Humboldt-Universität zu Berlin - Mathematisch-Naturwissenschaftliche Fakultät - Institut für Informatik

Humboldt-Universität zu Berlin | Mathematisch-Naturwissenschaftliche Fakultät | Institut für Informatik | Institutstermine | Vortrag: Automated Parameterization of Performance Models from Measurements

Vortrag: Automated Parameterization of Performance Models from Measurements

  • Wann 14.12.2015 von 15:15 bis 16:30
  • Wo Rudower Chaussee 25, Haus IV, Raum 410
  • iCal

Dr. Giuliano Casale, Imperial College London, UK, wird einen Vortrag über die Erstellung von Performanzmodellen für Softwaresysteme halten.

Alle Interessierten sind herzlich eingeladen


 

Abstract:

In this talk I will present the problem of estimating parameters of performance models from measurements of real systems and discuss algorithms that can support researchers and practitioners in this task. The focus will be on performance models based on queueing systems, where the estimation of resource consumption of requests is a required input to the model. I will present in particular novel estimation methods to obtain mean resource consumption values across types of requests, which leverage regression, maximum-likelihood and Markov-chain Monte Carlo methods. I will then discuss their applicability to cloud and enterprise software system modelling.

 

Short bio:

Giuliano Casale is a Senior Lecturer in modeling and simulation at the Department of Computing of Imperial College London. Prior to this, he was a full-time researcher at SAP Research UK. His research interests include cloud computing, modeling, and resource management, topics on which he has published more than 80 refereed papers in international conferences and journals. He has served as co-chair for various conferences in the area of performance engineering, such as ACM SIGMETRICS/Performance, MASCOTS, ICAC and ICPE. He is member of the IFIP WG 7.3 group on Computer Performance Analysis and the ACM SIGMETRICS Board of Directors.