一种面向随机性负载的计算网格资源预留定价策略

针对传统预留机制对本地任务服务质量造成的负面影响,提出了基于预留收益与损失均衡的资源预留机制。首先,基于本地任务相关统计特性利用概率论方法,给出并证明了一段时间内本地任务执行时间之和的概率分布函数。然后,在资源提供者的效用中考虑本地任务可能带来的损失,求解出了为保障本地任务服务质量的最低资源价格。实验结果表明,提出的预留机制能有效保障本地任务的服务质量,提高资源利用率、降低任务拒绝率。

Random load oriented pricing strategy for resource reservation in computational grid

Traditional reservation mechanisms often have the negative influences on the QoS of local tasks. To solve this problem, a resource reservation mechanism based on the reservation income and loss balance was proposed. First, according to the local task correlation statistical property, using the theory of probability method, the sum of the local task execution time probability distribution function in a period was given and proved. Then, when computing resources providers' utility, the loss brought by the local task was considered and the lowest resources price was resolved. The experimental results indicate that the proposed mechanism can safeguard local tasks effectively and effectively raise resource utilization and reduce task rejection rate.