This paper addresses the problem of job scheduling in volunteer computing (VC) systems where each computation job is replicated and allocated to multiple participants (workers) to remove incorrect results by a voting mechanism. In the job scheduling of VC, the number of workers to complete a job is an important factor for the system performance; however, it cannot be fixed because some of the workers may secede in real VC. This is the problem that existing methods have not considered in the job scheduling. We propose a dynamic job scheduling method which considers the expected probability of completion (EPC) for each job based on the probability of worker's secession. The key idea of the proposed method is to allocate jobs so that EPC is always greater than a specified value (SPC). By setting SPC as a reasonable value, the proposed method enables to complete jobs without excess allocation, which leads to the higher performance of VC systems. We assume in this paper that worker's secession probability follows Weibull-distribution which is known to reflect more practical situation. We derive parameters for the distribution using actual trace data and compare the performance of the proposed and the previous method under the Weibull-distribution model, as well as the previous constant probability model. Simulation results show that the performance of the proposed method is up to 5 times higher than that of the existing method especially when the time for completing jobs is restricted, while keeping the error rate lower than a required value.

Global computing system (GCS) harnesses the idle CPU resources of clients connected to Internet for solving large problems that require high volume of computing power. Since GCS scale to millions of clients, many projects usually adopt coarse-grained scheduling in order to reduce server-side contention at the expense of sacrificing the degree of parallelism and wasting CPU resources. In this paper, we propose a new type of client, i.e., a scheduling proxy that enables adaptive-grained scheduling between the server and clients. While the server allocates coarse-grained work units to scheduling proxies alone, clients download fine-grained work units from a relatively nearby scheduling proxy not from the distant server. By computation of small work units at client side, the turnaround time of work unit can be reduced and the waste of CPU time by timeout can be minimized without increasing the performance cost of contention at the server. In addition, in order not to lose results in the failure of scheduling proxies, we suggest a technique of result caching in clients.