Directory Information Recovery via Metadata Location Servers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional storage systems face challenges in efficiently locating data objects within hierarchical directory structures when intermediate nodes become unavailable, due to issues like hardware or software errors, leading to difficulties in accessing files despite their availability.
Innovation Solution
A system that determines characteristics of a file object, such as owner, creation date, and business organization, to obtain a subset of nodes using metadata location servers, allowing for the identification of available nodes containing the file object even if intermediate nodes in the directory structure are unavailable.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If conventional hierarchal directory structures are used to store data objects, then data organization and access are simplified, but system reliability deteriorates when intermediate nodes become unavailable
Solution Approach 1:
The system performs preliminary actions by creating alternative path records and backup directory entries during file creation and normal operations. When intermediate nodes fail, these pre-established alternative paths enable direct access to files without traversing through failed nodes, thus maintaining reliability while preserving ease of operation.
Solution Approach 2:
The patent introduces intermediary components such as backup directory servers and alternative path records that mediate between clients and failed intermediate nodes. These intermediaries store metadata and path information that allow clients to bypass failed nodes and directly locate files, resolving the contradiction between simple hierarchical access and reliability.
2Reliability
If the entire directory structure is searched to locate a file object, then complete coverage is ensured, but search time and system performance deteriorate
Solution Approach 1:
The patent segments the directory search space by organizing files into multiple directories and maintaining alternative path records for each file. Instead of searching the entire directory structure, the system divides the search into smaller segments by checking alternative paths first, then only searching specific subdirectories if needed, thus reducing search time while ensuring complete coverage.
Solution Approach 2:
The system applies partial action by searching only the necessary portion of the directory structure. By maintaining alternative path records that point directly to file locations, the system avoids excessive searching of unrelated directories. The search is limited to checking alternative paths and relevant subdirectories rather than traversing the entire directory tree.
3Productivity
If specialized hardware platforms are used to meet performance requirements, then I/O performance and reliability are improved, but system cost and complexity increase
Solution Approach 1:
The patent implements multi-functionality by enabling standard storage servers to perform both data storage and directory service functions. The system allows any server in the network to act as a directory server or data server depending on needs, eliminating the requirement for specialized hardware. This universal approach maintains high I/O performance while reducing hardware complexity and cost.
Solution Approach 2:
The system uses copying by maintaining multiple copies of directory information and path records across different servers. Instead of relying on specialized hardware for directory services, the patent distributes directory data copies across standard servers, allowing any server to provide directory services and improving performance through parallel access without increasing hardware complexity.
Data Source
AI summary
Locating a file object in a storage system includes determining at least one characteristic of the file object, obtaining a subset of nodes in the storage system by applying the at least one characteristic to criteria for storing file objects at particular ones of the nodes in the storage system, and examining the subset of nodes to determine if any contain the file object. A metadata location server may be used to obtain the subset of nodes. The criteria may include at least one policy relating to placement of objects corresponding to files at the time the files are created. The policies may relate to owners of the files. The characteristic information may include the owner of a target file corresponding to the file object, the date that the target file was created, a business organization associated with the target file, and/or the size of the target file.


