Cuckoo Hash Directory Indexing for Concurrent File Access
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing file systems face significant delays in accessing files from large directories due to the lack of concurrent processing support in their tree-based indexing structures, leading to inefficient file access paths.
Innovation Solution
Implementing a directory management system based on Cuckoo hash, which uses a hash table to create indexes for sub-directories or sub-files, allowing for concurrent processing and reducing the number of reading and writing operations through optimized hash functions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a tree structure (HTree, B+tree) is used to create indexes for sub-directories or sub-files in each directory, then the index structure is well-organized and searchable, but a large access delay is introduced when files are accessed in large directories due to sequential reading operations from root node to leaf nodes
Solution Approach 1:
The patent segments the sequential tree traversal into parallel hash function evaluations. Instead of traversing from root to leaf in a tree structure, the invention uses multiple independent hash functions (h1, h2, h3, h4) that can be evaluated simultaneously to determine file positions, eliminating the sequential dependency inherent in tree structures.
Solution Approach 2:
The patent replaces the mechanical tree traversal mechanism with a hash-based direct addressing mechanism. The tree structure's sequential reading and path-finding operations are substituted with parallel hash function evaluations that directly compute file positions without sequential dependencies.
2Ease of manufacture
If a tree structure is used for indexing, then the index can be created and maintained, but concurrent operations are not available during the search process from root node to leaf nodes
Solution Approach 1:
The patent divides the index operation into independent segments (multiple hash functions) that can execute in parallel. Each hash function h1, h2, h3, h4 operates independently on the filename to produce hash values, allowing concurrent evaluation without the sequential locking required by tree structures.
Solution Approach 2:
The patent transitions from the one-dimensional sequential path of tree traversal to a multi-dimensional parallel evaluation space where multiple hash functions operate simultaneously. This dimensional change enables concurrent operations by distributing the indexing work across multiple independent computational paths.
3Adaptability or versatility
If the depth of B+tree increases to accommodate more sub-files or sub-directories, then the extendibility of the file system is improved, but the search delay increases correspondingly
Solution Approach 1:
The patent changes the fundamental parameter of index structure from tree depth to hash function count. Instead of increasing tree depth which linearly increases search time, the invention increases the number of hash functions in parallel, maintaining constant-time complexity O(1) regardless of the number of files, thus achieving extendibility without increased delay.
Data Source
AI summary
A directory management method and system for a file system based on Cuckoo hash are provided, including the steps of reading metadata of a sub-directory or a sub-file, used as a target file, in a directory: receiving a request for reading the target file in the directory; for the target file, determining an ith candidate data block, in a hash table of the directory according to a hash calculation result obtained by performing hash calculation on a filename filename of the target file according to an ith hash function of Cuckoo hash iteratively; if the filename filename of the target file exists in the ith candidate data block, reading metadata of the target file and returning the metadata, and ending the process; otherwise, continuing to iterate until the iteration ends and returning a message that the target file does not exist.