CN103713963A - Efficient file backup and restoration method - Google Patents

Abstract

The invention discloses an efficient file backup and restoration method and aims to solve the technical problems of changing the traditional backup methods and quickening file backup or restoration to save storage space on the premise of system running normality, and necessarily, restoring required files accurately and efficiently. According to the technical scheme, with time lapse, changes in files result in new file versions; a new version is formed by combing change data of one file and a former file version; during storage, the change data of the file can be divided into continuous storage blocks to be stored in a medium; if the data of the storage blocks are equal to that of one existing storage block, re-storage is not required; a memory assigns each storage block a unique ID; through a set of IDs, the files are recovered. The method has the advantages that re-storing same data blocks is not required so that storage space is saved; operations such as data copying and pasting are not required so that recovery speed and recovery efficiency are increased.

Description

A kind of efficient file backup and restoration methods

Technical field

The present invention relates to a kind of file backup and Yan compound method, be a kind of fast, reliable, the backup Yan compound method that can save storage space.

Background technology

Popularizing to routine work life of computing machine brought a lot of facilities, but when computing machine is subject to some attacks or disaster, in the face of the significant data of losing, people are in a rush.In the face of such stubborn problem, the conventional several method once of people: 1. use external equipment to copy out data; 2. use CD writer that data are carved into CD; 3. save the data in outside system partitioning; 4. diskette backup data; 5. use GHOST Backup Data or system.Above-mentioned Lung biopsy no doubt can have been protected significant data to a certain extent, but the problems such as inconvenient operation, easily maloperation, external equipment loss, waste resource emerge in an endless stream, and is not suitable for enterprise's use of mass data.

Some Enterprises need to regularly back up, and when data volume is merciless, increases, and this certainly will need more storage space and more bandwidth, and managerial personnel can not wish to destroy easily original data framework, but this needs more hardware to drop into.

So the multiple software of current backup Yan need to solve, and except real-time, accuracy etc., also needs to consider storage space.

The present invention proposes a kind of file backup Yan compound method of efficiently and accurately, and the method backup can save no small storage space, the method is applied in backup product, can obtain precise and high efficiency, gratifying backup effect.

Summary of the invention

The object of the invention is to propose a kind of efficient file backup Yan compound method, the method can be saved storage space, is used in backup product.

The technical problem to be solved in the present invention is to change traditional backup mode, quickening file backup or the multiple speed of Yan, saving storage space, the file that Yan appears again and needs accurately and efficiently if desired.

For solving the problems of the technologies described above, the present invention takes following technical scheme: when needs are during to amended file backup, first get file compared with the variable quantity data of version last time, variable quantity data are divided into continuous piece to be stored in medium and returns to ID, at this, calculate eigenwert the storage of data block, data block needs its eigenwert and the eigenwert of the data block having existed to carry out that this is right before being stored, this value is identical, two data blocks are identical, do not need repeated storage, directly return to the ID of this data block having existed, thereby, operating system has all generated data block data block to the modification of each file and has been stored unique ID that device distributes correspondence, the set of these ID is file change amount data-mapping.

File data mapping is the data-mapping of last version and the merging of relative last version file variable quantity data-mapping of file.Thus, during backup file, first obtain last FileVersion data-mapping, itself and file change amount data-mapping are merged, just obtain this file data mapping, backup this document completes.When file backs up for the first time, File Mapping is exactly this file change amount data-mapping.

Concrete, the attribute information of the independent storage file of meeting during each backup file, such as filename, file size, file modification time etc., the file change of reentrying amount data, variable quantity data are divided into continuous data block, each blocks of data piece for wherein, stores in medium, obtains corresponding ID, generate thus the mapping of file change amount, by merging the File Mapping of last version, obtain complete File Mapping, the mapping of storage this document.

When user needs Yan again during corresponding file, according to FileVersion, can obtain corresponding file data mapping, each ID is wherein obtained to corresponding storage block, then by these storage block composing documents.

The method of the invention can be saved storage space, reduces backup content, improves backup rate.The present invention is mainly that piecemeal is preserved file change amount data to medium, for wherein there being not repeated storage (while having stored bulk data piece in medium of data block in medium, the likelihood ratio that data block is identical is higher, repeated storage not, can save many spaces), reduce required storage space, raise the efficiency; In whole process, do not relate to and the operation such as copy, the consumption time is few, more can guarantee real-time.

Accompanying drawing explanation

Fig. 1---backup file process flow diagram.

Fig. 2---the file attribute of FileVersion and the logic relation picture of File Mapping under different time points.

The logic relation picture that Fig. 3---storage block ID and storage block content indicate.

Embodiment

Enumerating example is below elaborated to technical scheme of the present invention.

