Offloading RAID Parity Reconstruction to Storage Concentrators
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In conventional storage systems, reconstructing data lost due to failed storage devices is resource-intensive and processor-heavy, particularly for XOR calculations, which can burden the storage server and hinder efficient data recovery.
Innovation Solution
Offloading XOR computations to storage concentrators, which retrieve data and parity from operational storage devices, compute XOR values, and provide context information to the storage server for data reconstruction, thereby distributing the computational load and enhancing efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the storage server performs XOR calculations for data reconstruction, then data recovery can be achieved, but the computational burden on the storage server increases significantly
Solution Approach 1:
The patent introduces storage concentrators as intermediary devices between storage devices and storage servers. These concentrators perform the computationally intensive XOR calculations locally at the storage device level, acting as mediators that offload processing tasks from the storage server. The concentrators retrieve data and parity from operational storage devices, compute XOR values, and provide context information back to the storage server, thereby resolving the contradiction by distributing computational workload.
Solution Approach 2:
The patent segments the data reconstruction process into multiple independent tasks that can be performed in parallel. Instead of the storage server handling all XOR calculations sequentially, the system divides the workload across multiple storage concentrators and operational storage devices. Each concentrator handles specific reconstruction tasks independently, enabling concurrent processing and reducing the computational burden on any single server.
2Reliability
If the storage server performs all XOR determinations for data reconstruction, then data can be recovered, but the processor resources are consumed excessively
Solution Approach 1:
Storage concentrators serve as intermediary processing units that perform XOR determinations locally rather than requiring the storage server to handle all computations. These concentrators retrieve necessary data and parity information from operational storage devices, perform the XOR calculations to determine missing data, and return results to the storage server, thereby improving processor efficiency by distributing computational tasks.
Solution Approach 2:
The storage concentrators perform self-service by autonomously retrieving data from operational storage devices, computing XOR values, and providing context information without requiring continuous intervention from the storage server. This self-service capability allows the system to maintain data reconstruction functionality while reducing the processor resources consumed by the storage server.
3Reliability
If conventional storage systems perform data reconstruction locally, then data can be recovered, but the system complexity increases
Solution Approach 1:
The patent segments the storage system into distinct functional components: storage devices for data storage, storage concentrators for computation and data retrieval, and storage servers for coordination. This segmentation allows each component to have specialized functionality, simplifying the overall system architecture. The concentrators handle complex XOR calculations locally, while the storage server focuses on coordination and data management, reducing the complexity burden on any single component.
Data Source
AI summary
Methods and systems for a storage environment are provided. For example, one method includes receiving a request from a storage server at an offload engine for reconstructing data lost due to a failed storage device of a parity group having a plurality of storage devices; retrieving data and parity by the offload engine from the parity group storage devices that are operational; determining by the offload engine XOR of the retrieved data and parity; presenting XOR of data and parity by the offload engine to the storage server with context information associated with the retrieved data; and reconstructing lost data by the storage server using the XOR of data and parity and the context information provided by the offload engine.