Composite Data Object Short Identifiers for Database Caching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing identifier systems for data objects, particularly those with composite identifiers, face challenges in memory consumption and system performance due to large identifier sizes and inefficient caching techniques, as conventional hash-based short identifiers fail to preserve grouping and are evenly dispersed, leading to inefficient database operations.
Innovation Solution
Generating short identifiers that consist of a sort order portion and a hash portion for composite component identifiers, ensuring adjacency and global uniqueness, which are stored in databases to facilitate efficient querying and caching.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If composite identifiers are used to identify data objects, then object identification accuracy is improved, but memory consumption increases
Solution Approach 1:
The composite identifier is segmented into multiple components (e.g., table identifier, record identifier, field identifier) that are stored separately. This allows the system to maintain the full identification capability while reducing the memory footprint of individual identifier references, as only the necessary components are loaded into memory based on query requirements.
Solution Approach 2:
The patent extracts and stores only the essential identification components in memory, while keeping the complete composite identifier structure in persistent storage. This extraction approach allows the system to maintain accurate object identification while minimizing in-memory storage requirements by keeping only the minimum necessary identifier data active.
2Quantity of substance
If hash value based short identifiers are used to reduce memory consumption, then memory consumption is reduced, but grouping of objects cannot be preserved
Solution Approach 1:
The patent applies different quality characteristics to different parts of the identifier system. The in-memory identifier components are optimized for speed and compactness, while the persistent storage maintains the complete grouping information. This local quality differentiation allows short identifiers to reduce memory consumption while the full composite identifiers in storage preserve object groupings and relationships.
Solution Approach 2:
The patent introduces a temporal dimension to the identifier system, where identifiers exist in multiple states: full composite identifiers in persistent storage and condensed in-memory representations. This dimensional approach allows the system to enjoy the benefits of both full identifiers (grouping preservation) and short identifiers (memory efficiency) by operating in different identifier dimensions as needed.
3Measurement precision
If hash values are evenly dispersed in the value range, then identifier uniqueness is improved, but database caching efficiency deteriorates
Solution Approach 1:
The patent performs preliminary organization of identifier components in persistent storage according to their grouping relationships and query patterns. By pre-organizing the data structure to reflect common query patterns and object groupings, the system enables efficient caching of relevant identifier components in memory without relying on hash dispersion, thus improving caching efficiency while maintaining identifier uniqueness.
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
A short identifier can be generated for each of a plurality of data objects. Each short identifier can include a first portion characterizing a sort order and a second portion characterizing a composite component identifier associated with the corresponding data object. The sort order can characterize a relationship (e.g., grouping, etc.) among at least a subset of the plurality of data objects. After the short identifiers have been generated, they can be stored in a database to enable queries of the database containing a short identifier to return data contained within the corresponding data objects. Related systems, apparatus, methods, computer program product and/or articles are also described.