On the Scalability of LISP Mappings Caches

TL;DR

This paper presents an analytical model demonstrating that LISP cache scalability remains constant despite Internet growth, validated by real packet traces, ensuring efficient edge router operation.

Contribution

It introduces a new analytical model for LISP cache scalability based on long-term popularity and validates it with real-world data.

Findings

LISP cache scalability is O(1) with respect to Internet growth.

Long-term popularity distribution influences cache performance.

Model validated using packet trace data.

Abstract

The Locator/ID Separation Protocol (LISP) limits the growth of the Default-Free Zone routing tables by creating a highly aggregatable and quasi-static Internet core. However, LISP pushes the forwarding state to edge routers whose timely operation relies on caching of location to identity bindings. In this paper we develop an analytical model to study the asymptotic scalability of the LISP cache. Under the assumptions that (i) long-term popularity can be modeled as a constant Generalized Zipf distribution and (ii) temporal locality is predominantly determined by long-term popularity, we find that the scalability of the LISP cache is O(1) with respect to the amount of prefixes (Internet growth) and users (growth of the LISP site). We validate the model and discuss the accuracy of our assumptions using several one-day-long packet traces.