able to “speak” across VLAN boundaries, we need fallback bridging.
Now, for the platforms. The 3550 switch regards all non-IPv4 traffic to be a candidate for fallback bridging.
I state it this way, because the 3550 specifically, treats IPv6 traffic as non-IP traffic.
The 3560 switch is clever enough to treat IPv6 (and related traffic, such as NDP) as IP traffic, and therefore not bridge it between VLANs.
Everything regarding fallback bridging is configured using bridge groups. We apply these bridge groups to SVI’s and routed ports. Non-IP Traffic between these VLAN’s and routed ports can then occur.
Configuration wise, its very simple:
Cat-1(config)# bridge 1 protocol vlan-bridge Cat-1(config)# int vlan 200 Cat-1(config-int)# bridge-group 1 Cat-1(config)# int vlan 300 Cat-1(config-int)# bridge-group 1
Thats all there is to establish the bridge group and apply it to your VLAN’s.
Now after the initial configuration, there are a few things you can do to tweak access to the bridge. By default the switch accepts connections from all mac addresses learned from the two (or more) VLAN’s. You can disable this behavior like this:
Cat-1(config)# no bridge 1 acquire
After this, your bridge wont learn any addresses automatically anymore. You then setup static assignments, on whether you want to forward/deny specific mac entries. This is configured like so:
Cat-1(config)# bridge 1 address 0000.1111.1111 forward
The show command to verify your bridge is:
Cat-1#sh bridge 1 Br Group Mac Address State Type Ports -------- ----------- ----- ---- ----- 1 0000.1111.1111 Forward STATIC -
As shown, we have verified that we have statically assigned the 0000.1111.1111 mac address to be forwarded through the bridge.
To test out this configuration (without the static, and allowing the bridge to automatically learn the mac addresses), what you can do is set up two different VLAN’s, put a port into each, hook up a router to each one and configure IPX on each router:
R1(config)# ipx routing R1(config-if)# ipx network AAAA R1(config-if)# mac-address 0000.1111.1111 R2(config)# ipx routing R2(config-if)# ipx network AAAA R2(config-if)# mac-address 0000.2222.2222 R1#ping ipx AAAA.0000.2222.2222 Type escape sequence to abort. Sending 5, 100-byte IPX Novell Echoes to AAAA.0000.2222.2222, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 8/15/24 ms
And this is all successfull even though they are in separate VLAN’s.
I think thats all I have for fallback-bridging for now. Different stuff for me to say the least 🙂