Parallel Failure Recovery Techniques in Cluster-based Media Servers

摘要

For large scale VOD service, cluster servers are spotlighted to their high performance and low cost. A cluster server consists of a front end node and multiple backend nodes. Though the increase of backend nodes provides the more QoS streams, the possibility of failures in backend nodes is proportionally increased. The failure causes not only the stop of all streaming services but also the loss of current playing positions. In this paper, when a backend node becomes a failed state, recovery mechanisms are studied to support the unceasing streaming service. Without considering the characteristic of cluster-based servers and MPEG media, the basic RAID techniques causes the performance bottleneck in the internal network and also results in the inefficiency CPU usage of backend nodes. To address these problems, a new failure recovery mechanism is proposed based on the pipeline computing concept. The proposed method distributes the network traffics needed in the process of recovery and utilizes the available CPU computing power of backend nodes. In experiments, it provides the improved performance of cluster-based VOD servers as well as the continuous streaming media service in the failed state of a backend node.