Fault Surviving Optimisation within Brick Based Storage System

The cost of providing an organization with enterprise class storage is continuing to rise, due primarily to the use of specialised components and the complex internal storage system structures. This high cost has resulted in researchers directing their efforts towards modular or 'brick' based storage, which utilise commodity components while employing autonomic principles for the management of the system, reducing the costs incurred during the lifetime of the storage system. While systems based upon this brick structure incorporate high levels of data reliability/availability, they fail to recognise the importance of ensuring the management services of a system are also tolerant of failures (such as optimisation). If system management services are not provided reliably, such storage systems which experience failures will degradedramatically in performance, especially when the node/s responsible for providing the management fails. This paper addresses the issue of providing fault tolerant management services within brick storage systems. We present in this paper our model of a system that can ensure the integrity and availability of data stored within brick based storage while ensuring the continual execution of the systems management services irrespective of failures within the system.