[go: up one dir, main page]
More Web Proxy on the site http://driver.im/

CN106648970A - File backup method and distributed file system - Google Patents

File backup method and distributed file system Download PDF

Info

Publication number
CN106648970A
CN106648970A CN201610974296.6A CN201610974296A CN106648970A CN 106648970 A CN106648970 A CN 106648970A CN 201610974296 A CN201610974296 A CN 201610974296A CN 106648970 A CN106648970 A CN 106648970A
Authority
CN
China
Prior art keywords
file
node
distributed
backup
level
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201610974296.6A
Other languages
Chinese (zh)
Inventor
倪建
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Huawei Digital Technologies Co Ltd
Original Assignee
Beijing Huawei Digital Technologies Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing Huawei Digital Technologies Co Ltd filed Critical Beijing Huawei Digital Technologies Co Ltd
Priority to CN201610974296.6A priority Critical patent/CN106648970A/en
Publication of CN106648970A publication Critical patent/CN106648970A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00Error detection; Error correction; Monitoring
    • G06F11/07Responding to the occurrence of a fault, e.g. fault tolerance
    • G06F11/14Error detection or correction of the data by redundancy in operation
    • G06F11/1402Saving, restoring, recovering or retrying
    • G06F11/1446Point-in-time backing up or restoration of persistent data
    • G06F11/1448Management of the data involved in backup or backup restore
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00Error detection; Error correction; Monitoring
    • G06F11/07Responding to the occurrence of a fault, e.g. fault tolerance
    • G06F11/14Error detection or correction of the data by redundancy in operation
    • G06F11/1402Saving, restoring, recovering or retrying
    • G06F11/1446Point-in-time backing up or restoration of persistent data
    • G06F11/1458Management of the backup or restore process
    • G06F11/1464Management of the backup or restore process for networked environments
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F21/00Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F21/60Protecting data
    • G06F21/602Providing cryptographic facilities or services
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F21/00Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F21/60Protecting data
    • G06F21/62Protecting access to data via a platform, e.g. using keys or access control rules
    • G06F21/6218Protecting access to data via a platform, e.g. using keys or access control rules to a system of files or objects, e.g. local or distributed file system or database

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • Computer Hardware Design (AREA)
  • Computer Security & Cryptography (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Bioethics (AREA)
  • Quality & Reliability (AREA)
  • Databases & Information Systems (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The invention provides a file backup method and a distributed file system. The method comprises the steps that automatic classifying detection is conducted by adopting the distributed file system arranged on business equipment to acquire a classifying mode composed of distributed nodes which are classified according to the detection efficiency from high to low, and when the distributed file system stores a backup file, the backup file is synchronously stored in the business equipment located at the first-class distributed nodes. According to the file backup method and the distributed file system, file backup is conducted by adopting the classifying mode, and therefore the reliability and the response efficiency in the file backup executing process can be guaranteed; in addition, in the whole automatic and synchronous file backup process, no redundant physical equipment is arranged, and therefore the cost of node equipment and telecom-class equipment cannot be increased.

Description

File backup method and distributed file system
Technical field
The present invention relates to wireless communication technology field, in particular, is related to a kind of file backup method and distributed text Part system.
Background technology
In existing telecom carrier-class device, during to avoid key file corruption or loss whole telecom carrier-class device is affected Use, for the storage of the node device to key file of the telecom carrier-class device, needs are accomplished to prevent destroyed also The characteristics of meeting high reliability.
At present, the key file in mainly adopting four kinds of modes to telecom carrier-class device is stored, first, to key File is being locally encrypted storage, while be controlled based on the authority of OS (Operating System, operating system), this Once kind of a mode password is cracked, substantially without safety and reliability;Second, artificially periodically key file is derived to webmaster Backed up Deng external equipment, this mode is periodically intervened by artificial, and real-time is poor, while the key text of external storage Part is also required to be managed, and reliability is unstable;Third, using APP (Application, computer in each node device Application program) voluntarily key file is backed up, for example, first in locally stored key file, resynchronize and set to redundancy It is standby, once local file lose after, APP voluntarily from redundance unit obtain backup critical file, this mode by APP voluntarily The operation for being stored and being obtained, flow process is excessively complicated, easily goes wrong, so as to cause poor reliability;Fourth, in carrier class Multiple physical equipments, such as magnetic battle array/RAID (Redundant Arrays of of redundancy are configured on each node device of equipment Independent Disks, Redundant Arrays of Inexpensive Disks) etc., although this mode improves reliability, but is the increase in whole The cost of telecom carrier-class device.
