A Data Reconstruction Method Based on Storage Objects

Abstract

Provides data reconstruction methods based on storage objects. The provided data reconstruction method includes, in response to the failure of the first drive in the storage pool, obtaining a storage object using the data block provided by the first drive; obtaining the first data block in an idle state from the storage pool and assigning it to the storage object; the storage object; storing the data stored in the data block of the first drive in the storage object to the first data block.

Description

technical field [0001] The present invention relates to storage systems, and more particularly, to a method and system for rebuilding data when a drive in the storage system fails. Background technique [0002] The existing RAID (Redundant Arrays of Independent Disks, redundant array of independent disks) technology aggregates multiple physical drives (for example, disks) into a disk group, divides the disk group into stripes, and uses redundancy technology on each stripe. Provide protection for data. In the RAID system of the prior art, when one or more of the multiple drives fails, a spare drive is activated and data reconstruction is performed to maintain the data protection capability and performance of the RAID system. After the failed drive is replaced, the data from the spare drive needs to be copied back to the new drive. RAID systems can provide various levels of data protection, for example, a RAID5 system can correct the failure of one drive, while a RAID6 syste...

AI Technical Summary

Problems solved by technology

In this process, all write operations will finally be concentrated on the spare drive 88 and the new drive, so that the data reconstruction performance of the RAID system in the prior art is limited to the performance of a single drive

[0004] There are at least the following problems in the RAID system of the prior art: (1) the data domain and the physical domain are a tightly coupled relationship (stripes directly correspond to physical storage locations), and this tightly coupled relationship will cause the RAID data reconstruction process to be extremely difficult. Long time, the performance of RAID system data reconstruction cannot grow linearly with the number of disks

And as the drive capacity increases, the data reconstruction process time will increase linearly

[0005] (2) Due to the long reconstruction time, a large amount of data reading and writing during the reconstruction process increases the burden on the drive, making the drive more prone to failure

[0006] (3) Partial reconstruction cannot be achieved. When a drive in the drive group fails, traditional RAID technology will perform reconstruction operations on all storage spaces of the entire drive, and the reconstruction of the entire drive will take a lot of time

[0007] (4) It is impossible to refactor according to priority, and traditional refactoring cannot perform priority refactoring according to fault level

[0008] (5) Data awareness cannot be achieved. The existing RAID system located at the block device layer has a fixed data layout and cannot know the data semantics of the upper-layer application. Therefore, data-aware reconstruction operations cannot be achieved. Recover data by means of a full drive rebuild

For the new solid-state drive (SSD, Solid State Drive), if the full drive reconstruction method is adopted, it will cause the spare drive or the replaced drive to be completely rewritten, which not only leads to a significant decline in performance, but also leads to SSD failure. Lifespan is affected

Images