The new IETF activity aimed at supporting enterprise multi-homing of their Internet connections may be taking a step in a much more valuable direction in the long run. The Locator/Identifier Separation Protocol or LISP is a means of breaking the current link between identifying a user and identifying the user’s location.
LISP effectively provides a kind of hierarchical routing where the primary encapsulation of a packet gets it to the correct network, and from that network it’s then routed to the correct user address on that network. A user, under LISP, would be known by a global “address” independent of the network address, and one that wouldn’t even change if the user changed ISPs.
We believe this could have a profound impact on the way the Internet works and could make IP a more suitable protocol for “infrastructure” services, as well as making its operation less costly. However, it’s clear that there is a lot of opposition to LISP, even in the IETF, and so it’s not at all clear how soon it might be deployed beyond experiments.