Entropy-Based Reduction Selection in Storage Controllers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data reduction techniques in storage systems often waste processing time and energy by attempting unsuitable reduction methods on certain data types, leading to inefficient performance.
Innovation Solution
A storage system with a controller that analyzes incoming data units for entropy values and compares them to thresholds to select suitable reduction techniques, such as deduplication or compression, thereby avoiding unsuitable reduction attempts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data reduction techniques are applied to all incoming data, then storage space is reduced, but processing time and energy are wasted on unsuitable data types
Solution Approach 1:
The patent applies preliminary action by calculating entropy values for incoming data units before committing to a reduction operation. The system performs a quick entropy analysis on sample data to predict suitability for deduplication or compression, and only then proceeds with the actual reduction operation if the entropy threshold indicates suitability. This preliminary assessment prevents wasted processing time on unsuitable data types.
Solution Approach 2:
The patent replaces mechanical trial-and-error reduction attempts with an entropy-based prediction system. Instead of mechanically applying reduction operations to all data and measuring results, the system uses entropy calculation (a different mechanical approach) to predict outcomes beforehand, substituting the trial-and-error mechanism with a predictive model that guides operation selection.
2Quantity of substance
If data reduction techniques are applied to all incoming data, then storage space is reduced, but energy consumption increases due to unsuccessful reduction attempts
Solution Approach 1:
The system performs preliminary entropy calculation on incoming data units to assess reduction suitability before executing energy-intensive reduction operations. By evaluating entropy values against predetermined thresholds, the system identifies data units likely to benefit from reduction, thereby avoiding energy waste on unsuitable data types while still achieving storage space reduction on appropriate candidates.
Solution Approach 2:
The patent changes the parameter of operation selection from a fixed approach (applying the same reduction technique to all data) to a dynamic approach based on entropy parameter analysis. The system adjusts which reduction operation to apply (or whether to apply any operation) based on the calculated entropy parameter of each data unit, optimizing energy consumption by matching operations to data characteristics.
3Quantity of substance
If reduction operations are attempted on unsuitable data types, then processing overhead increases, but storage space reduction is minimal or zero
Solution Approach 1:
The patent applies preliminary entropy analysis to assess data suitability for reduction before executing reduction operations. This preliminary action filters out data types unlikely to benefit from reduction, thereby minimizing processing overhead on unsuitable data while maximizing storage space reduction on suitable candidates. The entropy threshold comparison serves as a gatekeeping mechanism.
Solution Approach 2:
The patent introduces entropy calculation as an intermediary step between data reception and reduction operation execution. This intermediary mechanism evaluates data characteristics and mediates the decision of whether to proceed with reduction, preventing direct application of reduction operations on unsuitable data and thereby reducing unnecessary processing overhead.
Data Source
AI summary
Example implementations relate to storing data in a storage system. An example includes receiving, by a storage controller of a storage system, a data unit to be stored in persistent storage of the storage system. The storage controller calculates multiple entropy values for the data unit. The storage controller selects, based on the multiple entropy values, at least one reduction operation from multiple different reduction operations. The storage controller performs the selected at least one reduction operation on the received data unit.