From the foregoing, there is poor reliability in terms of file backup in the telecom carrier-class device of prior art, or can for guarantee Cause the too high problem of equipment cost by property.
The content of the invention
In view of this, the present invention provides a kind of file backup method and distributed file system, it is therefore intended that solve existing Telecom carrier-class device poor reliability in terms of file backup in technology, or to ensure that reliability causes the too high problem of equipment cost.
To solve above-mentioned technical problem, following technical scheme is present invention employs:
A first aspect of the present invention provides a kind of file backup method, it is adaptable to which distributed file system, this is distributed File system is arranged on the business device in each node device of telecom carrier-class device, is comprised the following steps:
Obtain backup file to be stored;
Backup file automatic synchronization is stored to one-level distributed node, is included at the one-level distributed node local Business device start when, automatically the periphery business device to being provided with distributed file system carries out classification detection, is determined The periphery business device.
It is advance using the distributed file system being arranged on business device in file backup method disclosed by the invention The hierarchical pattern of structure, backup file automatic synchronization is stored to the distributed node being classified, it is ensured that perform file standby Reliability and response efficiency during part, and on business device do not increase other hardware, be also not result in node device with And the cost of telecom carrier-class device increases.
In the first implementation that first aspect present invention is provided, also include:
Determine that one-level distributed node is completed after storage, notify the distributed section of other ranks via classification detection determination Point, the synchronous backup file, is stored from one-level distributed node.
Based on the first implementation of above-mentioned file backup method, after the completion of the storage of one-level distributed node, then by Layer is backed up, and can further guarantee to perform the reliability during file backup.
In second implementation that first aspect present invention is provided, the classification detection includes:
When the business device starts, the node centered on the business device, to being provided with distributed file system Periphery business device detected;
Based on detection efficient, using periphery business device described in grade highest as one-level distributed node, described in remaining Periphery business device is in turn divided into low level distributed node according to the detection efficient;
Wherein, the detection efficient detects one or more in file read-write speed and memory space by network state It is determined that, the grade of the detection efficient and the periphery business device is directly proportional.
Detected based on the automatic classification in second implementation of above-mentioned file backup method, obtain according to reliability and The hierarchical pattern for carrying out file backup of response efficiency smooth evolution, it can be ensured that the reliability during file backup is performed And response efficiency.
In the third implementation that first aspect present invention is provided, after backup file to be stored is obtained, also Including:
It is determined that the rank of backup file;
If the rank of backup file is low level file, backup file is carried out into secondary storage in local business device, The file of automatic synchronization backup simultaneously is stored to one-level distributed node;
If the rank of backup file is high-level file, directly performing will back up file automatic synchronization to the distributed section of one-level The step for point is stored;
Wherein, the rank of file is backed up, and backup file corruption or is lost to the shadow produced by the safety of the business device Sound is directly proportional.
Based on the third implementation of above-mentioned file backup method, according to the difference of backup file-level, using difference Mode backed up, make the backup of different files more targeted, further increase that different files are backed up can By property and security.
In the 4th kind of implementation that first aspect present invention is provided, also include:
Periodically to local node, the one-level distributed node and the low level distributed node are detected;
When the local node occurs abnormal, the consistent one-level distributed node of the backup file and institute are judged The sum of low level distributed node is stated whether not less than threshold value;
If the sum of the consistent one-level distributed node of the backup file and the low level distributed node is not Less than threshold value, obtain from the one-level distributed node it is described backup file the file on the local node is carried out it is extensive It is multiple;
Or, if the backup file is the consistent one-level distributed node and the low level distributed node Sum is not less than threshold value, when corresponding operating is carried out to the local node, directly obtains from the one-level distributed node The backup file of the correspondence operation;
When the one-level distributed node or the low level distributed node occur abnormal, other distributed sections are screened Point is replaced there is abnormal one-level distributed node or the low level distributed node.
The 4th kind of implementation based on above-mentioned file backup method, when a certain node file is damaged or is lost, is based on The backup file at other nodes is backed up, corresponding reading or recovery operation is performed, the process need not be relied on business device APP, you can realize damage or lose file automatic recovery.
