Authors: K. Kalaiselvan; P. Venkatakrishna
Addresses: School of Computing Sciences and Engineering, VIT University, Vellore Campus, Vellore – 632 014, Tamilnadu, India ' School of Computing Sciences and Engineering, VIT University, Vellore Campus, Vellore – 632 014, Tamilnadu, India
Abstract: Here we are going to see how to optimise the performance of the cloud resources using the specially designed mathematics-based quality of service (QoS) framework model. Existing cloud QoS frameworks use the parameters like cost, bandwidth and availability as a criteria for selecting the resource. These parameters are not enough when the application is compute and data intensive. In this scenario the service or resource, performing better in terms of 'latency', 'throughput' and 'response time' would be an ideal choice. Thus our QoS framework model has taken the performance metrics as a main criteria for SLA negotiation. The special weight-age is assigned to all the metrics here. Whenever the resource violates SLA, the rank of the resource would be reduced. The special weight-age is substituted in the formula to calculate the value called overall performance value. Based on the overall performance value best resource would be selected.
Keywords: infrastructure as a service; IaaS; platform as a service; PaaS; software as a service; SaaS; quality of service; QoS; SLA; service level agreements; latency; throughput; cloud computing; grid computing; schedulers; metaschedulers; latency; throughput; response time; mathematical modelling; resource selection.
International Journal of Cloud Computing, 2013 Vol.2 No.4, pp.325 - 339
Available online: 08 Dec 2013Full-text access for editors Access for subscribers Purchase this article Comment on this article