I disagree, although I image we will never agree on this.
The devices in an MPLS router are called label switching routers for a
reason, and in the most common form of operation the LFIB is populated
with data derived from the IGP and EGP.
There are many ways of thinking of MPLS labels, but one model is to
consider them IP address equivalents, hence the use of the term
Forwarding Equivalence Classes.
- Stewart
On 06/12/2018 14:04, Joe Touch wrote:
Your right - it’s a broken L2.
Only routers are supposed to decrement TTLs. Links and tunnels themselves aren’t.
Joe
On Dec 6, 2018, at 5:16 AM, Stewart Bryant <stewart.bryant@xxxxxxxxx> wrote:
On 06/12/2018 05:22, Joe Touch wrote:
On Dec 5, 2018, at 9:01 PM, Christopher Morrow <morrowc.lists@xxxxxxxxx> wrote:
How is it, for example, different to put ipv6 packets into an MPLS path doing nothing along 'many' hops (except forwarding the packets along), and then once you pop out of the tunnel start processing the packet as you (joe) would want.
The hopcount doesn’t get decremented by L2.
Joe
MPLS is not L2.
MPLS has two modes, one in which the TTL of the IP payload is decremented on ingress and the TTL across the MPLS path is ignored. In the other mode, the TTL of the IP packet minus one is copied into the MPLS label which is then decremented as the packet travels across the network at egress from the MPLS layer the TTL is copied back into the the IP packet.
- Stewart
