A Novel SLA based Task Scheduling in Grid Environment

Service Level Agreements (SLAs) are used in different domains and on different levels to establish agreements on Quality of Service (QoS) between a service provider and a customer. SLAs can be based on general agreements like framework agreements. In a grid environment SLA can be technical agreements which deal with properties of services and resources. Research and development in grid computing started addressing electronic Service Level Agreements where negotiation is usually done automatically by appropriate instances of the service provider and service consumer. Today’s internet facility is used to specify SLAs with new improvements which satisfy customer expectations. The main idea is to provide different levels of service by forging agreements between the user and resource owner. Such agreements are agreed on the basis of different constraints expressed by the user and/or the resource owner. The use of Service Level Agreement (SLA) gives rise to a fundamentally new approach for job scheduling on the grid. This paper provides an overview of the issues surrounding the design of a fundamentally new infrastructure for job scheduling on the grid, which is based on the notion of SLA’s. Such SLA’s are negotiated between a client and a provider based on job categorization and resource clustering.