Doubly Mapped Cluster Contraction for Storage Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional data storage techniques often result in underutilization of storage capacity in large storage groups, leading to inefficient use of processor, storage, memory, and network resources, especially when dealing with smaller data sets, as all disks in a node are considered part of the group, making it inefficient to allocate storage for smaller data amounts.
Innovation Solution
The implementation of a doubly mapped redundant array of independent nodes (RAIN) system, which allows for more granular storage by mapping data to portions of real disks, enabling data redundancy and flexibility in adding or removing storage components without losing access to data, using a double mapping topology and parity stripes to ensure data accessibility even with failures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If all disks of nodes are considered part of a storage group in conventional ECS systems, then large storage capacity is achieved, but storage utilization becomes inefficient for smaller data sets
Solution Approach 1:
The patent segments storage resources by introducing a hierarchical mapping structure where a logical storage group is divided into multiple logical disks, each mapped to portions of physical disks across multiple nodes. This segmentation allows flexible allocation of storage capacity without requiring all physical disks to be actively utilized, thereby resolving the contradiction between maintaining large storage capacity and achieving efficient utilization for varying data sizes.
2Productivity
If physically apportioning smaller groups with fewer nodes and disks is done, then storage allocation efficiency improves, but processor, storage, and memory resource usage becomes inefficient
Solution Approach 1:
The patent creates a universal mapping layer that allows the same physical storage infrastructure to serve multiple logical storage groups of different sizes. The mapping structure enables dynamic allocation where physical disks can be shared across multiple logical groups, allowing efficient storage allocation for smaller data sets while maintaining the ability to utilize the full physical infrastructure when needed, thus avoiding wasted computing resources.
3Quantity of substance
If a node with many disks is used, then large storage capacity is available, but the storage cannot be efficiently apportioned for smaller data amounts
Solution Approach 1:
The patent introduces a logical dimension between physical storage and data allocation by creating multiple levels of mapping (logical storage group → logical disks → physical disks). This dimensional transformation allows the system to present flexible, variable-size storage volumes to users while the underlying physical infrastructure maintains its fixed, high-capacity structure, thereby achieving both large capacity and operational flexibility.
Data Source
AI summary
Contraction of a doubly mapped redundant array of independent nodes, e.g., a doubly mapped cluster, is disclosed. Different mappings of data for a doubly mapped cluster corresponding to different uses of computing resources. Where a computing resource parameter indicates the computing resource is underutilized, an alternative mapping of the doubly mapped cluster can be undertaken. The alternative mapping can better utilize the computing resources. The contraction of the doubly mapped cluster can maintain access to stored data. The contraction can preserve data protection set integrity. The contraction can result in the doubly mapped cluster comprising fewer mapped nodes after the contraction but can avoid wholesale moving of corresponding data stored in a real cluster. As such, contraction of a doubly mapped cluster can be distinct from scaling-in of a doubly mapped cluster.