Tolerating Multiple Failures iti RAID Architectures with Optimal Storage and Uniform Declustering

We present DATUM, a novel method for tolerating multi-ple disk failures in disk arrays. DATUM is the first known method that can mask any given number of failures, re-quires an optimal amount of redundant storage space, and spreads reconstruction accesses uniformly over disks in the presence of failures without needing large layout tables in controller memory. Our approach is based on information dispersal, a coding technique that admits an efficient hardware implementation. As the method does not restrict the configuration parameters of the disk array, many existing RAID organizations are par-ticular cases of DATUM. A detailed performance com-parison with two other approaches hows that DATUM’S response times are similar to those of the best competi-tor when two or less disks fail, and that the performance degrades gracefully when more than two disks fail.