One of the most interesting features i’ve noticed about LISP is the ability to encapsulate IPv6 packets and transport them between IPv4 only RLOC’s. This can be used for a transition strategy.
Click to access white_paper_c11-629044.pdf
The best parts are you don’t have to configure tunnel endpoints, it works in a multi-homed environment and supports communications between LISP and non-LISP sites (with a PxTR). The document describes this as ‘IPv6 islands over an IPv4 core’.
I read about LISP online and watched the Cisco Live presentation from Berlin 2017 (BRKRST-3045). I wanted to see this for myself.
The lab is direct from http://lisp.cisco.com/lisp_tech.html If you want to lab this, go grab the PDF. The lab is pretty good and gives you the chance to see LISP in action, all the configs and step by step guides are provided.
Topology.
Look at the packet from wireshark. This is a ping between R111 host and R117 Host. This was captured at the exit of core R113 (before decapsulation). It has the IP header from R112 (ETR) to R116 (ITR), next is UDP dst.port 4341 (LISP), next LISP Data, then the IPv6 header and ICMP message. So its easy to see how this works.
For traceroute , the only hops that are visible are IPv6.
Before I could connect to anything in Site2, the ETR (R112) had to request mapping information from the MS/MR, once it was returned R112 then knew how to forward traffic towards 2001:DB8:B:1::3. From R112 perspective it now has a map for the prefix 2001:DB8:B::/48 and the RLOC is 10.0.9.2
If you want to know what LISP can be described as read this article by Ethan Banks http://ethancbanks.com/2013/07/30/lisp-not-exactly-a-routing-protocol.
RH.
all-things-networking 