[c-nsp] Basic understanding of 6PE and 6VPE

[Saku Ytti]

On (2012-06-26 09:23 +0200), adam vitkovsky wrote:

> Regarding the TTL hiding it's new to me that some Customers would like to
> see all the MPLS hops 
> However I guess if the network runs as supposed to the Customers wouldn't
> mind the missing extra hops in their traceroutes

If you are global/tier1 ISP mostly selling IP transit, it would be nice to
know more than packet ingressed in Finland and egressed in Pakistan.
Obviously the RTT will be incorrect, unless you pop them during transit,
which implies INET and not BGP-free core. The incorrect RTT can negatively
affect business of residential connection providers, as pseudo-clued gamers
might think you have 50ms latency inside two pops in city, when going to
another country (we did get also these complains when we didn't hide core).

So I can relate to Gert's problem also. I think it all depends on what you
are selling and to who.

Technically it would be possible to implement conditional core hiding. So
you could specify based on source IP when to hide or expose core (or when
not to).
Or heck, even based on what communities source IP has (like SCU type
mechanism in JunOS).
You could then decide that you don't hide core when you do traceroute from
your NMS network. Or maybe you'll hide it for VPN customers also.

-- 
  ++ytti