A second aspect of the present invention provides a kind of distributed file system, and the distributed file system is arranged at carrier class On business device in each node device of equipment, including:
Acquiring unit, for obtaining backup file to be stored;
One-level backup units, for backup file automatic synchronization to be stored to one-level distributed node, a fraction When including that local business device starts at cloth node, automatically the periphery business device to being provided with distributed file system enters Periphery business device determined by row classification detection.
In the first implementation that second aspect present invention is provided, also include:
Multi-level backup unit, for determining that one-level distributed node is completed after storage, notifies what is determined via classification detection The distributed node of other ranks, the synchronous backup file, is stored from one-level distributed node.
In second implementation that second aspect present invention is provided, the classification probe unit includes:
Detecting module, for the node centered on business device, the periphery business to being provided with distributed file system sets It is standby to be detected;
Diversity module, for based on detection efficient, using grade highest periphery business device as one-level distributed node, Remaining periphery business device is in turn divided into low level distributed node according to detection efficient;Wherein, detection efficient is by network-like State, one or more determinations in detection file read-write speed and memory space, the detection efficient and periphery business device etc. Level is directly proportional.
In the third implementation that second aspect present invention is provided, also include:
Rank confirmation unit, for determining the rank of backup file, if the rank of backup file is low level file, performs Local backup unit;If the rank of backup file is high-level file, one-level backup units are performed;The rank of the backup file, The impact produced by the safety of business device is directly proportional to backup file corruption or lose
Local backup unit, for backup file to be carried out into secondary storage in local business device, while automatic synchronization State backup file to be stored to one-level distributed node.
In second implementation that second aspect present invention is provided, also include:
Periodic detection unit, for periodically entering to local node, one-level distributed node and low level distributed node Row detection, the local node is local business device;When local node occurs abnormal, judging unit is performed, when a fraction When cloth node or low level distributed node occur abnormal, replacement unit is performed;
Whether judging unit, judge to back up the sum of the consistent one-level distributed node of file and low level distributed node Not less than threshold value, if the sum of the consistent one-level distributed node of backup file and low level distributed node is not less than threshold value, Perform recovery unit, or remote acquisition unit;
Recovery unit, is carried out extensive for the acquisition backup file from one-level distributed node to the file on local node It is multiple;
Remote acquisition unit, for when corresponding operating is carried out to local node, directly obtaining from one-level distributed node Take the backup file of respective operations;
Replacement unit, for screening other distributed nodes one-level distributed node or low level point that exception occur are replaced Cloth node.
The method and distributed file system of the file backup that the present invention is provided, using the distribution being arranged on business device Formula file system first carries out automatic classification detection, obtains and is constituted according to the distributed node of detection efficient divided rank from high to low Hierarchical pattern, when distributed file system carry out back up file storage when, the backup file synchronization is stored in positioned at one In business device at level distributed node;In the present invention file backup is carried out using hierarchical pattern, it can be ensured that performing Reliability and response efficiency during file backup, and whole automatic synchronization backs up the process of file, does not configure unnecessary thing Reason equipment, is also not result in that node device and the cost of telecom carrier-class device increase.
Description of the drawings
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing The accompanying drawing to be used needed for having technology description is briefly described, it should be apparent that, drawings in the following description are only this Inventive embodiment, for those of ordinary skill in the art, on the premise of not paying creative work, can be with basis The accompanying drawing of offer obtains other accompanying drawings.
Fig. 1 is the disclosed classification detection direction schematic diagram of the embodiment of the present invention;
Fig. 2 is a kind of schematic flow sheet of file backup method disclosed in the embodiment of the present invention one;
Fig. 3 is the schematic flow sheet of another kind of file backup method disclosed in the embodiment of the present invention one;
Fig. 4 is the disclosed schematic diagram of a scenario using above-mentioned file backup method of the embodiment of the present invention two;
Fig. 5 is the distribution map of distributed nodes at different levels disclosed in the embodiment of the present invention two;
Fig. 6 is a kind of structural representation of distributed file system disclosed in the embodiment of the present invention three.
Specific embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete Site preparation is described, it is clear that described embodiment is only a part of embodiment of the invention, rather than the embodiment of whole.It is based on Embodiment in the present invention, it is every other that those of ordinary skill in the art are obtained under the premise of creative work is not made Embodiment, belongs to the scope of protection of the invention.
From background technology, for the storage of key file in telecom carrier-class device, it is local that prior art is adopted Encryption storage, the mode that derivation and APP (Application, computer applied algorithm) are voluntarily backed up artificial periodically carry out file , there is poor reliability during file is stored in storage;And adopt and configured on the node device of telecom carrier-class device Multiple physical equipments are for file storage, although improve the reliability of file storage, but also increase telecom carrier-class device Cost.
