Application Performance Monitoring in Software Defined Networks

Abstract

Application Performance Monitoring (APM) is crucial to any enterprise to ensure that the performance of its application or service is meeting the service level agreements. There is an inherent need for the APM of traditional networks to evolve, with the advances in the internet technologies. In this paper, we present a novel approach and framework for realizing a futuristic APM for Software Defined Networks (SDN). The approach is based on analyzing the Gartner's dimensions of APM, in the SDN scenario. The five distinct dimensions of APM as per Gartner's matrix are i) End User Experience monitoring. ii) Runtime application architecture discovery, modeling, and display. iii) User defined transaction profiling. iv) Component deep-dive monitoring in an application context. v) Reporting & application data analytics. The contributions of the paper are a) Mapping of APM dimensions to SDN measurements. b) An emulation framework development involving the identification of APIs for the measurements and capturing the measurements. c) Results depicting the feasibility of this approach.