Mapped Redundant Array Independent Nodes Storage Failure Mitigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional data storage systems, such as ECS systems, face inefficiencies in resource usage due to large real clusters being underutilized for smaller data storage needs, leading to suboptimal processor, network, and storage resource allocation.
Innovation Solution
The implementation of a mapped redundant array of independent nodes (RAIN) technology, which allows for the creation of smaller logical storage groups within larger real clusters, enabling more granular data storage and efficient resource utilization by logically apportioning storage space and using reserved real storage components to mitigate real node failures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If a large real cluster is used for data storage, then storage capacity is increased, but resource utilization efficiency deteriorates for smaller data storage needs
Solution Approach 1:
The patent divides the large real cluster into multiple smaller logical clusters through mapping. Each logical cluster corresponds to a subset of real storage devices, allowing flexible allocation of storage resources. This segmentation enables the system to provide smaller storage groups for different data storage needs while maintaining the underlying large real cluster infrastructure, thereby improving resource utilization efficiency.
Solution Approach 2:
The patent introduces a mapping layer as an intermediary between the real cluster and logical clusters. This mapping mechanism translates logical storage requests into physical storage operations on the real cluster. The mapping layer enables efficient resource allocation by allowing multiple logical clusters to share the same real cluster resources dynamically, resolving the contradiction between storage capacity and resource utilization efficiency.
2Device complexity
If real nodes are used directly for storage, then hardware simplicity is maintained, but failure mitigation capability deteriorates
Solution Approach 1:
The patent implements reserved real storage devices that are pre-configured and ready to replace failed real nodes. When a real node fails, the system can immediately allocate a reserved device to take over its role, minimizing data loss and service disruption. This preliminary preparation of replacement storage resources significantly enhances failure mitigation capability without requiring complex hardware changes.
Solution Approach 2:
The patent creates virtual copies of real nodes through mapping. Each real node can be mapped to multiple logical nodes, and when a real node fails, the system can redirect data access to copies stored on other real nodes. This copying mechanism provides redundancy and failover capability while maintaining the simplicity of the underlying hardware architecture.
3Productivity
If smaller real groups are created, then resource allocation efficiency is improved, but system complexity increases
Solution Approach 1:
The patent segments the storage system into multiple levels: real clusters at the physical level and logical clusters at the virtual level. This segmentation allows the system to present simplified smaller real groups to applications while maintaining the complexity management at the mapping layer. The segmentation enables efficient resource allocation through logical grouping without increasing the complexity of the underlying hardware.
4Reliability
If reserved real storage components are used to mitigate failures, then reliability is improved, but storage space consumption increases
Solution Approach 1:
The patent makes reserved real storage devices multi-functional. These reserved devices can serve multiple purposes: they act as replacement storage for failed nodes, they can be allocated to logical clusters when needed, and they maintain the same storage capacity as regular devices. This universality allows the system to improve reliability through reserved storage while minimizing the impact on overall storage space consumption by utilizing the same physical resources for multiple functions.
Data Source
AI summary
Mitigating the effects of a real node failure in a mapped redundant array of independent nodes, e.g., mapped cluster is disclosed. In response to a change in an accessibility to data stored on a real storage device of a real node of a real cluster, wherein the real storage device corresponds to a mapped storage device of a mapped node of a mapped cluster, substituting a reserved real storage device for the real storage device. The substituting the reserved real storage device can correspond to a change in a topology of the mapped cluster, wherein the change in the topology comprises replacing the mapped storage device with a substitute mapped storage device that corresponds to the replacement real storage device. The changed topology can enable writing of data to the substitute mapped storage device that can cause writing of corresponding data to the reserved real storage device.