Therefore, a kind of technical scheme of file backup is embodiments provided, the technical scheme is applied to carrier class Equipment, telecom carrier-class device here refers to cluster device, comprising the node device for being arranged at same city or different cities, the node Equipment is generally it can be thought that be the data center for arranging a certainly;The embodiment of the present invention is in advance in each section of the telecom carrier-class device Distributed file system is set on multiple business devices of point device, the distributed file system by Embedded multiple veneers or Multiple VM (Virtual files, virtualized file) of cloud are constituted;
When business device starts, using itself as local node, each node device based on whole telecom carrier-class device In business device, other are provided with the business device of distributed file system as the distributed node for backup, base The priority of each distributed node is automatically determined in factors such as efficiencies of transmission, each distributed node is the distal end of local node Backup node, then, is performed in the node device based on the api interface (, interface operation file) that distributed file system is provided When corresponding document creation or write, distributed file system carries out the file for creating or write accordingly according to file-level Backup operation.
Based on for whole telecom carrier-class device, classification as illustrated in FIG. 1 is detected the grade classification of above-mentioned distributed node Direction schematic diagram, when business device starts, node A, by periphery distributed field system is provided with centered on the business device The business device of system carries out classification detection, by result most as distributed node B based on network state, speed and memory space , used as the one-level distributed node of this business device, remaining is successively as two grades of distributed nodes, three fractions for good business device Cloth node etc.;The distributed node of same rank can select multiple according to demand, therefore the division of each rank can To consider;
If after classification detection, having node to exit, then can directly exit;If there is new node to add, can be automatic From the synchronous corresponding backup file of all mid-side nodes of the new node;
Under normal circumstances, being divided to one-level distributed node, or be divided to three-level distributed node can meet currently Demand of the business device to file backup;It is of course also possible to more;The embodiment of the present invention is not defined to this;
Embodiment one
As shown in Fig. 2 disclosed in the embodiment of the present invention one a kind of schematic flow sheet of file backup method;The present embodiment By on the business device of each node device of telecom carrier-class device, pre-setting distributed file system, and in each industry Business equipment start after, by itself centered on, automatically using itself as local node, with each node device of telecom carrier-class device The business device of distributed file system is provided with as distributed node, carries out backing up the division of grade;The distributed document System provide api interface, business device by api interface perform create or write file when, the distributed file system according to File-level performs corresponding backup;
In the node device, file backup method mainly comprises the steps:
S101:Distributed file system determines the current rank for creating by api interface or writing file, if high-level File, then be S102~S103 for the process that one-level distributed node is backed up;If low level file, then for one-level The process that distributed node is backed up is S104~S106;
In S101, the rank of file is created or write, damage or lose produced by the safety to business device with it Impact is directly proportional, and affects bigger, and rank is higher;High-level file includes high classified papers, high sensitive file and high reliability File;Low level file includes:General susceptibility file, general reliability file and ordinary file;
Scene alternatively, it is also possible to be determined by distributed file system is divided to the rank of file, ordinary circumstance Under, the file for creating under high safety scene or writing is high-level file, and the file for creating under common security scenario or writing is Low level file;
For the backup of high-level file:
S102:The file of the establishment of distributed file system automatic synchronization or write is backed up to one-level distributed node Storage;
The backup of high-level file needs high safety scene, therefore, in S102, directly open VW patterns, i.e., local nothing File mode;Specially:The high-level file that local node is created or write directly is backed up into one-level distributed node Interior, when making no presence of the high-level file on local node, and using the high-level file every time, then distributed file system is straight Capable remote access is tapped into, the high-level file is obtained;When local node is trespassed or is destroyed, because of local reality not There is file, therefore do not result in big loss;
Here one-level distributed node is typically chosen the business device for not opening user port;Such as existing large-scale collection Group facility, typically has two main controlled nodes, that is, Centroid, and remaining is non-main controlled node, that is, all mid-side nodes, It is general to detect distributed nodes at different levels from the non-main controlled node;The non-main controlled node is then general not to open user port, it is impossible to enter The access of row external network, may only carry out network interaction with Centroid, and security is higher;
