Parallel RAID Rebuilding via Stripe Mapping

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Traditional RAID rebuilding methods are slow and unreliable, especially as the number of disks grows, due to limited bandwidth and rebuild speed, leading to potential system failures when multiple disks fail during the rebuilding process.

Innovation Solution

The method involves dividing a RAID group into multiple stripes and mapping these stripes across other RAID groups for parallel rebuilding, ensuring that data is rebuilt efficiently without compromising system reliability by redirecting read and write operations and utilizing spare disks effectively.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of time

If traditional sequential RAID rebuilding is used, then system simplicity is maintained, but rebuilding time becomes excessively long (several days to weeks)

Engineering Contradiction:
Improverebuilding timeVSAvoidrebuilding process complexity
Core Design Contradiction:
Loss of timeVSDevice complexity

Solution Approach 1:

The patent divides the RAID group into multiple stripes (first stripe, second stripe, third stripe, etc.) and rebuilds them in parallel across different RAID groups. This segmentation transforms the single sequential rebuild process into multiple parallel operations, dramatically reducing total rebuilding time from weeks to hours while managing complexity through systematic organization of the rebuild operations.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces a new dimension to the rebuild process by mapping stripes from a failed RAID group to multiple other RAID groups simultaneously. Instead of rebuilding within a single RAID group sequentially, the solution distributes rebuild operations across multiple RAID groups in parallel, effectively adding a spatial dimension to the rebuild process that accelerates recovery.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Productivity

If parallel rebuilding across multiple RAID groups is implemented, then rebuilding speed increases significantly, but system reliability may be compromised during the rebuilding process

Engineering Contradiction:
Improverebuilding speedVSAvoidsystem reliability during rebuild
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent performs preliminary actions by first mapping the stripes to target RAID groups and establishing the parallel rebuild framework before actually executing the data reconstruction. This preliminary setup ensures that all necessary resources and pathways are prepared in advance, enabling fast parallel rebuilding while maintaining system stability and reliability throughout the process.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent uses spare disks as intermediaries in the parallel rebuild process. Data is reconstructed through these intermediary spare disks before being written to the final destinations, providing an additional layer of reliability and error checking that protects against data loss even while performing high-speed parallel operations across multiple RAID groups.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Quantity of substance

If disk capacity is increased to 2T or 4T, then storage space is improved, but rebuilding time extends to several days or weeks due to limited bandwidth and IOPS

Engineering Contradiction:
Improvestorage capacityVSAvoidrebuilding time
Core Design Contradiction:
Quantity of substanceVSLoss of time

Solution Approach 1:

The patent segments large-capacity disks (2T or 4T) into multiple smaller stripes, allowing parallel processing of the rebuild operation. This segmentation enables the system to handle large storage capacities without proportionally increasing rebuild time, as multiple stripes are reconstructed simultaneously across different RAID groups rather than sequentially.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent adds a parallel processing dimension to handle large-capacity disks by distributing stripe reconstruction across multiple RAID groups. This dimensional approach allows the system to scale storage capacity to 2T or 4T per disk while maintaining reasonable rebuild times through concurrent operations.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

4Reliability

If one or more disks fail during the prolonged rebuilding process, then RAID reliability is greatly degraded, but shortening rebuild time requires complex parallel operations

Engineering Contradiction:
ImproveRAID reliabilityVSAvoidrebuild operation complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments the RAID group into multiple independent stripes that can be rebuilt in parallel. This segmentation isolates failures to specific stripes rather than affecting the entire RAID group, allowing other stripes to continue rebuilding successfully even if one disk fails during the process. The systematic organization of stripes manages the complexity of parallel operations while improving reliability.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent distributes stripe reconstruction across multiple RAID groups in parallel, creating a multi-dimensional rebuild architecture. This approach reduces the window of vulnerability during rebuilding by completing operations faster, while the distributed nature of the operations across different RAID groups provides redundancy that protects against total failure.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Data Source

PatentUS10496483B2Method and apparatus for rebuilding redundant array of independent disks
Publication Date: 2019.12.03 EMC IP HLDG CO LLC
  • US10496483B2 patent drawing
  • US10496483B2 patent drawing
  • US10496483B2 patent drawing

AI summary

Embodiments of the present disclosure relate to a method and apparatus for rebuilding redundant array of independent disks (RAID). The method comprises dividing a RAID group including a disk into a plurality of stripes in response to detecting a failure of the disk, and each stripe includes a disk slice across all disks of the RAID group. The method further comprises mapping the plurality of stripes to other RAID groups respectively and rebuilding data on the plurality of stripes in other RAID groups. Embodiments of the present disclosure rebuild in parallel the stripes of the failed RAID group in other RAID groups, thereby effectively increasing the rebuilding speed for RAID and ensuring reliability of the RAID group.