Storage Data Fingerprinting for Hot-Data Cache Decisions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing computer systems face inefficiencies in data accessing without significant hardware upgrades, necessitating improved data processing efficiency without increasing costs.
Innovation Solution
A hot-data-based decision-making method for accessing data in storage devices, utilizing a data pre-processing step, old datum comparison, old datum updating, new datum updating, and new datum replacing steps, with fingerprint hashing and double comparison to determine data hotness and optimize storage unit usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the cache capacity is increased to store more frequently accessed data, then data accessing efficiency is improved, but hardware cost increases
Solution Approach 1:
The patent implements dynamic data classification by continuously updating hotness scores for data items based on access patterns. The system dynamically determines which data should be cached based on current access behavior rather than using static capacity allocation, allowing efficient use of limited cache space.
Solution Approach 2:
The patent introduces hotness scores as a parameter to classify data into hot, warm, and cold categories. By changing the parameter of data classification from simple presence/absence to hotness-based categorization, the system optimizes cache usage without increasing physical capacity.
2Speed
If the cache size is enlarged to improve data retrieval speed, then data accessing efficiency is improved, but system cost increases
Solution Approach 1:
The patent performs preliminary classification of data into hot, warm, and cold categories before actual data retrieval operations. By pre-sorting data based on hotness scores and access patterns, the system ensures that most frequently accessed data is positioned for fastest retrieval without requiring larger hardware capacity.
3Quantity of substance
If more storage units are added to increase data capacity, then data storage capability is improved, but data accessing efficiency deteriorates
Solution Approach 1:
The patent segments data into different categories (hot, warm, cold) based on access patterns and stores or manages these segments differently. This segmentation allows the system to maintain efficient access to hot data while accommodating larger overall data capacity through structured organization rather than uniform expansion.
Data Source
AI summary
A hot-data-based decision-making method for using in a computer having a storage device that includes plural storage units. The method includes processing a new datum to obtain a new data fingerprint and at least two storage index values; and determining whether a datum in the storage unit corresponding to any of the storage index values matches the new datum, and if so, generating an update hotness score by updating a pre-stored hotness score of an old data fingerprint of an old datum, wherein the old datum is hot if the update hotness score is not lower than a hotness threshold score, or is cold if it is lower; or, otherwise, determining whether the storage unit is blank, and if so, simply storing the new data fingerprint into the storage unit, or, otherwise, replacing an existing data fingerprint in the storage unit with the new data fingerprint.