An SSD-Based Accelerator Using Partitioned Bloom Filter for Directory Parsing

With the advancement in IT including mass storage device and super-fast communication, cloud computing is now in high demand. At the same time, the explosive increase in the volume of diverse and extensive data has brought forth the age of big data organized as countless directories within a file system. The exponential increase in data volumes causes slow retrieval of metadata of the requested files. This paper proposes a novel directory parsing technique using an SSD-based Bloom filter. The proposed scheme employs partitioned Bloom filter for accurate and fast directory parsing. Computer simulation reveals that the proposed scheme significantly reduces the rate of false positive and shortens the time spent on directory parsing in comparison to the existing Bloom filter-based scheme.