Abstract: Systems need to run a larger and more diverse set of applications, from real time to interactive to batch, on uniprocessor & multiprocessor platforms. The problem of inferring application resource requirements is difficult because the relationship between application performance and resource requirements is complex and workload dependent. This study investigates a measurement-based approach to resource inference � employing online measurements of workload characteristics and system resource usage to estimate application resource requirements. A scheduling algorithm which provides low latency for real-time and interactive application is presented. The schedulability of each process is enforced by a guaranteed cpu service rate, independent of the demands of other processes. The resulting scheduler is implemented in the Linux kernel and evaluate its performance using various application and benchmarks.