NDN Log Analysis Using Big Data Techniques: NFD Performance Assessment

Named Data Networking (NDN) is presented as the future Internet Architecture. It brings several properties, including security and content dissemination. The measurement of such properties needs a deep analysis of forwarding behavior during the runtime. The analysis presented in this paper stands on two pillars: the logs/traces generation and the big data analysis. This paper proposes an instrumentation framework composed by an extension of the dumping tool ndndump. This extension generates big data friendly logs that are essential to capture the behavior of NDN forwarding and tools for hardware performance measurement. Based on these logs, the second part of this work is a discussion on some properties of NDN using big data techniques that show the performance and content dissemination. We explore a potential solution where the NFD (Named Data Networking Forwarding Daemon) code might be reused to build a high-performance forwarder. The main goal is the NFD performance evaluation. Execution scenarios are realized using the emulator minindn and compared to native network scenarios.