cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
453
Views
0
Helpful
8
Replies

bridging question

sweston
Level 1
Level 1

I have configured a pair of 1841 routers with WIC T1's for bridging between 2 small offices to extend their network. I can browse the network from either location, telnet out the remote gateway at office 1 from the branch. However when attempting any HTTP traffic the browsers just spin their wheels and eventually error out. I can ping and traceroute just fine. I have tried firefox and ie both with the same results.

Any ideas? Config below pretty much the same on both ends.

interface FastEthernet0/0

description $ETH-LAN$$ETH-SW-LAUNCH$$INTF-INFO-FE 0$

no ip address

no ip route-cache

duplex auto

speed auto

bridge-group 1

!

interface FastEthernet0/1

no ip address

no ip route-cache

shutdown

duplex auto

speed auto

!

interface Serial0/0/0

no ip address

encapsulation lapb multi

no ip route-cache

bridge-group 1

!

ip classless

no ip http server

!

!

!

control-plane

!

bridge 1 protocol ieee

8 Replies 8

rais.ahmad
Level 1
Level 1

Can you ping across using a 1500 byte packet?

Thanks.

I can ping with a 1458 byte packet but a 1459 is the cutoff and stops.

I think the lapb multi accounts for the extra 2 bytes. What is the MTU on the S0/0? (Look for it in the show int). It is probably 1500. Maybe you simply need to increase it by 2. mtu 1502.

Carefully though, 'cos it might cause a brief interruption in service, depending on your router. (From experience ;-) I would wait for a maintenance slot.

Kevin Dorrell

Luxembourg

Thanks will try that. This is not in production yet and is still in my shop. We deployed and found these problems and pulled it back. We are still a week ahead of schedule so it's not big deal. If required to do so we will route a diff subnet instead of bridging but was trying to avoid doing so for just 2 workstations. Will try increasing the MTU and post results.

I tried increasing the MTU but did not work per your suggestion. Then out of curiosity I tried HDLC encapsulation instead and everything works great.

I didn;t use it to begin with because it was not mentioned in this doc I found:

http://www.cisco.com/univercd/cc/td/doc/product/software/ios122/122cgcr/fibm_c/bcfpart1/bcftb.htm

Thanks for leading me in the right direction though as the encapsulation was the cause of the problem.

Do you forsee any problems in bridging using HDLC encap?

I had wondered about the choice of lapb multi when I first saw your post and assumed that you had some particular reason to need it. But I see now that was not the case.

I think you will do very well with HSLC encapsulation. I do not see any problems to be likely. HDLC is the default encapsulation. The lapb multi encapsulation is fairly rarely used. I am confident that HDLC is a better choice.

HTH

Rick

HTH

Rick

I agree, HDLC is normally the one to go for.

I must confess, I was mislead by the multi keyword, and I thought it was referring to multilink LAPB, which would have entailed an extra multilink header. However, now I read the documentation (which is what I should have done in the first place), I see that it refers to carrying multiprotocol payloads over LAPB. Nevertheless, as it does multiplex the protocols, it still entails an extra two-byte protocol tag header.

http://www.cisco.com/en/US/products/sw/iosswrel/ps1835/products_command_reference_chapter09186a00800b5fa0.html#wp1017836

Anyway, I would stick to the HDLC. With LAPB you have to think about which end is logical DCE and which is logical DTE. With HDLC, you don't have to consider that.

Kevin Dorrell

Luxembourg

You have mtu issues then. Your 1500B ping should succeed. I believe you can resolve it using tcp mss.

http://www.cisco.com/en/US/products/sw/iosswrel/ps1839/products_feature_guide09186a00804247fc.html

You can also try clearing the df-bits:

interface F0/0

ip policy route-map clear-df

!

route-map clear-df permit 10

match ip address 101

set ip df 0

!

access-list 101 permit tcp any any

HTH