Unlock instant, AI-driven research and patent intelligence for your innovation.

A Greedy Prefetching Data Recovery System and Recovery Method Based on Metadata Classification

A system and metadata recovery technology, applied in database distribution/replication, structured data retrieval, database design/maintenance, etc., can solve the problems of wasting disk space, inefficient access, affecting system recovery performance, etc., to reduce recovery time , The effect of improving recovery performance and saving storage overhead

Active Publication Date: 2019-05-14
广州摩翼信息科技有限公司
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Since the logically continuous data is physically dispersed in different disk locations, the fragmentation generated by the traditional deduplication method seriously affects the recovery performance of the system
Some previous optimization methods tried to use the rewriting algorithm during backup to improve the later recovery performance. However, the disadvantage of the rewriting algorithm itself is that better data recovery performance must be obtained at the expense of the deduplication rate. result in wasted disk space
Moreover, the traditional method only generates a copy of backup metadata for later recovery after the backup is completed, resulting in frequent and inefficient access to the metadata on the disk during the recovery process, which affects the recovery performance of the system

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A Greedy Prefetching Data Recovery System and Recovery Method Based on Metadata Classification
  • A Greedy Prefetching Data Recovery System and Recovery Method Based on Metadata Classification

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0033] This embodiment discloses a greedy prefetch data recovery system based on metadata classification.

[0034] The purpose of data backup is to restore. Since logically contiguous data is physically dispersed in different disk locations, the fragmentation generated by the traditional deduplication method seriously affects the recovery performance of the system. Some of the existing optimization methods try to use the rewrite algorithm during backup to improve the later recovery performance. However, the drawback of the rewrite algorithm itself is that it must sacrifice the deduplication rate to obtain better data recovery performance. result in wasted disk space. And the traditional method only generates a copy of backup metadata for later recovery after the backup is completed, which causes the system to frequently and inefficiently access the metadata on the disk during the recovery process.

[0035] like figure 1 As shown, this embodiment discloses a greedy prefetchi...

Embodiment 2

[0048] This embodiment discloses a greedy prefetch data recovery method based on metadata classification, and the data recovery method specifically includes the following steps:

[0049] S1. When the data recovery system receives a request to restore a certain backup data, the metadata read module opens the backup metadata information file from the disk, and pre-reads all the file metadata in the backup metadata information file into the LRU cache of the memory. , through the content of the file metadata in the metadata file, the block metadata required for restoring the backup can be known, and the fingerprint of the block, the serial number of the container where the block is located, and the size of the block can be obtained.

[0050] S2, greedy reading of block metadata, the file metadata pre-read into memory is divided into multiple segments, each segment contains backup metadata information of multiple files, and each segment will induce a read block manipulation of meta...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a greedy prefetch type data recovery system and recovery method based on meta-data classification; the data recovery system comprises a read meta-data module, a block data searching module and a recovery data module connected in sequence, wherein the read meta-data module is used for parallelly reading backup meta-data information; the block data searching module is used for providing a block searching function based on LRU buffer memory in a memory, and managing recovery backup relating hot data; the recovery data module is used for rebuilding the desired data in a local disk. The system and method can organize the backup meta-data into file meta-data and block meta-data, can fully prefetch the file meta-data into the memory and segment the same, and can carry out greedy prefetch on the block meta-data, thus realizing high efficiency data recovery while ensuring the dereplication rate and fully utilizing hardware resource; data block rewrite operation is not needed in the backup data dereplication process, thus ensuring data dereplication rate.

Description

technical field [0001] The invention relates to the technical field related to data recovery in a deduplication backup system, in particular to a greedy prefetch data recovery system and recovery method based on metadata classification. Background technique [0002] With the advent of the era of big data, the rapid growth of data volume has brought great challenges to the limited storage space of data centers. Data deduplication technology greatly reduces the disk overhead required to store data and the bandwidth required for network transmission, and has gradually become a key data reduction technology for today's backup systems. The purpose of storing data in the deduplication backup system is to restore data in a timely manner in the future. If an enterprise cannot restore data in time when a disaster such as a disk failure or database-related file damage occurs, the loss will be immeasurable, and the data backup will also change. It doesn't make much sense. Since logic...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06F16/27G06F16/21
CPCG06F16/21G06F16/27
Inventor 邓玉辉杨儒
Owner 广州摩翼信息科技有限公司