Until now little research has addressed the problem of on line scheduling of sporadic parallel tasks with hard deadlines in partitionable multiprocessor sys tems In this paper we present two on-line scheduling al gorithms -Buddy/RTand Stacking, for such environments Both the algorithms either guarantee or reject a task at the time of its arrival Buddy/RT is a straight-forward extension of the well-known Buddy strategy to the real time environment, while Stacking is a more sophisticated algorithm based on the lessons learned from Buddy/RT The underlying concept behind the Stacking algorithm is to reduce fragmentation by 'stacking' equal-sized jobs in the time dimension The Stacking algorithm is found to perform significantly better than Buddy/RT over a wide range of workloads, even though both the algorithms have the same time complexity

Due to its simplicity, regularity and suitability for VLSI implementation, the mesh topology for multiprocessors has drawn considerable attention. Several processor allocation strategies for mesh-connected multiprocessors have been proposed in recent years. In this paper; we present the results of a perform" study of all the proposed strategies known to authors. Originally each of these allocation strategies was proposed for use with First-Come-FirstServed job scheduling. In this paper we also propose and evaluate new variants of these strategies using the Scan scheduling discipline. We find Scan to signijicantly improve the performance of all the allocation strategies. A wide range of workloads and system sizes are considered. In addition, we compare the overheads of these algorithms and study the effects of overheads on pe~ormance.

No abstract