Distributed storage data recovery scheduling method, system, device and storage medium

Abstract

The invention discloses a distributed storage data recovery scheduling method, system, equipment and storage medium. The method comprises: selecting a check block and a data block to construct an inversion matrix as a reference matrix, labeling the rows of the reference matrix in order, and recording addresses; calculating the inverse matrix of the reference matrix, and labeling the columns of the inverse matrix of the reference matrix in order; Input the inversion matrix to be scheduled, match the inversion matrix to be scheduled with the elements of the reference array, define the same label for the row with the same element, and record the new address; according to the row address of the reference array with the same label and the inversion to be scheduled The relative change of the row address of the matrix corresponds to the column elements of the inverse matrix of the scheduling reference matrix, and the inverse matrix after the scheduling of the input to-be-scheduled inverse matrix is ​​obtained, which is used for distributed storage data recovery. The invention can realize simple scheduling of elements of the reference array based on any reference array under the same error condition, and improve the decoding speed and efficiency of the erasure correction circuit under different disk reading sequences.

Description

technical field [0001] The data storage field of the present invention specifically relates to a distributed storage data recovery scheduling method, system, equipment and storage medium. Background technique [0002] Faced with the storage requirements of massive data, distributed storage has gradually replaced the dominant position of unified storage due to its advantages of low cost and good scalability, and has received more and more attention in theoretical research and practical applications. However, the reliability of distributed storage is insufficient. On the one hand, distributed storage systems mostly use cheap disks as storage nodes, and the reliability of each storage node is often not high. On the other hand, a distributed storage system usually includes many Nodes, due to software and hardware failures, human errors and other reasons, the system often has node failures. In order to improve the data reliability of the distributed storage system and ensure tha...

AI Technical Summary

Problems solved by technology

But the disadvantage is that it is not optimized for the characteristics of matrix coding, so although the calculation is regularized, it will introduce a lot of redundant operations

Especially when recovering data in a distributed storage scenario, for the same error code block, although the row and column elements are the same using the Gaussian elimination method, as long as the order in which the hardware reads data changes, the inverse matrix needs to be recalculated, which will cause huge resource consumption

Images