Scheduling continuous media in a Video-On-Demand server

Abstract
Advances in storage, compression, and network technology are making support for Video-On-Demand applications feasible. Continuous and real time data needs of this application require resource management and scheduling of storage devices. The paper discusses a model for scheduling storage devices to guarantee rate requirements for continuous media. The authors present an analysis of this class of schedulers and derive a feasibility condition and its buffer requirements. The condition is used for admission control of new requests, operations on existing requests, and to configure block sites and main memory requirements. The analysis presented here yields solutions in the continuous domain. However, due to the discrete nature of the disk scheduler a solution from this analysis cannot be implemented. The authors discuss transformations of the derived solution into one in the discrete domain while ensuring guaranteed data rate.

This publication has 3 references indexed in Scilit: