Computing Services Grid Resource Management Techniques

Through investigating on the key techniques in grid resource management, resource description, discovery, scheduling and QoS under computational grid environment are discussed in this paper. A computing expectation based grid resource management model is presented based on the theory of grid economy model. Researches in detail of the computing expectation based grid resource management model are:1. A mechanism of autonomous resource price fluctuation is put forward. Resource load balancing is implemented by the balancing of resource supply and demand which is reflected by the resource price fluctuation.2. Computing expectation, which reflects the computation desire and demand of grid user, is designed. The grid user's detailed demand for the attribute of resource is abstracted as a computing expectation of job, which is explained by the grid system and mapped to the corresponding attribute of grid resource. The operation of user is simplified and user's computing demand is described flexibly by the computing expectation.3. The performance between query efficiency and system load is balanced by database based hierarchical resource discovery mechanism, which implements grid resource description and discovery.4. Different resource scheduling algorisms are designed for different computing expectation in order to achieve flexible and efficiency grid resource schedule.The investigations for grid resource management in this paper achieve the load balance of grid resource, simplify the operation of grid user and implement the flexible and efficient management and schedule of grid resource by characteristic description for user's computation demand.