Load management in distributed video servers

Abstract

In this paper, we define and formulate various policies for load management in distributed video servers. We propose a predictive placement policy that determines the degree of replication necessary for popular videos using a cost-based optimization procedure based on a priori predictions of expected subscriber requests. For scheduling requests, we propose an adaptive scheduling policy that compares the relative utilization of resources in a video server to determine an assignment of requests to replicas. To optimize storage utilization, we also devise methods for dereplication of videos based on changes in their popularities and in server usage patterns. Performance evaluations indicate that a load management procedure which uses a judicious combination of the diflerent policies performs best for most server configurations.Advances in storage technologies are making highperformance video servers a reality. These video servers are being deployed over emerging broadband networks to deliver a variety of interactive, digital video services to thousands of residential subscribers. To meet the scalability requirements in such large deployments, distributed video server architectures are being considered [3]. In this paper, we propose various methods for load management that are targeted at improving the cost-effectiveness of distributed video servers. The paper is organized as follows: Section 1 motivates the need for load management and discusses its implications on a video server's cost-performance.Section 2 presents the different policies for load management while Section 3 discusses more optimizations for load management. Section 4 presents a comparative analysis of the different policies. We conclude in Section 5 and outline areas for future research.

Directory-based composite routing and scheduling policies for dynamic multimedia environments

“…video/audio streaming, conferencing. Approaches to load-based task assignment in distributed server environments have been developed [VR97,HCM98].…”

Optimal video replication and placement on a cluster of video-on-demand servers

“…The video replication and placement problem has been studied extensively in the literature [3,4,5,15,16]. Chervenak et.…”

“…The relative popularity of videos follows Zipf-like distributions with a skew parameter of . Typically, ¼ ¾ ½ ½ [1,15]. The probability of choosing the Ø video is Ô È Å ½ .…”

“…For cost-effectiveness and scalability, the research interest in cluster-based streaming servers (briefly, VoD cluster) has increased dramatically in the past few years [2,9,10,12,15]. A VoD cluster consists of a number of back-end servers controlled by a dispatcher that makes admission decisions.…”

Mapping media streams onto a network of servers

“…The problem studied in this paper has also been described in [10] and [11]. In these two papers presented by Venkatasubramanian and Ramanathan no solutions for the mapping problem were discussed, but some general principles for duplicating media assets on a network of servers were presented.…”