Computational Storage Compression Using Deferred Threshold Triggers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data storage technologies face a tradeoff between energy savings and performance, with current computational storage solutions not offering adaptive compression that minimizes CPU usage and energy consumption without performance impact.
Innovation Solution
Implementing deferred adaptive compression in computational storage, where data is initially uncompressed, and compression is initiated only when storage space utilization reaches certain thresholds, with background processing and intelligent data handling to minimize energy usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data is compressed continuously to save storage space, then storage efficiency is improved, but energy consumption and CPU usage increase
Solution Approach 1:
The system implements periodic compression actions based on storage threshold levels rather than continuous compression. When storage utilization exceeds a predefined threshold, compression is triggered; when below the threshold, compression is suspended. This periodic approach maintains storage efficiency while significantly reducing energy consumption during low-utilization periods.
Solution Approach 2:
The compression strategy dynamically adjusts based on real-time storage utilization metrics. The system transitions between compressed and uncompressed states according to threshold levels, making the compression behavior adaptive rather than static. This dynamic approach optimizes the balance between storage efficiency and energy consumption based on actual system conditions.
2Quantity of substance
If compression is applied to all data to maximize storage capacity, then storage efficiency is improved, but performance and response time deteriorate
Solution Approach 1:
The system applies compression selectively to specific data blocks rather than uniformly to all data. Only data blocks that exceed storage thresholds are compressed, while other blocks remain uncompressed for fast access. This local application of compression maintains high storage capacity utilization while preserving performance for frequently accessed or non-critical data.
Solution Approach 2:
The system performs partial compression actions only when necessary to meet storage requirements, rather than applying full compression to all data. This partial approach achieves sufficient storage capacity optimization without the performance penalty of comprehensive compression, allowing the system to do just enough compression to meet storage goals while maintaining productivity.
3Quantity of substance
If compression operations are performed frequently to maintain storage efficiency, then storage space optimization is improved, but system complexity and operational overhead increase
Solution Approach 1:
The system implements a feedback mechanism that monitors storage utilization levels and automatically triggers or suspends compression operations based on threshold comparisons. When storage utilization exceeds the upper threshold, compression is activated; when it falls below the lower threshold, compression is deactivated. This feedback-driven approach simplifies management by automating compression decisions based on simple metric comparisons rather than complex operational rules.
Data Source
AI summary
Provided are a method, system, and computer program product for performing deferred adaptive compression via a computational storage. On a drive that is initially empty, all data is written as uncompressed. In response to a first threshold level of storage space utilization being reached on the drive, compression of existing data on the drive is initiated. In response to a second threshold level of storage space utilization being reached on the drive, operations are performed for compressing incoming data and then writing the compressed incoming data to the drive, wherein the second threshold level is higher than the first threshold level.