The design of efJicient storage hierorchies generally involves the repeated running of "typical" program address traces through a simulated storage system while various hierarchy design parameters are djusted. This paper describes a new and eficient method of determining, in one pass of an address trace, performance measures for a large class of demand-paged, multilevel storage systems utilizing a variety of mapping schemes and replacement algorithms. The technique depends on an algorithm classification, called "stack algorithms," examples of which are "least frequently used," "least recently used," "optimal," and "random replacement" algorithms. The techniques yield the exact access frequent-v to each storage device, which can be used to estimate the overall performance of actual storage hierarchies.

No abstract

No abstract

No abstract

Resource management in distributed multimedia systems (DMS) is a central issue, due to the real-time requirements of continous media. However, in complex and mobile systems effective resource reservation becomes almost impossible. Recent research proposes adaptive applications as a complementary solution. This article is a review and generalization of issues surrounding adaptation in DMS. As a background, we first discuss heterogeneous networks and the rationale behind adaptation. Then we explain the basic common notions and mechanisms involved, followed by implementation examples. This reveals the importance of standardized system support for adaptation. We conclude by discussing some outstanding research issues.