Storage Compression Verification with Parallel Decompression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data compression in storage systems leads to a deterioration of write performance due to the need for immediate decompression and data checks, which affects the reliability and efficiency of data storage.
Innovation Solution
A storage system with a dual controller configuration that includes a compression circuit to generate compressed data and a decompression circuit to verify data consistency, allowing both compression and decompression processes to be executed in parallel, thereby preventing the deterioration of compression throughput and ensuring high reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data compression processing is executed sequentially before storage, then data reliability is improved through decompression verification, but write performance deteriorates due to sequential processing bottleneck
Solution Approach 1:
The decompression circuit performs verification processing in advance on compressed data before it is fully written to storage. By preparing verification early and using pipelined parallel processing where compression and decompression occur simultaneously in different circuit stages, the system avoids sequential bottlenecks while ensuring data reliability through pre-verification.
2Reliability
If decompression verification is performed after compression, then data consistency is confirmed, but compression throughput deteriorates due to sequential execution
Solution Approach 1:
The decompression circuit begins verification processing on compressed data as soon as it is generated, before the compression process is fully complete. This preliminary verification approach allows the system to confirm data consistency without waiting for sequential completion of all compression operations, thereby maintaining high compression throughput while ensuring reliability.
Solution Approach 2:
The patent transitions from sequential single-dimension processing to parallel multi-dimension processing by implementing independent compression and decompression circuits that operate simultaneously. This dimensional change allows verification processing to occur in parallel with compression operations, eliminating the sequential bottleneck and maintaining high throughput while ensuring data consistency.
3Reliability
If immediate decompression and data check are executed after compression, then reliability is ensured, but write performance deteriorates further
Solution Approach 1:
Verification processing is performed in advance on compressed data within the decompression circuit before data is fully written to storage. This preliminary verification eliminates the need for separate post-compression checking steps, ensuring reliability while maintaining write performance through efficient pipelined parallel processing.
Solution Approach 2:
The patent implements parallel processing across independent compression and decompression circuit dimensions, allowing verification to occur simultaneously with compression and write operations. This multi-dimensional parallel execution ensures reliability through immediate verification while preventing deterioration of write performance by eliminating sequential processing stages.
Data Source
AI summary
Deterioration of compression throughput including a decompression check after data compression is suppressed. Provided is a storage system including an interface and a controller. The controller includes a compression circuit configured to generate compressed data by compressing received data received via the interface; and a decompression circuit configured to decompress the compressed data before storing the compressed data in a storage drive to confirm data consistency. The compression circuit sequentially executes a compression task of the received data, sequentially generates packets of the compressed data, and transfers the packets to the decompression circuit. The decompression circuit decompresses the received packet in parallel with the compression task.