Abstract: A method for adaptive mapping for data compression includes determining an input/output (I/O) request pattern, dynamically switching between a segment mapping mode and a flat hash table mapping mode based on the determined I/O request pattern, updating a shared mapping table for the segment mapping mode and the flat hash table mapping mode, and adjusting an entry of the mapping table based on the determined I/O request pattern and a status of the entry.

Abstract: The present disclosure describes techniques of metadata management for transparent block level compression. A first area may be created in a backend solid state drive. The first area may comprise a plurality of entries. The plurality of entries may be indexed by addresses of a plurality of blocks of uncompressed data. Each of the plurality of entries comprises a first part configured to store metadata and a second part configured to store compressed data. Each of the plurality blocks of uncompressed data may be compressed individually to generate a plurality of compressed blocks. Metadata and at least a portion of compressed data associated with each of the plurality of compressed blocks may be stored in one of the plurality of entries based on an address of a corresponding block of uncompressed data. A second area may be created in the backend solid state drive for storing the rest of the compressed data.

Abstract: Systems and methods for space allocation for block device compression are provided. In particular, a computing device may receive an allocation request to write the compressed data, select a range list adequate for serving the allocation request from a plurality of range list, dequeue a range entry from the selected range list to allocate free space for the compressed data, and allocate the free space corresponding to the range entry to the compressed data to serve the allocation request.

Abstract: A system and method are described to efficiently allocate memory space with low latency overhead by allocating blocks of non-volatile memory on a storage device according to a tree data structure comprising a plurality of counter sets, each counter set including one or a plurality of counters indicating numbers of unallocated blocks of memory space within the non-volatile memory.

Abstract: Systems and methods for accessing block storage devices are provided. In particular, a computing device may receive a write request including an uncompressed data and an uncompressed block address associated with the uncompressed data, generate compressed data by compressing the uncompressed data, determine a plurality of mapping candidates of compressed data blocks in the block storage devices based on the uncompressed block address, select a compressed data block from the plurality of mapping candidates that has sufficient capacity to store the compressed data, write the compressed data to the selected compressed data block, update metadata of the selected compressed data block to link the uncompressed block address to a compressed block address of the selected compressed data block, and write the selected compressed data block back to a respective block storage device of the block storage devices.

Abstract: A flat hash table includes a plurality of entries, and each entry includes a hash function index and a usage bitmap. A method for block device level compression mapping using the flat hash table includes compressing uncompressed data to compressed data, retrieving an entry of the flat hash table using an uncompressed block address of the uncompressed data, determining a compressed block address of the compressed data by executing at least one hash function and by determining a hash function in the at least one hash function for mapping the uncompressed block address to the compressed block address that corresponds to a space in a block storage device, storing the compressed data to the space that corresponds to the compressed block address, and updating the hash function index of the entry of the flat hash table with an index indicative of the hash function.

Abstract: The present disclosure describes techniques of metadata management for transparent block level compression. A first area may be created in a backend solid state drive. The first area may comprise a plurality of entries. The plurality of entries may be indexed by addresses of a plurality of blocks of uncompressed data. Each of the plurality of entries comprises a first part configured to store metadata and a second part configured to store compressed data. Each of the plurality blocks of uncompressed data may be compressed individually to generate a plurality of compressed blocks. Metadata and at least a portion of compressed data associated with each of the plurality of compressed blocks may be stored in one of the plurality of entries based on an address of a corresponding block of uncompressed data. A second area may be created in the backend solid state drive for storing the rest of the compressed data.

Abstract: Systems and methods for space allocation for block device compression are provided. In particular, a computing device may receive an allocation request to write the compressed data, select a range list adequate for serving the allocation request from a plurality of range list, dequeue a range entry from the selected range list to allocate free space for the compressed data, and allocate the free space corresponding to the range entry to the compressed data to serve the allocation request.

Abstract: A method for adaptive mapping for data compression includes determining an input/output (I/O) request pattern, dynamically switching between a segment mapping mode and a flat hash table mapping mode based on the determined I/O request pattern, updating a shared mapping table for the segment mapping mode and the flat hash table mapping mode, and adjusting an entry of the mapping table based on the determined I/O request pattern and a status of the entry.

Abstract: A system and method are described to efficiently allocate memory space with low latency overhead by allocating blocks of non-volatile memory on a storage device according to a tree data structure comprising a plurality of counter sets, each counter set including one or a plurality of counters indicating numbers of unallocated blocks of memory space within the non-volatile memory.

Abstract: A hybrid approach for performing edit distance searching used for fuzzy string searches. A system is disclosed that includes an FPGA (field programmable gate array) appliance, having: a data input manager that receives an m-byte input pattern and loads an n-byte substring of the m-byte input pattern into a first set of registers, and streams input strings of searchable data through a second set of registers; an edit distance calculation engine having an array of processing elements (PEs) implemented using FPGAs coupled to the first and second set of registers, wherein the array of PEs calculate an edit distance for each input string of searchable data relative to n-byte substring; and an output manager that identifies matching input strings having an edit distance less than a threshold, and forwards matching input strings to a CPU for software-based edit distance processing relative to the m-byte input pattern.