Further Developments of a Dynamic Distributed Video Proxy-Cache System

We investigate the dynamics of a distributed video proxy-cache system that is able to adapt the number of running nodes depending on conditions like client request patterns, network load etc. Since we've already examined the split operation used to expand the system by adding new nodes, we explore in detail two operations (hibernate and shut down) used to reduce the number of active nodes. Also, several scenarios for object (video data) movement and replication between participants are taken into consideration. Further more, we study the system's efficiency by measuring the byte hit rate under different test scenarios