S103:Distributed file system is backed up successfully after the completion of the backup of one-level distributed node to node device feedback Information;
In S103, once the process for creating or writing file terminates, then it is assumed that the backup on one-level distributed node Also complete, then distributed file system then generates backup success message, point out the backup of the business device remote storage node into Work(;
For the backup of low level file:
S104;Distributed file system is backed up the file for creating by api interface or write in local node;
In S104, local node refers to the business device set by the distributed file system, in the business device When APP is created or write file using the api interface that distributed file system is provided, dividing in the business device is pre-set at Cloth file system, while API is created or writes file, is backed up in local node;
That is, in this business device, outside file directory that removing itself is created or write file writes direct, should Distributed file system simultaneously writes this document in another file directory or storage catalogue of local node, and the backup is simultaneously Need not rely on network;
S105:The file of the establishment of distributed file system automatic synchronization or write is backed up to one-level distributed node Storage;
In S105, when the one-level distributed node is that this business device starts, distributed text is provided with to periphery automatically The business device of part system carries out being classified what detection was obtained;
Low level file can perform backup under common security scenario, therefore, the S105 is performed in the distributed section of one-level The process backed up on point is:The file of the establishment or write is encrypted, is then provided using distributed file system API write interface, the file of the establishment or write is write into one-level distributed node, stored;
It should be noted that S105 can be performed while local node is backed up S104 is performed, both do not have Obvious priority execution sequence;
S106:Distributed file system obtains the file of the backed up in synchronization on the one-level distributed node, and saves with local File on point is compared, if file is consistent, generates backup success message and feeds back to business device;
In order to ensure whether the backed up in synchronization on remote backup, i.e. one-level distributed node is successful, then S106 is performed, be distributed Formula file system obtains the establishment of backup or the file of write, then using the reading interface of API from the one-level distributed node Contrasted with the current file for creating or writing, if the file backed up on one-level distributed node is created or write with current File is consistent, it is determined that the backed up in synchronization success on one-level distributed node;
It should be noted that for backup mode under common safe mode, it is standby on accessed one-level distributed node Part file, can be being contrasted with the file backed up in local node;
Complete on one-level distributed node after backup for above-mentioned high-level file and low level file, based on demand, If the distributed node for also dividing other ranks is backed up, then following step is continued executing with;
S107:Distributed file system notifies the distributed node of other ranks, synchronously should from the one-level distributed node The file for creating or writing, is backed up;
In S107, distributed file system one-level distributed node backup success after, to two grades of distributed nodes Synchronization message is sent with three-level distributed node, this document is obtained by two grades of distributed nodes and three-level distributed node and is deposited Storage, completes backup;
Wherein, two grades of distributed nodes and three-level distributed node can adopt grade distributed node of async backup, i.e., two Obtain after this document stored from one-level distributed node, three-level distributed node is obtained again from two grades of distributed nodes should File is stored, and completes async backup;Can also be by the way of backed up in synchronization, i.e. two grades of distributed nodes and three fractions Cloth node obtains this document from one-level distributed node and is stored simultaneously, completes backed up in synchronization;
The embodiment of the present invention need not increase any hardware, but adopt the distributed document elder generation being arranged on business device Carry out automatic classification detection, obtain according to reliability and the hierarchical pattern for carrying out file backup of response efficiency smooth evolution, really Protect perform file backup during reliability and response efficiency, and the number of the grade can with by node device based on need Ask and selected;Then, different modes are selected to carry out file backup according to the height of file-level, e.g., to high-level file By the way of the backup of VW patterns, by the way of being backed up on local node and distributed node to low level file; And the process of whole automatic synchronization backup file is independent of the APP on node device;It follows that the embodiment of the present invention is by setting The distributed file system being placed on business device carries out the file backup of successively hierarchical pattern, not only ensures file backup Reliability, is also not result in that node device and the cost of telecom carrier-class device increase.
