Failure recovery algorithms for multimedia servers

In this paper, we present two novel disk failure recovery methods that util ize the inherent characteristics of video streams for efficient recovery. W hereas the first method exploits the inherent redundancy in video streams ( rather than error-correcting codes) to approximately reconstruct data store d on failed disks, the second method exploits the sequentiality of video pl ayback to reduce the overhead of online failure recovery in conventional RA ID arrays. For the former approach, we present loss-resilient versions of J PEG and MPEG compression algorithms. We present an inherently redundant arr ay of disks (IRAD) architecture that combines these loss-resilient compress ion algorithms with techniques for efficient placement of video streams on disk arrays to ensure that on-the-fly recovery does not impose any addition al load on the array. Together, they enhance the scalability of multimedia servers by (1) integrating the recovery process with the decompression of v ideo streams, and thereby distributing the reconstruction process across th e clients; and (2) supporting graceful degradation in the quality of recove red images with increase in the number of disk failures. We present analyti cal and experimental results to show that both schemes significantly reduce the failure recovery overhead in a multimedia server.