Scheduling Data Placement Jobs

In recent years, scientific applications have become increasingly data intensive. Besides, data management has perpetually remained one of the crucial problems in every stage of computer engineering, from micro (CPU chip design) level to macro (Internet and Grid infrastructure) level. For example, accessing data in a transparent and efficient manner is a major issue, both in operating system design and in microprocessor architecture. In operating systems, efficiently moving pages from disk to memory is crucial; in microprocessor architecture, instruction fetch time plays an important role; on large-scale distributed systems, transferring data files between geographically-separated storage sites, and optimizing data access in supercomputers, have major effects on overall performance. Even in the very recent multi-core era, importance of data access and data management cannot be overemphasized. I/O happens to be one of the major bottleneck for end-to-end application performance especially for Peta-scale applications. Hence, the importance of data management research in computer science, especially on large-scale systems, is deeply felt. - more

"A supercomputer is a device for turning compute-bound problems into I/O-bound problems" - Seymour Cray.