Virtual Copy Address Mapping for Scale-Out Storage Nodes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In scale-out storage systems, transferring data between nodes for data utilization purposes leads to performance degradation and increased storage capacity consumption, as either all data is copied or data is transferred incrementally upon demand.
Innovation Solution
A storage system architecture that includes virtual volumes, pools, physical drives, memory for management information, and processors, allowing for the creation of virtual copies and log-structured writes across nodes without copying all data, thereby reducing performance degradation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If all data is copied from the first node to the second node, then data utilization is enabled, but storage capacity consumption increases and data transfer time is excessive
Solution Approach 1:
The patent extracts only the necessary mapping information (logical-to-physical address mappings) from the first node and stores it in the second node's memory, rather than copying entire data volumes. This allows the second node to access data logically while the actual data remains on the first node, dramatically reducing storage capacity consumption while enabling data utilization.
Solution Approach 2:
The patent introduces mapping information as an intermediary layer between the second node and the actual data on the first node. The mapping information acts as a mediator that enables the second node to access and utilize data without physically storing or transferring the actual data blocks, thus reducing storage capacity consumption while maintaining data accessibility.
2Quantity of substance
If data is transferred incrementally upon demand from the first node to the second node, then storage capacity consumption is reduced, but system performance deteriorates due to frequent data transfers
Solution Approach 1:
The patent performs preliminary action by pre-transferring and storing mapping information in the second node's memory before any data access occurs. This preliminary setup enables the second node to quickly access data through the stored mappings without needing to request or transfer actual data blocks from the first node during operation, thus maintaining high system performance while reducing storage capacity consumption.
Solution Approach 2:
The patent copies only the essential mapping information (not the actual data) from the first node to the second node. This selective copying of control information rather than full data copying reduces storage capacity consumption while enabling fast data access through the copied mappings, thereby maintaining system performance.
3Speed
If mapping information is copied from the first node to the second node, then data access performance is improved, but control information storage is required in the second node
Solution Approach 1:
The patent applies partial action by copying only the essential mapping information needed for data access rather than copying all control information or actual data. This partial copying improves data access performance in the second node while minimizing the storage requirements for control information, as only the critical address mapping data is transferred.
Data Source
AI summary
A first node performs copy (virtual copy) of address mapping between a virtual volume and a pool to a first virtual volume to create a third virtual volume in the first node. A second node performs mapping from a first pool volume in the second node to the third virtual volume in the first node, and links an address of the first pool volume, which is mapped to the third virtual volume, to an address of a second virtual volume in the second node.