Cached Data Mirroring With Transitional Pairs for Node Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In a four-control storage system, when a faulty node returns to normal, the re-teaming process for mirrored pairs can lead to data confusion and loss, compromising data security due to the need for one node to exit and abandon corresponding data.
Innovation Solution
A data caching method that adds a faulty node to a recombined mirrored pair, forming a transitional mirrored pair with two surviving nodes, initiates a cached data re-mirroring process, and then deletes the redundant node to switch back to the initial mirrored pair, ensuring data continuity and security.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a faulty node returns to normal and re-teaming is performed according to the conventional mirrored pair method, then the node can be restored to normal operation, but one node must exit and abandon corresponding data, leading to data confusion and loss
Solution Approach 1:
The patent designates a target mirrored pair in advance before the faulty node returns to normal. When the node recovers, it is directly integrated into the pre-designated target mirrored pair rather than requiring dynamic re-teaming, which prevents data loss by having the restoration path prepared beforehand
Solution Approach 2:
The patent introduces a target mirrored pair as an intermediary structure that mediates between the faulty node and the original mirrored pair. This intermediary allows the faulty node to be gradually reintegrated without disrupting the original data relationships, avoiding data confusion and loss
2Adaptability or versatility
If a node exits during the re-teaming process to allow faulty node integration, then the faulty node can be restored, but data corresponding to the exiting node is abandoned, compromising data integrity
Solution Approach 1:
The system designates a target mirrored pair in advance before node failure or recovery occurs. This preliminary designation ensures that when a faulty node returns, there is already a prepared structure to receive it, eliminating the need for nodes to exit and preventing data confusion
Solution Approach 2:
The patent applies different roles to different nodes within the mirrored pair structure. The target mirrored pair has a specific designated role to receive the faulty node, while other mirrored pairs maintain their original data relationships. This local differentiation allows restoration without affecting other data
Data Source
AI summary
A data caching method, including: in response to a faulty node returning to normal, adding the faulty node to a recombined mirrored pair according to a mirrored pair where the faulty node is initially located to obtain a transitional mirrored pair whereby in a cached data re-mirroring process, when one node of the mirrored pair is faulty, data mirroring continues between other nodes; initiating, by a main node in the transitional mirrored pair, a cached data re-mirroring operation process to the faulty node; and in response to the main node in the transitional mirrored pair completing the cached data re-mirroring to the faulty node, deleting a redundant node in the transitional mirrored pair based on an initial mirrored pair to switch the transitional mirrored pair back to the initial mirrored pair for data caching.