A Mass Storage System Administrator Autonomic Assistant

摘要

System administrators of today's high performance computing systems are generally responsible for managing the large amounts of data traffic and archival querying that mass storage systems must provide to users who compute on hundreds or thousands of processors at once. The file management systems that have been engineered to handle this workload generally consist of a reliable compute server, high- and low-speed disks, robotic tape silos with thousands of cartridges, and various network interconnects. As a result of the wide variety of mechanical components used, storage system administrators maintain the functional aspects of system operations and troubleshoot the day-to-day lower level physical system and software failures under severe constraints. The Mass Storage System Administrator Autonomic Assistant (MSSAAA) is being developed to reduce the burden on storage system administrators and improve the reliability, availability and serviceability of mass storage systems