A Study On Resource Scheduling Scheme In Resource Kernel

Because the time constrains of many system become tighter, the use of the real-time system get wilder. The duty of real-time system is to provide time constraint guarantees for the real-time applications, in other words, it make the applications access to the resources they need timely. Many works have devoted into individual resource scheduling (i.e. processor capability, disk bandwidth and network bandwidth ). Obtaining simultaneous and timely access to multiple resource is known to be a NP-complete problem. In this paper, we investigate the problem of co-scheduling different resources. First, we analysis the Cooperative Scheduling Server (CSS), an dedicated server that manage controlled resource (like disk bandwidth). The CSS is implement with the Resource kernel model. Then present a dynamic slack sharing algorithm to replace the fixed one used in CSS. The new slack sharing algorithm is based the current resource workload. A simple analysis is make to determine the performance of the dynamic slack sharing algorithm. Last, a Resource Kernel related work, share memory allocate manager, is introduced. This manager can provide efficient data access.