Fine-grained models of Bursty Workloads for Self-adaptive Systems Analysis

Diego Pérez

University of Zaragoza

Date: 17th April 2012 (Tuesday)
Time: 14:00 - 15:00
Venue: North N104B

Software is often embedded in dynamic contexts where it is subjected to high variable, non-stable, and usually bursty workloads.
A key requirement for a software system is to be able to self-react to workload changes by adapting its behaviour dynamically, to
ensure both the correct functionalities and the required performance. Research on fitting variable workload traces into formal models has been carried out using Markovian Modulated Poisson Processes (MMPP). These works concentrate on
modelling stable workload states, but accurate modelling of transient times still deserves attention since
they are critical moments for the self-adaptation. The work of this presentation is built on research in the area of MMPP trace fitting. It proposes a Petri net fine-grained model for highly variable workloads that also accounts for transient times. It shows a comparison of performance and availability analysis results of an adaptive system using workload models that accurately represent workload state changes and models that do not.

Employable Graduates; Exploitable Research