Resource failures risk assessment modelling in distributed environments

摘要

Purpose: To propose a mathematical model to predict the risk of failure of resources in distributed environments through formal service level agreements (SLA) to maximize the quality of service (QoS) and minimize SLA violations. Summary: Grid/cloud computing has to incorporate risk management methods to move ahead of the present best-effort approach. Risk management addresses the possibility of occurrence of events in future that may adversely affect a process. Since the grid/cloud computing supports various parties that work within contractual agreements. The use of SLAs is becoming more and more popular among service providers and end users. An SLA violation may result in a node outage or network failure. Hence a service provider would like to assess the risks in advance before implementation. This paper discusses the risk management in grid/cloud computing that can be used by the provider to identify the bottlenecks in the infrastructure and the possibility of SLA violations. A mathematical model for the prediction of risk of failure is proposed using discrete-time analytical model. Though the work is basically on grid computing, the same can be applied to cloud computing environments as well (40 refs.)