RAID Extent Metadata for Incremental Data Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional storage device data management schemes do not support incremental data recovery or rebuild for disk slice-based extents, leading to inefficient data recovery and unnecessary data read/write operations, which reduces the service life of disks.
Innovation Solution
A data management method that acquires metadata of RAID extents to determine the state of disk slices and copies identification and degradation numbers into a metadata page, allowing for incremental data recovery by matching the RAID extent with its corresponding metadata for valid data recovery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If full data recovery is performed when a disk slice fails, then data reliability is restored, but data recovery time increases and disk service life decreases
Solution Approach 1:
The system performs preliminary actions by maintaining metadata that records the state of each disk slice (normal, degraded, or failed) and the degradation number indicating how many times a slice has changed from normal to degraded state. This preliminary information preparation enables the system to quickly determine whether incremental recovery is possible without performing full recovery, thus reducing data recovery time while maintaining reliability.
Solution Approach 2:
The patent segments the data recovery process into two types: full recovery and incremental recovery. By dividing the recovery approach based on the failure mode and metadata state, the system can select the appropriate recovery method. When only individual disk slices fail (rather than entire disks), incremental recovery is performed, which recovers only the affected portions rather than the entire dataset, thereby reducing recovery time.
2Reliability
If full data recovery is performed for individual disk slice failures, then data完整性 is ensured, but the number of data read/write operations increases
Solution Approach 1:
The system applies partial action by performing incremental recovery instead of full recovery when appropriate. The metadata tracks which disk slices have failed and their degradation history, allowing the system to recover only the specific portions of data that are actually needed. This avoids the excessive action of reading and writing entire disks when only small portions need recovery, thereby reducing unnecessary data read/write operations and energy consumption.
Solution Approach 2:
The system uses feedback from metadata about disk slice states and degradation numbers to determine the appropriate recovery scope. The metadata provides real-time information about which slices are normal, degraded, or failed, enabling the system to adjust the recovery operation dynamically. This feedback mechanism ensures data integrity by recovering only what is necessary while avoiding unnecessary operations.
3Productivity
If traditional data management scheme is used without incremental recovery support, then system complexity remains low, but data recovery efficiency decreases
Solution Approach 1:
The system prepares preliminary metadata structures in advance that track disk slice states and degradation numbers. This preliminary action adds some complexity to the data management system, but it enables significantly higher data recovery efficiency by allowing incremental recovery operations. The metadata structure is once-created and then reused to quickly determine recovery scope, making the initial complexity worthwhile for the ongoing efficiency gains.
Data Source
AI summary
A technique involves: in response to receiving a request to write data to a storage device, acquiring metadata of a redundant array of independent disks (RAID) extent that is associated with the request to write data, the RAID extent being built based on disk slices comprised in a plurality of disks of the storage device; determining, based on the metadata, a disk slice of data write-enabled RAID extent to write the data; and copying an identification and a degradation number into a metadata page associated with a mapper for data recovery, the identification and degradation number being comprised in the metadata and associated with the data write-enabled RAID extent, the mapper being used to map a physical space and a virtual logical space, and the degradation number indicating the number of times that the disk slice comprised in the RAID extent changes from being normal to being degraded.