For certain file F, As time goes on, new variation causes new FileVersion to produce, F (t0)-->F (t1)---> ...---->F (ti) ..., F (ti) represents ti FileVersion constantly, Delta (ti) represents the variable quantity data constantly with respect to ti-1, and Map (Delta (ti)) represents file change amount data-mapping, and four following expression formulas are set up.

1.F(t0)＝0+Delta(t0)

2.F(ti)＝F(ti-1)+Delta(ti)

3.Delta(ti)＝∑B(k)

4.Map(Delta(ti))＝∑AID(B(k))

For the version F (ti) of the file F of moment ti, F (ti) be equivalent to ti-1 version constantly with respect to ti-1 constantly file change amount and, therefore expression formula 2 set up, special circumstances when expression formula 1 is initial back-up.

Fig. 1 is the process flow diagram of backup file.AlD (B (k)) represents the storage ID of storage block B (k), when storage change amount data are Delta (ti), Delta (ti) is divided into continuous storage block B (k), expression formula 3 is set up, AID (B (k)) is that B (k) distributes unique storage lD by storer for it while storing on medium, eigenwert the storage of calculating B (k), expression formula 4 is set up.If there is certain storage block B (j) in storer, its data are identical with B (k), storer will return to the AID of B (j) so, AID (B (k))=AID (B (j)), whether concrete relatively two data block contents are the same, can compare both eigenwerts.According to AID (B (k)), storer can obtain the data content of B (k).Thus, version F (ti) for the file F of any time ti, storage Delta (ti) obtains the corresponding with it mapping of file change amount Map (Delta (ti)), so Map (F (ti))=Map (F (ti-1))+Map (Delta (ti)).

The method of backup file: backup ti FileVersion F (ti) constantly, first obtain the variable quantity data Delta (ti) constantly with respect to F (ti-1), Delta (ti) is divided into continuous storage block, for each piece B (k) wherein, storage B (k) is in medium, obtain AID (k), generate thus Map (Delta (ti)), by merging Map (F (ti-1)), obtain complete Map (F (ti)), store M ap (F (ti)).

The multiple ti of the method for the multiple file of Yan: Yan FileVersion F (ti) constantly, obtains corresponding Map (F (ti)), for each AID (k) wherein, obtains storage block B (k), generates accordingly F (ti).

Fig. 2 expresses the file attribute of FileVersion and the logic relation picture of File Mapping under different time points.For example want the multiple file F constantly backing up at ti of Yan, can the first step obtain corresponding attribute (title and a series of attribute), second step obtains corresponding File Mapping.

Fig. 3 is the logic relation picture that storage block ID and storage block content indicate, and according to File Mapping, by wherein each No. ID, obtains corresponding storage block content this No. ID.Wait to draw each storage block content, content splicing just can be formed to file content, the attribute of file obtains, and has had file attribute and file content, and the object of the multiple file of Yan just can reach.

FilePath--preserves the file path of storage block content

Offset-storage block content is with respect to the skew of first address

The size of Length-storage block

Introduction of the present invention be a kind of multiple method of file backup Yan, the storage block of using in example is preserved hereof, in actual application, storage medium does not limit to.

Claims (1)

1. efficient file backup and a Yan compound method, is characterized in that comprising the following steps:

The first step, when file has the part of modification and need to back up, obtains file attribute information and file change amount data;

Second step, file attribute information is stored in medium, by file change amount data block storage, generates unique No. ID of each piece, and the mapping of spanned file variable quantity, comprises following step:

Step 1, file attribute information storage, corresponding with this time point backup file version;

Step 2, storage file variable quantity data, by variable quantity data allocations in continuous storage block, storage block stores in medium one by one, be stored device and distribute unique storage ID, if there is certain data block in storer, both data identical (can by feature at comparatively), the ID of the data block that storer has existed this so returns, storage allocation again not, the set that is these ID at this defined file variable quantity data-mapping, by above operation, otherwise utilizes data block ID can obtain its storage content;

The 3rd step, obtain the File Mapping of the last backup operation of file, shine upon with file change amount the mapping that merges the file that obtains backing up, if file is first backup, the File Mapping of the last backup operation does not exist, File Mapping is exactly the file change amount mapping of this time, preserves File Mapping, corresponding with this time point backup file version.So far, file backup completes;

The 4th step, when user needs, the multiple file of Yan, comprises following step:

(1) obtain needing respective file attribute and the File Mapping of the time point FileVersion that Yan arrives again;

(2) obtain after File Mapping, each ID is wherein obtained to corresponding data block, then by these data block composing documents;

(3) by the file forming, its setup of attribute is the file attribute that step (1) obtains, the multiple file success of Yan.

Images