Scheduling Remote Access to Scientific Equipments in Cyber-infrastructure for Education and Research
- J. Yin, J Cao, YX Wang
- International Symposium on Cluster Computing and…
Equipment grid aims to facilitate easy access to expensive scientific instruments by grid services and consists of the following three components: service pool alliance, service pool, and geographically distributed physical instruments. When users submit experiments to equipment grid, service pool alliance will allocate instruments in related service pools to conduct the experiments, which may need coordination and cooperation of several physical instruments that constitute service chains. After experiments have finished and results have returned, users evaluate the performance of related service chains. In this work, a scheduling algorithm using provenance information is proposed to enhance performance of equipment grid by increasing dispatch probability of instruments with high QoS. In this algorithm, we express QoS of instruments and user appraisals in fuzzy linguistic values, taking the vagueness of user opinions on experiment results and various criteria to evaluate instrument QoS into account. Simulation results show that with this algorithm, equipment grid can better satisfy the users.