Storage System Virtual Volume Group Data Loss Localization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In RAID storage systems, when a malfunction occurs in a storage apparatus with a redundancy level exceeding the capacity pool, data loss spreads across the entire virtual volume, leading to prolonged data restoration times and deteriorated load balancing performance.
Innovation Solution
A storage system configuration where virtual volumes are created with multiple redundancy levels, limiting drive combinations to ensure uniform distribution across storage pools, and using different combinations for each virtual volume group to localize data loss while maintaining load balancing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If data is distributed across a capacity pool to shorten rebuild time, then rebuild speed improves, but data loss spreads to the entire virtual volume when drive failures exceed redundancy level
Solution Approach 1:
The capacity pool is segmented into multiple independent RAID groups, each with its own redundancy level. When a drive fails, only the affected RAID group needs rebuilding, not the entire virtual volume. This segmentation localizes data loss and rebuild operations to specific segments rather than affecting the whole system.
Solution Approach 2:
Different RAID groups are assigned different redundancy levels based on their specific requirements. Critical data receives higher redundancy (e.g., RAID 6) while less critical data uses lower redundancy (e.g., RAID 5). This local quality approach optimizes both reliability and rebuild performance for different data portions independently.
2Reliability
If the pool is divided into multiple pools to localize data loss, then data loss localization improves, but load balancing performance of host I/O and rebuilding deteriorates
Solution Approach 1:
The RAID group structure serves multiple functions simultaneously: it localizes data loss to specific groups, maintains load balancing across all drives in the capacity pool, and provides flexible redundancy levels. Each RAID group is a universal building block that can be configured differently while participating in the same overall capacity pool management system.
Solution Approach 2:
Instead of dividing the capacity pool into separate isolated pools, the invention introduces a new dimension of organization by creating RAID groups as intermediate layers within the unified capacity pool. This hierarchical structure allows data loss localization at the RAID group level while maintaining global load balancing across the entire pool through uniform drive assignment rules.
3Quantity of substance
If storage apparatus capacity increases to meet growing data needs, then storage capacity improves, but rebuild time increases proportionally
Solution Approach 1:
Large capacity pools are divided into multiple smaller RAID groups. When rebuild is needed, only the data within the failed drive's RAID group must be reconstructed, not the entire capacity pool. This segmentation reduces the effective rebuild scope and time while allowing the overall system to scale to larger capacities.
Solution Approach 2:
Instead of rebuilding the entire virtual volume or capacity pool when a drive fails, the system performs partial rebuild only on the affected RAID group. This partial action approach reconstructs only the necessary portion of data, significantly reducing rebuild time while maintaining data integrity for the entire system.
Data Source
AI summary
A storage apparatus includes a plurality of storage devices, and a control unit for providing a predetermined storage area of the plurality of storage devices to the host computer as a virtual volume group including one or more virtual volumes. The control unit configures one or more data sets having one or more redundancy levels from the plurality of storage devices, provides a storage area of a storage pool including the plurality of data sets to a part of a storage area of the virtual volume, limits a combination of the storage devices configuring the data sets to be assigned to the virtual volume to a given number of combinations of two combinations or more, uniformly distributes the storage devices, and uses a given number of different combinations of the storage devices to be assigned to the virtual volume in units of the virtual volume group.