Fast Reconfiguration of Data Placement in Parallel Disks

The “How much information?” study produced by the school of information management and systems at the University of California at Berkeley [10], estimates that about 5 exabytes of new information was produced in 2002. It estimates that the amount of stored information doubled in the period between 1999 and 2002. It is believed that more data will be created in the next five years than in the history of the world. Clearly we live in an era of data explosion. This data explosion necessitates the use of large storage systems. Storage Area Networks (or SANs) are the leading [13] infrastructure for enterprise storage. A SAN essentially allows multiple processors to access several storage devices. They typically access the storage medium as though it were one large shared repository. One crucial function of such a storage system is that of deciding the placement of data within the system. This data placement is dependent on the demand pattern for the data. For instance, if a particular data item is very popular the storage system might want to host it on a disk with high bandwidth or make multiple copies of the item. The storage system needs to be capable of handling flash crowds [8]. During events triggered by such flash crowds, the demand