T1 to E1 Problems

[mlinman]

If anyone has any input it would be a great help.
Have a T1 circut in Chicago to E1 in London.
The circut is from Qwest, but goes Qwest, Teleglobe to BT.

Chicago side:
Cisco 2600
T1 WIC

London side
Cisco 2600
PowerQUICC Serial
X.21 cable to BT NTU

We currently have Serial up Line protocol down. If telco puts a loop facing Chicago anywhere in the line the Chicago router goes Serial up Line protocol down (looped).
Any loops sent to the London router are not seen. If London puts a hard loop on the NTU in their office the Chicago router again shows Serial up Line protocol down (looped).

Since the Chicago router can see the loop on the NTU does this prove the line is good? Also the NTU is putting out 2.048 to the London router, could that be the problem?

How should the London router be configed, I've never done any X.21 stuff before.

Thanks for any help.

 

[wybnormal]

If the Chicago router sees the NTU in London, then the raw link is good TO THAT POINT. At least the layer2 is mostly functioning at this point.

Have you run any debugs on the London serial port to see what it's whinning about? What about the Chicago side? Error lights on the serial port of the London router? Has the cabling been checked to make sure it's the proper cabling?

MikeS
Find me at

http://www.packetattack.com

"Take advantage of the enemy's unreadiness, make your way by unexpected routes, and attack unguarded spots."
Sun Tzu

 

[mlinman]

Thanks for the reply.

The router in London has been swapped out as well as the cable to the NTU several times.

Chicago side shows no errors, London side shows framing and crc errors.

 

[wybnormal]

This is why I asked about the debug. We need more detailed information about the link itself and debuging the serial port will help with this. Framing/CRC can be caused by serveral different things... NTU is not configured correctly, bad line(unlikely if it's been swapped), incorrect framing, incorrect Telco config and so on.

Here are some links that may help

E1 Layer 1 Troubleshooting

http://www.cisco.com/warp/public/116/E1_layer1.html

E1 Error Event Troubleshooting Flowchart

http://www.cisco.com/warp/public/116/e1_flchrt_error.html

Troubleshooting Serial Lines

http://www.cisco.com/univercd/cc/td/doc/cisintwk/itg_v1/tr1915.htm#xtocid195570

MikeS
Find me at

http://www.packetattack.com

"Take advantage of the enemy's unreadiness, make your way by unexpected routes, and attack unguarded spots."
Sun Tzu

 

[LiNuXGuRL]

they should try to give you a hardware loop from the modem/ntu in London. also, i would change the config from down when looped to default (no down when looped), like this you can ping your own interface. a overseas line runs across several different components and hubsites. if and e2e test of the line is not possible i would have the telco test every single part of this line. if all they are doing is loop tests, at least request hardware loops because they are more accurate. crc and frame errors indicate a problem in the lower layers, however, if your line prot is down and no data is being sent you should not see any data.... as far as the congi on the london router with an x.21 int. if it shows line prot up (after taking out down when looped...you can put back in later, for pro active monitoring if you got snmp checking the link) when the external modem is looped you have already counted out the cabling, the physical router interface and the modem itself. they should also try to give you a loop from the closest telco hub to london, in direction london router. the configured bandwidth does not make a difference for the lower layers, but it matters for routing calculations. if this is a static route link it does not matter. there could also be a problem with the clocking of the modem. both routers should be set as dte...no clockrate configured.

 

[CCIEWANNABE]

framing and crc errors tells me you have a bad cable on the london side. Route once; switch many

 

[CCIEWANNABE]

Check to see if the maximum line length has not been exceeded. Or one of the following: Loose or unconnected shield ground wires, Bridge taps, Timing problems, or Electrical interference. Elevator room, or a possible florescent fixture within four feet etc. The list can go on an on. Route once; switch many

 

[wybnormal]

LiNuXGuRL is on a good track with banging on the telco's to test harder. The normal loopback test the telco's like to do it virtually useless. I have seen their *test* pass until I press them to run all ones and zeros testing along with marching ones. All of sudden I get a call saying it's fix.. a bad card at the switch was found. Very common in telco. I had one circuit where it failed every day about the same time.. turned out a sprinkler would hit the telco's box everyday and the box seal was bad. Water would get in and short out the lines. That was a HUGE pain since by the time they tested each day, it would be working again till the next day.

MikeS

Find me at

http://www.packetattack.com

"Take advantage of the enemy's unreadiness, make your way by unexpected routes, and attack unguarded spots."
Sun Tzu

 

[mlinman]

Thanks for all the tips. Since this is a T1/E1 conversion telco is now telling us we need something to channelize the circuit on the E1 side. So we are changing out the X.21 card in the Cicso and getting a NM-1CE1U card so we can take the E1 into the router.