摘要:
A defect management system is disclosed for disc storage systems which avoids the access latency associated with conventional linear replacement techniques by dispersing spare segments throughout each track at a regular interval and buffering sectors inbetween a defective sector and the corresponding spare segment during read and write operations. In one embodiment, a spare segment is an entire sector which replaces a defective data sector; and in an alternative embodiment, a spare segment stores only the defective portion of a data sector which is more efficient, but also more complicated in implementation. In both embodiments, the defect management system comprises a defect locator for locating a defective segment within a data sector. Once located, the defect management system maps the defective sector (or the defective portion thereof) to the nearest available spare segment. Then when accessing the track that comprises the defective sector, the data sectors between the defective segment and corresponding spare segment are buffered in a data buffer, and an area in the data buffer is reserved for storing the data associated with the spare segment. In this manner, data can be written to and read from the track in a contiguous sequence without requiring an extra revolution of latency as in the conventional linear replacement defect mapping techniques.