Resyncing Mirrored Metadata Without Checksum Writes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
RAID-6 data storage systems face inefficiencies in write operations and data recovery due to the need for multiple read and write operations across primary and parity disks, which reduces capacity and increases latency.
Innovation Solution
A storage system utilizing a meta object and a capacity object to perform full-stripe writes, where the meta object stores data until a threshold is reached and then executes batch writes to the capacity object, updating a logical map for data location and using data integrity tokens for resynchronization of mirrored disks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If RAID-6 is used for fault tolerance, then data reliability is improved, but write efficiency deteriorates due to multiple read/write operations
Solution Approach 1:
The patent performs preliminary actions by pre-calculating and storing checksums with each data block during the write operation. When a disk fails, the resynchronization process can directly use these pre-stored checksums to identify and restore correct data blocks without performing time-consuming checksum calculations on the fly, thus improving write efficiency while maintaining data reliability
2Reliability
If RAID-6 is used for fault tolerance, then data reliability is improved, but capacity efficiency deteriorates due to parity disks
Solution Approach 1:
The patent implements self-service by embedding checksums directly within the data blocks stored on disks. Each data block contains its own integrity verification information, allowing the system to perform self-diagnosis and self-recovery without requiring separate checksum storage structures. This eliminates the need for dedicated checksum storage space, improving capacity efficiency while maintaining data reliability through the same fault tolerance mechanism
3Reliability
If separate checksum writes are performed for data recovery, then data integrity is improved, but write efficiency deteriorates
Solution Approach 1:
The patent merges the data storage and checksum storage operations into a single atomic write operation. The checksum is calculated and written together with the data block in one I/O operation, eliminating the need for separate checksum write operations. This combining approach maintains data integrity verification capability while significantly improving write efficiency by reducing the number of I/O operations
Data Source
AI summary
Methods, systems, and apparatus, including computer programs encoded on computer storage media, for resynchronizing data in a storage system. One of the methods includes receiving, by a first storage subsystem, a plurality of write requests corresponding to respective meta data blocks, wherein the first storage subsystem comprises a meta object; storing, by the first storage subsystem and for each write request, in each disk of the meta object, a version of the corresponding meta data block; determining that a particular disk of the meta object has failed; determining whether one or more valid versions of the meta data block are stored in respective other disks of the meta object; and in response to determining that one or more valid versions of the meta data block are stored in respective other disks of the meta object, resynchronizing the meta data block in the particular disk.