Comparison of priority-based and first price sealed-bid auction algorithms of job scheduling in a geographically-distributed computing system

The paper discusses the problem of scheduling computational jobs in a geographically-distributed computing system with decentralized resource management model. Computational jobs in this model have absolute priorities, and high priority job can interrupt low priority running jobs. Straightforward approach for this problem in this kind of systems is based on the simple priority-based algorithm. Another approach is based on economic scheduling methods; authors consider the first-price sealed-bid auction. The paper presents results of experimental comparison of the two algorithms.