Modeling Virtualized Network Infrastructures for Capacity Management
Performance of the cloud system is determined by performance of three subsystems: computing, storage and networking. In this topic we model the networking infrastructure of a single cloud datacenter in order to increase predictability of the services running in the datacenter and to provide the mechanisms needed to make the cloud system self-aware. The self-awareness mechanisms allow better utilization of the cloud resources and provide the flexibility to satisfy the needs of varied range of customers.
The goal of this project will be obtained by performing the following steps: 1) reviewing networking virtualization techniques used in cloud; 2) developing descriptive model of the networking infrastructure in the data center; 3) developing a transformation from the descriptive model into a predictive model to predict the quality of service (QoS) in the network; 4) use the models and the prediction results to proactively allocate the resources of the network in order to meet the QoS requirements and SLAs. The results of the research will be: models, transformations and network-management mechanisms allowing to precisely predict and control the quality of service levels in the cloud data center environment.
Main research interests:
- Quality of Service provisioning in networks and distributed service-based systems
- Resource Management
- QoS-aware composite service composition
- System Modeling and Simulation
- Discrete optimization methods
- Computational complexity
Rygielski P., Kounev S.: Data Center Network Throughput Analysis using Queueing Petri Nets. In 34th IEEE International Conference on Distributed Computing Systems Workshops (ICDCS 2014 Workshops). 4th International Workshop on Data Center Performance, (DCPerf 2014), Madrid, Spain, June 30, 2014, pages 100-105, PDF
Rygielski P., Kounev S., Zschaler S.: Model-Based Throughput rediction in Data Center Networks. In Proceedings of the 2nd IEEE International Workshop on Measurements and Networking (M&N 2013), October 7-8, Naples, Italy, 2013, pp. 167-172, PDF
Rygielski P., Zschaler S., Kounev S.: A Meta-Model for Performance Modeling of Dynamic Virtualized Network Infrastructures (Work-In-Progress Paper). In Proceedings of the 4th ACM/SPEC International Conference on Performance Engineering (ICPE 2013), April 21-24, Prague, Czech Republic, ACM, New York, NY, USA, April 2013, pp. 327-330, PDF
Rygielski, P., Kounev, S.: Network Virtualization for QoS-Aware Resource Management in Cloud Data Centers: A Survey. In PIK – Praxis der Informationsverarbeitung und Kommunikation, 36(1):55-64, February 2013, de Gruyter, PDF
Milenkoski A., Iosup A., Kounev S., Sachs K., Rygielski P., Ding J., Cirne W., Rosenberg F.: Cloud Usage Patterns: A Formalism for Description of Cloud Usage Scenarios. Technical Report SPEC-RG-2013-001 v.1.0.1, SPEC Research Group – Cloud Working Group, Standard Performance Evaluation Corporation (SPEC), 7001 Heritage Village Plaza Suite 225, Gainesville, VA 20155, April 2013, HTML