Skyscraper broadcasting: a new broadcasting scheme for metropolitan video-on-demand systems
SIGCOMM '97 Proceedings of the ACM SIGCOMM '97 conference on Applications, technologies, architectures, and protocols for computer communication
Design and implementation of scalable continuous media servers
Parallel Computing - Special issues on applications: parallel data servers and applications
On the network impact of dynamic server selection
Computer Networks: The International Journal of Computer and Telecommunications Networking
Segment-based proxy caching of multimedia streams
Proceedings of the 10th international conference on World Wide Web
Enabling dynamic content caching for database-driven web sites
SIGMOD '01 Proceedings of the 2001 ACM SIGMOD international conference on Management of data
A Redundant Hierachical Structure for a Distributed Continuous Media Server
IDMS '97 Proceedings of the 4th International Workshop on Interactive Distributed Multimedia Systems and Telecommunication Services
Geographic Load Balancing for Scalable Distributed Web Systems
MASCOTS '00 Proceedings of the 8th International Symposium on Modeling, Analysis and Simulation of Computer and Telecommunication Systems
Coordinated Placement and Replacement for Large-Scale Distributed Caches
WIAPP '99 Proceedings of the 1999 IEEE Workshop on Internet Applications
Dynamic Load Balancing in Geographically Distributed Heterogeneous Web Servers
ICDCS '98 Proceedings of the The 18th International Conference on Distributed Computing Systems
Clustering and indexing of experience sequences for popularity-driven recommendations
Proceedings of the 3rd ACM workshop on Continuous archival and retrival of personal experences
Hi-index | 0.00 |
Because of their size, service times, and drain on server resources, multimedia objects require specialized replication systems in order to meet demand and ensure content availability. We present a novel method for creating replication systems where the replicated objects' sizes and/or per-object service times are large. Such replication systems are well-suited to delivering multimedia objects on the Internet. Assuming that user request patterns to the system are known, we show how to create replication systems that distribute read load to servers in proportion to their contribution to system capacity and experimentally show the positive load distribution properties of such systems. However, when user request patterns differ from what the system was designed for, system performance will be affected. Therefore, we also report on results that reveal (i) how server loads are affected and (ii) the impact two system design parameters (indicators of a system's load distribution qualities) have on server load when request patterns differ from that for which a system was designed.