High read block clustering at deduplication layer
Abstract:
Methods, systems, and computer program products are provided for deduplicating data. In one embodiment, a method comprises mapping a plurality of file blocks of selected data to a plurality of logical blocks, deduplicating the plurality of logical blocks to thereby associate each logical block with a corresponding physical block of a plurality of physical blocks located on a physical memory device, two or more of the corresponding physical blocks being non-contiguous with each other, and determining whether one or more of the corresponding physical blocks are one or more frequently accessed physical blocks being accessed at a frequency above a threshold frequency and being referred to by a common set of applications.
Public/Granted literature
Information query
Patent Agency Ranking
0/0