Selective RAID Rebuild Using Time Counters for Interruption Recovery
Solution Overview
Problem
In multi-device storage systems, temporary interruptions such as power fluctuations or device disconnections can cause storage devices in a RAID set to go out of sync, leading to prolonged and costly rebuild times as data storage capacity increases.
Innovation Solution
Implementing time counters for each RAID block to determine which devices are out of sync, allowing for selective partial rebuilds of only the affected stripes, rather than a full rebuild of the entire RAID set.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a full rebuild of the RAID set is performed when a device goes out of sync, then data consistency across all devices is restored, but the rebuild time becomes very large and costly from service loss, system design, and energy perspectives
Solution Approach 1:
The patent divides the RAID set into individual stripes and identifies only the affected stripes that need rebuilding. Instead of rebuilding the entire RAID set, the system segments the rebuild operation to target only the specific stripes where data inconsistency occurred, significantly reducing rebuild time while maintaining data consistency.
Solution Approach 2:
The patent applies local quality by treating different parts of the RAID set differently based on their sync status. Devices or stripes that are out of sync receive rebuild operations, while synchronized devices are left untouched. This localized approach restores data consistency without unnecessarily processing the entire RAID set.
2Reliability
If a full rebuild of the RAID set is performed, then data consistency is restored, but energy consumption increases significantly
Solution Approach 1:
The patent segments the rebuild operation to affect only the specific stripes that are out of sync rather than the entire RAID set. This segmentation reduces the total data volume that needs to be read, processed, and written back, thereby significantly reducing energy consumption while still restoring data consistency where needed.
Solution Approach 2:
The patent applies partial action by performing rebuild operations only on the minimum necessary portions of the RAID set (the affected stripes) rather than conducting a complete rebuild. This partial rebuild approach consumes less energy while achieving the goal of restoring data consistency.
3Reliability
If a full rebuild of the RAID set is performed, then data consistency is restored, but service loss increases
Solution Approach 1:
The patent segments the rebuild scope to only include affected stripes, which reduces the duration of the rebuild operation. This shorter rebuild time minimizes service loss and maintains higher productivity, as the storage system can return to full operational capacity faster compared to a complete RAID set rebuild.
Solution Approach 2:
The patent performs partial rebuilds on only the necessary stripes rather than the entire RAID set. This partial action reduces the time the system spends in a degraded or rebuilding state, thereby minimizing service loss and maintaining better productivity and availability.
Data Source
AI summary
Systems and methods for selective rebuild of interrupted data storage devices in storage arrays are described. A controller determines an operating interruption of a data storage device in a redundant array of independent disks (RAID) configuration. In response to the interruption, the controller determines a last block time value for the last successfully stored RAID block in the interrupted storage device and one or more incomplete RAID stripes that the interrupted storage device did not complete. The controller then selectively rebuilds the incomplete RAID stripes from the other storage devices in the RAID configuration.