[c-nsp] VRF question

[Per Carlson]

On 2004-10-08 18:42, dave bernardi wrote:
> So for those service providers running a (native?) MPLS backbone you'd 
 > likely carry your global internet routing table in a VRF

I'd not recommend running a global internet routing table in a VRF. In a 
VRF, all prefixes are prepended with a 64-bit RD creating 96-bit 
prefixes. Doing that with a full internet routing table will eat your 
memory.

> From what I can tell most IP features would still be available by the IP 
> edge functionality of your router (7600 platform) even if it is also the 
 > PE, correct?

That's very dependent of the type of cards you've got in the 7600. We
have seen examples where you can apply an ACL to a interface, but it's
not used at all (classic ge-wan osm).

> Depending on what IP features are lost it might be a better approach 
 > to run a hybrid MPLS backbone where you might leave Internet traffic
 > as IP switched (untagged) and label switch L2/L3-VPNs, for example.

There is no way to run some packages untagged and some tagged on the 
same interface. What you probably is thinking of is running the internet 
service in the global address space (one label) and L2/L3 VPNs on top of 
MPLS (two or more labels). This would work fine without any expansion of 
IP-adresses.

> How many of you are running a native MPLS backbone and was the major 
 > determining factor VPN, QoS, or other?

MPLS don't give you any (apparent) advantages beyond the ability to 
enable new types of services, like L2/L3-VPNs. That's the only reason we 
are using MPLS today.

Per