Partitioned Data Search Across Storage Modules to Reduce Conflicts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Network devices face conflicts and increased difficulty in parallel data searches due to the likelihood of entries being located in the same storage module during parallel processing, especially with increasing network traffic.
Innovation Solution
Implement a data search apparatus with multiple storage modules, where each module stores a unique part of the data table, and the storage addresses of entries are inconsecutive, using a processing module to determine storage modules based on key information and mapping relationships to reduce conflicts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If data is stored in consecutive addresses in storage modules, then storage efficiency is improved, but parallel search conflict probability increases
Solution Approach 1:
The patent divides the data table into multiple partitions and stores them across multiple storage modules. Each storage module contains a specific partition of the data table, so that parallel search requests are distributed to different storage modules, reducing conflict probability while maintaining storage efficiency.
Solution Approach 2:
The patent introduces a new dimension of data organization by using partition IDs and storage module mappings. Instead of storing data in simple consecutive addresses, it adds a partition dimension that maps data to specific storage modules, thereby reducing parallel search conflicts without sacrificing storage efficiency.
2Reliability
If multiple storage modules are used to store data table, then parallel search conflict probability is reduced, but storage resource overhead increases
Solution Approach 1:
The patent makes each storage module universal by enabling it to store different partitions of the data table. The same storage module structure can handle any partition, eliminating the need for dedicated storage resources for each partition and reducing overall storage overhead.
Solution Approach 2:
The patent changes the parameter of data distribution by using hash functions and partitioning strategies. This allows efficient utilization of storage resources across multiple modules while maintaining low storage overhead through optimized data placement and indexing.
3Productivity
If data is distributed across multiple storage modules, then parallel search performance is improved, but processing logic complexity increases
Solution Approach 1:
The patent performs preliminary actions by pre-computing partition assignments and creating mapping relationships between data entries and storage modules. This preliminary organization simplifies the actual search process, as the system only needs to follow pre-established paths rather than making complex decisions during parallel searches.
Solution Approach 2:
The patent introduces mapping relationships and partition tables as intermediaries between the search requests and the actual data storage. These intermediaries simplify the processing logic by providing a clear lookup mechanism that directs searches to the correct storage modules without requiring complex distributed search algorithms.
4Reliability
If inconsecutive storage addresses are used, then parallel search conflict is reduced, but storage access efficiency decreases
Solution Approach 1:
The patent segments the data table into partitions that are distributed across storage modules. Each partition maintains its own consecutive address space, preserving storage access efficiency within each module while the overall distribution across modules reduces parallel search conflicts.
Solution Approach 2:
The patent resolves the address consecutiveness issue by adding a partition dimension. Data addresses are no longer just linear offsets but include partition identifiers, allowing inconsecutive global addresses while maintaining consecutive local addresses within each storage module, thus preserving access efficiency.
Data Source
AI summary
A data search apparatus includes a plurality of storage modules configured to store a data table. Any one of the plurality of storage modules stores a part of entries of the data table. Any entry in the data table is unique in the plurality of storage modules, and storage addresses of entries stored in at least one of the plurality of storage modules are inconsecutive.