Effects of Dynamic Cloud Cluster Load on Differentiated Service Availability

Accredited to the diverse nature of cloud services, there is a need for cloud providers to offer differentiated availability for different service types in their SLAs. This can be achieved by using different redundancy strategies and fault-tolerance techniques. The availability resulting from these techniques is highly dependent upon the load on cloud datacenters and their clusters. This load is dynamic, caused by both variations in demand and failures in servers, network, and other cloud infrastructure. In this paper, the effect of dynamic load in a cloud cluster on the service availability is studied, using analytical models and simulations. The results are thus obtained for different loads and compared among different service classes. The analytical models are not able to grasp the interaction between different classes, and hence a simulation is performed. The results show that the cluster load has a quantifiable effect on service availability, and it increases with decreasing level of priority assigned to a service class.

[1]  Randy H. Katz,et al.  Above the Clouds: A Berkeley View of Cloud Computing , 2009 .

[2]  Rajkumar Buyya,et al.  Article in Press Future Generation Computer Systems ( ) – Future Generation Computer Systems Cloud Computing and Emerging It Platforms: Vision, Hype, and Reality for Delivering Computing as the 5th Utility , 2022 .

[3]  Dutch T. Meyer,et al.  Remus: High Availability via Asynchronous Virtual Machine Replication. (Best Paper) , 2008, NSDI.

[4]  Albert G. Greenberg,et al.  VL2: a scalable and flexible data center network , 2009, SIGCOMM '09.

[5]  Jin B. Hong,et al.  Availability Modeling and Analysis of a Virtualized System , 2009, 2009 15th IEEE Pacific Rim International Symposium on Dependable Computing.

[6]  Luiz André Barroso,et al.  The Datacenter as a Computer: An Introduction to the Design of Warehouse-Scale Machines , 2009, The Datacenter as a Computer: An Introduction to the Design of Warehouse-Scale Machines.

[7]  P. Mell,et al.  The NIST Definition of Cloud Computing , 2011 .

[8]  Ameen Chilwan,et al.  Dependability Differentiation in Cloud Services , 2011 .

[9]  Graham M. Birtwistle,et al.  DEMOS A System for Discrete Event Modelling on Simula , 1979, Springer New York.

[10]  Dong Seong Kim,et al.  Candy: Component-based Availability Modeling Framework for Cloud Service Management Using SysML , 2011, 2011 IEEE 30th International Symposium on Reliable Distributed Systems.

[11]  Bruce Christensen.  Dependability and Performance , 2003 .

[12]  Poul E. Heegaard,et al.  Differentiated Availability in Cloud Computing SLAs , 2011, 2011 IEEE/ACM 12th International Conference on Grid Computing.

[13]  Alexandru Iosup,et al.  The Failure Trace Archive: Enabling Comparative Analysis of Failures in Diverse Distributed Systems , 2010, 2010 10th IEEE/ACM International Conference on Cluster, Cloud and Grid Computing.

[14]  W. H I T E P A P,et al.  Protecting Mission-Critical Workloads with VMware Fault Tolerance , 2009 .