Cisco Support Community
cancel
Showing results for 
Search instead for 
Did you mean: 
Announcements

Welcome to Cisco Support Community. We would love to have your feedback.

For an introduction to the new site, click here. If you'd prefer to explore, try our test area to get started. And see here for current known issues.

New Member

AToM ATM pvc endpoints

At my place of employment we have been implementing AToM in order to back haul some legacy ATM network through the MPLS cloud until we can push the IP termination points to the appropriate market aggregation routers. My Question is in regards to dissimilar VPI/VCI on the xconnect endpoints, while using aal0 encapsulation. We have some xconnects that were built with different vpi/vci on two of our MPLS enabled routers. I wouldn't have thought that these xconnects would successfully pass traffic with aal0 encapsulation; however, they have been up and working for quite awhile now. There is concern within our organization that this could be a bug, and a code upgrade could render all of these non-functional. I have included sample configs to help illustrate this. I would appreciate any help understanding what may be going on here. Perhaps more importantly, determining if this is correct behavior for Cisco Routers or if we may have stumbled upon a bug.

Thanks,

RouterA#

interface Loopback0

description MPLS-Loopback

ip address 2.2.2.2 255.255.255.255

no ip redirects

no ip unreachables

no ip proxy-arp

ip pim sparse-dense-mode

no clns route-cache

end

interface ATM1/0/0.84 point-to-point

description client 444

bandwidth 256

no atm enable-ilmi-trap

no snmp trap link-status

pvc 0/84 l2transport

encapsulation aal0

xconnect 1.1.1.1 444 encapsulation mpls

!

end

RouterB#

interface Loopback0

description MPLS-Loopback

ip address 1.1.1.1 255.255.255.255

no ip redirects

no ip unreachables

no ip proxy-arp

ip pim sparse-dense-mode

end

interface ATM2/1/1.186 point-to-point

description client 444

bandwidth 256

no atm enable-ilmi-trap

no snmp trap link-status

pvc 0/186 l2transport

encapsulation aal0

xconnect 2.2.2.2 444 encapsulation mpls

!

end

It may also be pertinent to mention that these endpoints connect to ATM switches in each market that are configured with the respective vpi/vci for that xconnect endpoint.

(IP/Frame/CPE)-->(Frame/ATM switch)-->(MPLS/PE/xconnect)--CORE-->(MPLS/PE/xconnect)-->(ATM switch)-->(nonMPLS/IP/router)

  • MPLS
3 REPLIES

Re: AToM ATM pvc endpoints

Hi,

As far as i know there are no restrictions on this point, you can use different VPI/VCI on both sides in VC mode, as soon as the VCID is the same.

Even if you use MPLS cell-packing to insert multiple concatenated ATM cells in an MPLS packet.

We already implement ATM over MPLS with different VPI/VCI's on 7200 Series with different NPEs and different IOS S train versions and it's working great.

Please rate helpful posts.

Best Regards,

Mounir Mohamed

New Member

Re: AToM ATM pvc endpoints

Thank you Mounir,

You have helped me to confirm some of my suspicions. I'm glad to hear that we are not the only ones successfully using this scenario. Incidentally, most of these xconnects exist between 7613 and 7606 routers running different IOS code. We have had cases where these types of xconnects did not work and rebulding them with matching vpi/vci seemed to have resolved the issue. This presents a different question. Perhaps we shouldn't be asking ourselves why differing VPI/VCIs worked, but why, in some cases, they don't? I would be curious to know what the order of operations is within the router logic in regard to ATMoMPLS using the AAL0 ecapsulation.

Re: AToM ATM pvc endpoints

Hi,

Well this should not happen unless there is misconfigurations or specific IOS limitations across your gears.

Look there are some limitations for MPLS AToM Implementations across different platforms and of interfaces types for specific series, for example not all SIP types support all ATOM L2VPN features, for example Cisco 7600 SIP-400 Support 4 SPA Bayes and up to 4G bandwidth with VPLS/AToM/GRE/CBWFQ/WRED, also you may configure OAM so the PVC doesn't get up because it's supported over AAL5 on 7200 and 7500 only.

Does there is any OAM configurations?

Would you provide more details about such unsuccessful test (Series / IOS / PA or SPA ...etc)?

Can you send the config with some debug commands.

R1#debug mpls l2transport vc event

Please rate helpful posts.

Best Regards,

Mounir Mohamed

197
Views
7
Helpful
3
Replies
This widget could not be displayed.