Concurrent Routing & Bridging

Answered Question
Apr 13th, 2012

Hi,

I've done a test in the following topology:

R1(10.10.20.10)-----------------(f0/0)R3(s0/0)-------------------(s0/0)R4-----loop0(4.4.4.4/24)

R3 is a C3725 w/ IOS 12.4(21a) configured for CRB with f0/0 and f0/1 in bridge-group 1

R3:

.......

no ip cef

!

bridge crb

!

!

interface FastEthernet0/0

ip address 10.10.20.3 255.255.255.0

no ip route-cache

speed 100

full-duplex

bridge-group 1

!

interface Serial0/0

ip address 1.1.1.3 255.255.255.0

no ip route-cache

ip ospf 1 area 0

clock rate 2000000

!

interface FastEthernet0/1

no ip address

speed 100

full-duplex

bridge-group 1

!

!

bridge 1 protocol ieee

...

R3#sh int f0/0 crb

FastEthernet0/0

Bridged protocols on FastEthernet0/0:

  appletalk  clns       decnet     ip

Software MAC address filter on FastEthernet0/0

  Hash Len    Address      Matches  Act      Type

  .....

R3#

R3#

R3#sh int f0/1 crb

FastEthernet0/1

Bridged protocols on FastEthernet0/1:

  appletalk  clns       decnet     ip

Software MAC address filter on FastEthernet0/1

  Hash Len    Address      Matches  Act      Type

  ...........

R3#

R3#sh ip route 4.4.4.0

Routing entry for 4.4.4.0/24

  Known via "ospf 1", distance 110, metric 65, type intra area

  Last update from 1.1.1.4 on Serial0/0, 00:22:23 ago

  Routing Descriptor Blocks:

  * 1.1.1.4, from 1.1.1.4, 00:22:23 ago, via Serial0/0

      Route metric is 65, traffic share count is 1

R4#sh ip route 10.10.20.0

Routing entry for 10.10.20.0/24

  Known via "static", distance 1, metric 0 (connected)

  Routing Descriptor Blocks:

  * directly connected, via Serial0/0

      Route metric is 0, traffic share count is 1

R4#

As you can see IP protocol is bridged (only) on f0/0 and f0/1. However I can ping from R1 to loopback 4.4.4.4 configured on R4:

R1#ping 4.4.4.4 rep 2

Type escape sequence to abort.

Sending 2, 100-byte ICMP Echos to 4.4.4.4, timeout is 2 seconds:

!!

Success rate is 100 percent (2/2), round-trip min/avg/max = 108/168/228 ms

R1#

Why this behaviour on R3 ?

thanks

I have this problem too.
0 votes
Correct Answer by Richard Burts about 2 years 3 days ago

Carlo

You are correct that within a particular bridge group all the interfaces need to treat a protocol the same way. Within the bridge group if IP is bridged on one interface that it should be bridged on all interfaces. Of course a protocol like IP can be bridged in one bridge group and routed in another bridge group - or bridged in the bridge group and routed on interfaces not in a bridge group.

I would also mention that CRB was an early effort to provide a combination of bridging and routing of a protocol on IOS routers. Cisco pretty quickly came up with IRB Integrated Routing and Bridging which provides a much more flexible solution for combining bridging and routing of a protocol on a router, and provides the ability to have IP traffic from the bridged environment to be forwarded to the routed environment, which is a capability that does not exist in CRB.

HTH

Rick

  • 1
  • 2
  • 3
  • 4
  • 5
Average Rating: 5 (1 ratings)
Richard Burts Fri, 04/13/2012 - 10:33

Carlo

I am not clear what you are trying to do with CRB and also not clear what behavior you expected. Packets from R1 headed for R4 pass through R3 and to me that would be the expected behavior. So I wonder what in it surprised you?

I will also note that the configuration on R3 is inconsistent with the implementation of CRB. If the FastEthernet and the Serial interfaces are in the same bridge group then logically their IP addresses should be in the same subnet. Your configuration has them in different subnets which is more consistent with routed interfaces than with bridged interfaces.

HTH

Rick

ccianfarani Fri, 04/13/2012 - 11:59

Rick,

in this R3 CRB configuration f0/0 and f0/1 belong same bridge-group (bridge-group 1) (s0/0 is not configured in any bridge-group !)

AFAIK (see for example http://www.netcraftsmen.net/resources/archived-articles/428.html ) without bridge 1 route ip command (see R3 config) the default behaviour for bridge group 1 is to perform only IP bridging between bridged interfaces (f0/0 and f0/1). So I expect IP packets (ping) entering from f0/0 (targeted for 4.4.4.4) shouldn't routed by R3 (basically for this IP packets R3 shouldn't perform IP lookup in RIB/FIB)

Does it make sense ?

Richard Burts Sat, 04/14/2012 - 10:38

Carlo

When I read the original post I though that I saw the bridge-group on the serial interface. If I mis-read the post then I am sorry.

The article that you reference is very nice, and was written by a colleague of mine, so thanks for pointing it out.

I believe that the explanation for the behavior that you see is in the fact that an IP address is configured on one interface and not on the other interface belonging to the bridge group. I found this explanation about CRB on the Cisco web site:

All of the interfaces or subinterfaces belonging  to a specific bridge-group need consistent configuration in regards to  configuring or not configuring IP addresses. Mixing interfaces  configured with IP addresses and interfaces not configured with IP  addresses in the same bridge group can cause inconsistent or  unpredictable routing at the network level.

Here is the link if you want more details:

http://www.cisco.com/en/US/partner/docs/optical/15000r9_0/ethernet/454/guide/45490a_configbridges.html

HTH

Rick

jpvh12345 Sun, 04/15/2012 - 06:47

Richard is correct. What is behind fa0/0 and fa0/1 on r3 that would make you want to bridge those interfaces?

Sent from Cisco Technical Support iPad App

ccianfarani Mon, 04/16/2012 - 00:18

It's only an 'experiment'  to see how CRB works....

So, if I understand correctly from raccomandation, a certain network protocol (IP,IPX ecc..) has to be bridged OR routed in the same way on ALL interfaces belonging to a CRB bridge-group..(in other words in a bridge-group is not possibile, for instance, route IP on some interfaces and bridge it on others) ??

thanks

Correct Answer
Richard Burts Mon, 04/16/2012 - 09:21

Carlo

You are correct that within a particular bridge group all the interfaces need to treat a protocol the same way. Within the bridge group if IP is bridged on one interface that it should be bridged on all interfaces. Of course a protocol like IP can be bridged in one bridge group and routed in another bridge group - or bridged in the bridge group and routed on interfaces not in a bridge group.

I would also mention that CRB was an early effort to provide a combination of bridging and routing of a protocol on IOS routers. Cisco pretty quickly came up with IRB Integrated Routing and Bridging which provides a much more flexible solution for combining bridging and routing of a protocol on a router, and provides the ability to have IP traffic from the bridged environment to be forwarded to the routed environment, which is a capability that does not exist in CRB.

HTH

Rick

Actions

Login or Register to take actions

This Discussion

Posted April 13, 2012 at 8:52 AM
Stats:
Replies:6 Avg. Rating:5
Views:618 Votes:0
Shares:0
Tags: No tags.

Discussions Leaderboard

Rank Username Points
1 15,007
2 8,150
3 7,730
4 7,083
5 6,742
Rank Username Points
160
77
70
69
50