In embodiments of the present invention, based on it is above-mentioned high-level file and low level file are carried out in different ways it is standby After part, as shown in figure 3, the local node and each distributed node to performing backup is regularly verified, noting abnormalities When, the file based on backup is performed and read or recovery operation, and the process does not need the APP process on business device, by text Part distributed system is voluntarily performed, and detailed process comprises the steps:
S201:Distributed file system periodically detects to local node and each distributed node, when detecting this There is abnormal, execution S202 and S203 in ground node, when distributed node appearance is abnormal, performs S204;
S202:If occurring abnormal for local node, the file to backing up in each level node is compared, if being not less than Backup file on the node of threshold value is consistent, then perform S203;
In S202 is performed, the span of the threshold value is not less than the 50% of the node number for performing backup;
When backup file on to each level node is compared, can be with by the way of global verification, will be whole The backup file performed on the local node and/or distributed node of backup is compared, and judges the node with same file Whether number is not less than threshold value;
Can also be verified for local node and part distributed node using preferred;For example, in common safety Under scene, backup file that can be locally only on node and one-level distributed node is compared, and judges there is same file Node number whether be not less than threshold value;Under high safety scene, the backup file on one-level distributed node is compared, Judge whether the node number with same file is not less than threshold value;
During S202 is performed, if consistent not less than the backup file on the node of threshold value, then it is assumed that the text of backup Part is substantially errorless;
S203:For low level file, distributed file system obtains backup file from one-level distributed node to this File on ground node is recovered;For high-level file, distributed document is then obtaining standby from one-level distributed node Part file performs corresponding operating;
In S203, under common security situation, if the file of local node goes wrong, the distributed document directly from Backup file is obtained on one-level distributed node, the file on local node is recovered;
But, for the file of high safety rank, if damaging, can not be recovered, but when needs are used Wait, directly obtained from one-level distributed node by distributed file system, by such mode, Gao An can be avoided Full rank or high sensitive file, when local node is trespassed or is destroyed, are illegally accessed;
S204:When a certain rank distributed node occurs abnormal, distributed file system screens other distributed nodes Replace the abnormal distributed node of the appearance;
In S204, abnormal situation occur includes:Response speed is slack-off, or breaks down;It is general to be adjusted using dynamic The mode of grade is processed the abnormal distributed node of the appearance;
For the distributed node that response speed is slack-off, screen other distributed nodes and replace the distributed of appearance exception The mode of node has two kinds:
First, distributed file system is degraded the slack-off distributed node of the response speed, will be original rudimentary Other distributed node is upgraded;
Second, distributed file system directly rejects the slack-off distributed node of the response speed, to other low levels Distributed node carry out overall upgrading;
For the distributed node for breaking down, then directly reject, entirety is carried out to other low-level distributed nodes Upgrading;
The embodiment of the present invention, the distributed file system in business device is based on the above-mentioned local node for carrying out file backup Regularly verified with distributed node, if exception occurs in local node, local file is carried out based on aforesaid operations extensive Multiple, if exception occurs in distributed node, the grade of dynamic adjustment distributed node, the process of whole reparation automatically all need not Rely on the APP in node device;Certainly, if after file is lost, it is above-mentioned that APP can actively trigger distributed file system execution Recover the operation of file.
Carry out in the operating process of file backup, entering one based on the above-mentioned distributed file system being arranged in node device Step, distributed file system has authentication mechanism disclosed in the embodiment of the present invention, and for the APP of the business device authority is carried out The APP that certification, only certification pass through just can actively trigger the distributed file system and perform corresponding file backup correlation Operation;
If distributed file system is by purview certification, confirmation currently login the user of the business device be advanced level user or During keeper, then file backup method disclosed in the embodiments of the present invention can be performed, and to each distributed node Management;If the user for confirming currently to login the business device is domestic consumer, it is common that the distributed file system limits this User's establishment file, when reading file or written document, can carry out the rank of the distributed node of file backup;And, restriction is held Deletion of the row to local file, e.g., only the user of establishment file can delete this document.
