IP CEF w/ MPLS

accely

I just started the MPLS nuggets from the ISCW series and he talks about how MPLS is kinda like CEF for everyone else not running Cisco equipment.

When he talks about turning MPLS on and configuring it, the first step is to make sure CEF is on. I guess I'm confused because if you're not using Cisco equipment, then how could you turn CEF on? and then why would you need MPLS if you have CEF?

I'm glad I finally get to learn more about MPLS since some of the customers we support use MPLS through AT&T

kalebksp

CEF is required for MPLS on Cisco equipment, but not for other vendors. I'm not familiar with other vendors MPLS implementation, but they may have a CEF-like feature as well.

MPLS is like CEF in that it generates a table with mappings from incoming labels to outgoing labels and next hop, whereas CEF generates a table mapping the incoming packets destination to the outgoing interface and next hop. Both are done based on the routing table and generated on start up, allowing very fast switching of packets.

CEF and MPLS work together on edge routers. On the ingress edge router the IP destination network of an unlabeled packet will be looked up in the CEF table which will hold a mapping to the outgoing label. This is for efficiency so that the destination doesn't have to be looked up in the CEF table then again in the LFIB. On the other side PHP will likely be preformed before the packet reaches the egress router, so a LFIB lookup won't be necessary.

As for why you would use MPLS at all, there are many things that can be done on MPLS that can't be done or would be more difficult with IP routing. The major application is MPLS VPNs, which is likely what AT&T is providing to your customers.

accely

kalebksp wrote: »

CEF is required for MPLS on Cisco equipment, but not for other vendors. I'm not familiar with other vendors MPLS implementation, but they may have a CEF-like feature as well.

MPLS is like CEF in that it generates a table with mappings from incoming labels to outgoing labels and next hop, whereas CEF generates a table mapping the incoming packets destination to the outgoing interface and next hop. Both are done based on the routing table and generated on start up, allowing very fast switching of packets.

CEF and MPLS work together on edge routers. On the ingress edge router the IP destination network of an unlabeled packet will be looked up in the CEF table which will hold a mapping to the outgoing label. This is for efficiency so that the destination doesn't have to be looked up in the CEF table then again in the LFIB. On the other side PHP will likely be preformed before the packet reaches the egress router, so a LFIB lookup won't be necessary.

As for why you would use MPLS at all, there are many things that can be done on MPLS that can't be done or would be more difficult with IP routing. The major application is MPLS VPNs, which is likely what AT&T is providing to your customers.

Thanks!! This is clear and I actually understand it all which is even more awesome

And after watching the 3rd video, MPLS VPNS, this definitely makes more sense.

kalebksp

No problem, glad I could help.

cisco_trooper

Man, I hate to look like a douche by posting a link to my own site but it's pretty relevant. I wrote two or three MPLS articles on my CCIE blog. This one is probably the best for you -> Cisco CCIE CCIP CCNP CCNA Certification: Multiprotocol Label Switching Part II - Frame Mode MPLS Configuration