Metadata fault-tolerant recovery method in distributed storage system

Abstract

The present invention provides a metadata fault-tolerant recovery method in a distributed storage system. When power of a metadata server is cut off, the metadata in a memory may be inconsistent with the metadata in metadata files because that some operations have been performed on the files in advance and each file has a log after successful operation, and the operations on the files have resulted in metadata update which has not been timely written into the metadata files. When a recovery operation is performed to recover the metadata, first, a latest metadata file is loaded into a memory; then, a log from which the recovery begins is determined according to a version value, an execution on a log being actually a re-execution on the previous operation of a file; and when executions on all logs are completed, the metadata in the memory is written into the metadata files, newly generated metadata in the files are consistent with the metadata in the memory when the power is cut off.

Description

technical field [0001] The invention relates to the field of information storage, in particular to a metadata fault-tolerant recovery method in a distributed storage system. Background technique [0002] Traditional network storage systems use centralized storage servers to store all data. Storage servers become the bottleneck of system performance and the focus of reliability and security, which cannot meet the needs of large-scale storage applications. The distributed network storage system adopts a scalable system structure, uses multiple storage servers to share the storage load, and uses the location server to locate and store information. It not only improves the reliability, availability and access efficiency of the system, but is also easy to expand. More and more enterprises are starting to use distributed storage systems to store all data. [0003] However, for most distributed file systems, metadata and data are usually separated and stored in different servers, ...

AI Technical Summary

Problems solved by technology

[0004] Usually, the traditional metadata fault-tolerant recovery scheme is to store most of the metadata on the local disk of the metadata server or backup server, and only a small part of the metadata is stored in the memory of the metadata server to meet the needs of real-time metadata request performance. and the reliability of metadata, the problem brought about by this is that the performance of the metadata server will be greatly affected. At the same time, if the metadata server fails, the metadata stored in the memory cannot be written to the disk in time. cause loss, the consequences are more serious

Images