Embodiment two
Based on a kind of file backup method disclosed in the embodiments of the present invention one, in the present embodiment concrete application field is based on Scape is described in detail;
As shown in figure 4, the telecom carrier-class device is made up of 2 node devices, 1 node device is arranged on Beijing and (uses in Fig. 4 " Beijing DC/ machine frames 1 " represents), 1 node device is arranged on Hangzhou (being represented with " Hangzhou DC/ machine frames 1 " in Fig. 4);
DC, refers generally to the data center of different geographical, such as Beijing DC, and machine frame 1 then refers to the logic machine frame 1 in a DC; It is also assumed that DC is globally addressable;
As shown in figure 4, in the D1 node devices, partial service equipment is provided with by multiple embedded on (11,12,13) The distributed file system that the VW of veneer or cloud is constituted, remaining is common business device 14;Likewise, setting in the D2 nodes In standby, partial service equipment is provided with the distributed text being made up of the VW of multiple Embedded veneers or cloud on (21,22,23) Part system, remaining is the business device 24 of part;
It is assumed that as shown in figure 5, with business device 11 as local node, when business device 11 starts, in business device 11 Distributed file system, classification detection periphery is provided with the business device of distributed file system, according to detection file read-write The size of speed, the most fast business device 12 of read or write speed and business device 13 is distributed as the one-level of this business device 11 Node, business device 21 and business device 22 are as two grades of distributed nodes, business device 23 as three-level distributed node;
In addition, the distributed file system locally divides the sheet that another storage region is as 0.5 grade in business device 11 The backup of ground two-region;
APP in business device 11 is created using the api interface that distributed file system is provided, read/write, and is deleted Except file;
Based on the hierarchical pattern shown in above-mentioned Fig. 5, when the APP in business device 11 uses distributed file system offer When api interface writes low level file in the business device 11, file backup disclosed in the embodiments of the present invention one is performed Method, detailed process is:
S301, the low level file that the distributed file system on business device 11 will be created or write by api interface Backed up in another storage catalogue of the business device 11;
S302:Distributed file system automatic synchronization on the business device 11 low level file is to business device 12 and industry Business equipment 13 is backing up and storing;
S303:Distributed file system on business device 11 obtains same on the business device 12 and/or business device 13 The file of step backup, and be compared with the low level file on business device 11, if file is consistent, generation is backed up and successfully disappeared Breath feeds back to business device 11;
In S303, at least obtain the low level file backed up on an one-level distributed node and be compared;
S304:Distributed file system on business device 11 notifies business device 21, business device 22 and business device 23, the synchronous low level file is backed up from the business device 12;
S305:Distributed file system on business device 11 is periodically to business device 11, business device 12, business device 13rd, business device 21, business device 22 and business device 23 are detected exception occur when business device 11 is detected;
S306:To backing up on business device 12, business device 13, business device 21, business device 22 and business device 23 File compare, if the backup file on the business device of more than 3 is consistent, the distributed document on business device 11 System obtains backup file from business device 12 or business device 13 and the file on business device 11 is recovered;
This embodiment of the invention is based on specific application scenarios, using the distributed document elder generation being arranged on business device 11 Automatic classification detection is carried out, is obtained according to reliability and the hierarchical pattern for carrying out file backup of response efficiency smooth evolution, energy Enough guarantee the reliability and response efficiency during file backup is performed, and on business device do not increase other hardware, because This also assures that the cost for being not result in node device and telecom carrier-class device increases on the basis of the reliability for guaranteeing file backup Plus.
Embodiment three
Based on a kind of file backup method, the embodiment of the present invention three disclosed in the embodiments of the present invention one and embodiment two Also correspond to and disclose the distributed file system for performing above-mentioned file backup method;The distributed file system can be using embedded Multiple veneers of formula or the VW of multiple clouds are constituted, and the distributed file system 100 is as shown in fig. 6, mainly include:Acquiring unit 101st, one-level backup units 102 and multi-level backup unit 103;
Acquiring unit 101, for obtaining backup file to be stored;
One-level backup units 102, for backup file automatic synchronization to be stored to one-level distributed node;
Multi-level backup unit 103, for determining that one-level distributed node is completed after storage, notifies to determine via classification detection Other ranks distributed node, the synchronous backup file, is stored from one-level distributed node.
It is automatically right when including that local business device starts at the one-level distributed node in one-level backup units 102 Being provided with the periphery business device of the distributed file system carries out classification detection, determined by the periphery business device;
The classification probe unit includes:
Detecting module, for the node centered on business device, the periphery business to being provided with distributed file system sets It is standby to be detected;
Diversity module, for based on detection efficient, using grade highest business device as one-level distributed node, remaining Business device is in turn divided into low level distributed node according to detection efficient;
Disclosed in the embodiment of the present invention in the distributed file system, can with according to the file that creates or write Rank, or current high safety scene and common security scenario determined by distributed file system, perform different standby Part unit, specifically includes:
Rank confirmation unit, for determining the rank of backup file, if the rank of backup file is low level file, performs Local backup unit;If the rank of backup file is high-level file, one-level backup units 102 are performed;
Local backup unit, for the backup file to be carried out into secondary storage in local business device, while automatically Backed up in synchronization file to one-level distributed node is stored.
