Multi-SSD Storage with Reliability-Driven Erasure Coding and Replication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing storage systems with multiple SSDs face challenges in maintaining data reliability and performance due to varying wear leveling, leading to deteriorated data reliability and service quality when fixed fault tolerance or failure tolerance methods are applied.
Innovation Solution
A storage system with a system controller that monitors SSD reliability and dynamically adjusts data protection methods between erasure coding and replication based on SSD reliability status, ensuring data reliability and performance by managing SSDs with erasure coding for high reliability and replication for lower reliability SSDs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a fixed fault tolerance or failure tolerance method is applied to multiple SSDs, then data protection is provided, but data reliability and service quality deteriorate over time due to varying wear leveling
Solution Approach 1:
The system dynamically adjusts the data protection method for each SSD based on its real-time reliability status. SSDs are monitored continuously, and when their reliability status changes (e.g., from normal to degraded), the system automatically switches their protection method between erasure coding and replication, adapting to the changing conditions rather than using a fixed method for all SSDs
Solution Approach 2:
Different data protection methods are applied to different SSDs based on their individual reliability statuses. High-reliability SSDs use erasure coding while degraded SSDs use replication, allowing each SSD to receive the appropriate level of protection tailored to its specific condition rather than applying a uniform approach
2Productivity
If erasure coding is applied to all SSDs for efficient storage, then storage efficiency is improved, but performance degradation occurs when SSDs deteriorate
Solution Approach 1:
The system changes the protection parameter (method type) based on SSD reliability status. When an SSD's reliability status deteriorates below a threshold, the system changes its protection method from erasure coding to replication, thereby adapting the storage efficiency vs. performance trade-off to the current state of each SSD
Data Source
AI summary
Methods of operating storage systems that include a plurality of solid state drives, and the storage systems, are described. For example, a method may include, configuring the plurality of solid state drives into an erasure coding set for applying erasure coding, monitoring a reliability status of each of the plurality of solid state drives, selecting at least one solid state drive whose reliability status is lower than a reference value from among the plurality of solid state drives to exclude the at least one solid state drive from the erasure coding set, and constructing a replication set for applying data protection of a replication method by using the at least one solid state drive excluded from the erasure coding set.


