Thanks for the question Yue. I included the diagram so others can follow easier:



> How the 1st sequence of the packet flow (1st paragraph on Page 7) can be 
> achieved please?

Device 1 is sending packets no matter where it is connected. So it will use its 
default router to send packets. The first-hop router in Site D will typically 
have the same Mac address of the default router used in in Site A. 

> The Device 1 keeps its EID=1.0.0.1, but the prefix of Site D is 2.0.0.0/24. 
> Since 1.0.0.1 is not belongs to 2.0.0.0/24. How the packets can be exchanged 
> between xTR D and Device 1?

xTR D discovers EID 1.0.0.1 when Device 1 sends packets. Or, if by chance, xTR 
D is directly connected to where Device 1 moved to, can use ARP/ND (if Device 1 
felt the need to send an ARP request during the move). But xTR D will see 
packets from 1.0.0.1, apply policy to either permit or deny the roaming of 
1.0.0.1, and if permitted, registers EID 1.0.0.1/32 with RLOC IP_D to the 
mapping system. 

When Device 1 is not directly connected to xTR D, the first-hop-router learns 
1.0.0.1 (when Device 1 sends an IP or ARP packet) and must inject the route 
1.0.0.1/32 into the IGP so xTR D can learn about the existence of 1.0.0.1 so it 
can register it.

Does that help?

Dino



_______________________________________________
lisp mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/lisp

Reply via email to