Searchable Data Structures for Storage System Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional storage systems face performance degradation and congestion when handling multiple client requests for information about storage devices managed by multiple servers, due to the use of linked lists which do not allow efficient indexing and require scanning through all elements to access data, leading to delays and inefficiencies.
Innovation Solution
Implementing a system with searchable data structures and a lock data structure at a management console to store information about storage devices, allowing for efficient retrieval of data using search keys and preventing unauthorized access by managing locks at the data structure level, thereby improving performance and handling multiple requests concurrently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If linked lists are used to store storage device information, then the system structure is simple, but data retrieval efficiency deteriorates due to the need to scan through all elements
Solution Approach 1:
The patent segments the flat linked list structure into hierarchical indexed structures with multiple levels. Data is organized into groups and subgroups with indexes at each level, allowing the system to jump directly to relevant data sections rather than scanning sequentially through all elements.
Solution Approach 2:
The patent introduces dimensional indexing by organizing data in multi-level hierarchical structures with indexes at different levels. This transforms the one-dimensional sequential access of linked lists into multi-dimensional direct access through index keys, enabling efficient retrieval without linear scanning.
2Quantity of substance
If linked lists are used for storing storage device information, then memory usage is low, but access time increases due to sequential scanning requirements
Solution Approach 1:
The patent performs preliminary organization of data into indexed hierarchical structures during data insertion or system initialization. Indexes are pre-computed and stored with the data, so when retrieval is needed, the system can directly access the relevant data through the pre-built indexes without performing sequential scanning at access time.
3Adaptability or versatility
If multiple client requests are handled using conventional storage systems, then system coverage is comprehensive, but performance degradation occurs due to congestion
Solution Approach 1:
The patent segments the monolithic data access path into multiple independent indexed structures. Each index can be searched and accessed independently, allowing multiple client requests to be processed simultaneously through different index paths without causing congestion in a single sequential access bottleneck.
4Productivity
If searchable data structures with locks are implemented, then data access efficiency improves, but system complexity increases due to lock management
Solution Approach 1:
The patent implements locking at the granular level of individual index entries or data nodes rather than locking the entire data structure. This allows fine-grained concurrency control where locks are acquired only for the specific data elements being accessed, minimizing lock contention and reducing the impact of lock management complexity.
Data Source
AI summary
Method and system for providing information regarding a plurality of storage devices managed by a plurality of storage servers are provided. The storage space at the storage devices is presented to a plurality of computing systems as logical storage space. A plurality of searchable data structures having a plurality of data object types are stored at a temporary memory storage device of a management console that interfaces with the plurality of computing systems and the storage servers. Each data object type stores information regarding the storage device. The searchable data structure includes information regarding the storage devices and the logical storage space presented to the computing systems. A lock data structure for tracking locks that are assigned for accessing information pertaining to a storage server and a data object type is maintained to prevent unauthorized access to at least one of the searchable data structures.