Disclosed in the embodiment of the present invention in the distributed file system, also include performing each node periodic detection and The unit of abnormal nodes is processed, such as:
Periodic detection unit, for periodically entering to local node, one-level distributed node and low level distributed node Row detection, local node is local business device;When local node occurs abnormal, judging unit is performed, when one-level distribution When formula node or the low level distributed node occur abnormal, replacement unit is performed;
Whether judging unit, judge to back up the sum of the consistent one-level distributed node of file and low level distributed node Not less than threshold value, if the sum of the consistent one-level distributed node of backup file and low level distributed node is not less than threshold value, Then according to current scene demand, recovery unit is performed under common security scenario, long-range acquisition is performed under high safety scene Unit;
Recovery unit, is carried out extensive for the acquisition backup file from one-level distributed node to the file on local node It is multiple;
Remote acquisition unit, for when corresponding operating is carried out to local node, directly obtaining from one-level distributed node Take the backup file of respective operations;
, there is abnormal one-level distributed node or described rudimentary for screening other distributed nodes and replacing in replacement unit Other distributed node.
The distributed file system disclosed in the embodiment of the present invention, is also included for creating, reading and writing and deleting backup text The user of part carries out the purview certification unit of purview certification.
Concrete operations involved by the disclosed modules and unit for performing file backup of the above embodiment of the present invention, The appropriate section in file backup method disclosed in the embodiments of the present invention is may refer to, is also no longer repeated here.
In sum, the embodiment of the present invention need not increase any hardware, but adopt and be arranged at dividing on business device Cloth file first carries out automatic classification detection, obtain according to reliability and response efficiency smooth evolution carry out file backup point Level pattern, it is ensured that the reliability and response efficiency during file backup is performed, and the number of the grade can be with by node Equipment is selected based on demand;Then, different modes are selected to carry out file backup according to the height of file-level, and it is whole The process of automatic synchronization backup file is independent of the APP on node device;It follows that the embodiment of the present invention is by being arranged at industry Distributed file system on business equipment carries out the file backup of successively hierarchical pattern, not only ensures the reliability of file backup Property, also it is not result in that node device and the cost of telecom carrier-class device increase.
The preferred embodiments of the present invention are the foregoing is only, the present invention is not limited to, for the skill of this area For art personnel, the present invention can have various modifications and variations.It is all within the spirit and principles in the present invention, made any repair Change, equivalent, improvement etc., should be included within the scope of the present invention.

Claims (10)

1. a kind of file backup method, it is characterised in that suitable for distributed file system, the distributed file system is arranged On business device in each node device of telecom carrier-class device, including:
Obtain backup file to be stored;
The backup file automatic synchronization is stored to one-level distributed node, at the one-level distributed node this is included When the business device on ground starts, automatically the periphery business device to being provided with the distributed file system carries out classification detection, Determined by the periphery business device.
2. method according to claim 1, it is characterised in that also include:
Determine that the one-level distributed node is completed after storage, notify the distribution of other ranks via the classification detection determination Formula node, the synchronous backup file, is stored from the one-level distributed node.
3. method according to claim 1 and 2, it is characterised in that the classification detection includes:
When the business device starts, the node centered on the business device, the week to being provided with distributed file system Side business device is detected;
Based on detection efficient, using periphery business device described in grade highest as one-level distributed node, remaining described periphery Business device is in turn divided into low level distributed node according to the detection efficient;
Wherein, the detection efficient detects one or more determinations in file read-write speed and memory space by network state, The grade of the detection efficient and the periphery business device is directly proportional.
4. method according to claim 1, it is characterised in that after obtaining backup file to be stored, also include:
Determine the rank of the backup file;
If the rank of the backup file is low level file, the backup file is carried out into secondary depositing in local business device Storage, while file is backed up described in automatic synchronization being stored to one-level distributed node;
If the rank of the backup file is high-level file, directly performs and be distributed the backup file automatic synchronization to one-level The step for formula node is stored;
Wherein, the rank of the backup file, is produced with the backup file corruption or loss to the safety of the business device Raw impact is directly proportional.
5. method according to claim 1, it is characterised in that also include:
Periodically to local node, the one-level distributed node and the low level distributed node are detected;
When the local node occurs abnormal, judge the consistent one-level distributed node of file and described low of backing up Whether the sum of rank distributed node is not less than threshold value;
If the sum of the consistent one-level distributed node of the backup file and the low level distributed node is not less than Threshold value, the backup file is obtained from the one-level distributed node file on the local node is recovered;
Or, if the backup file is the sum of the consistent one-level distributed node and the low level distributed node Not less than threshold value, when corresponding operating is carried out to the local node, correspondence is directly obtained from the one-level distributed node The backup file of the operation;
When the one-level distributed node or the low level distributed node occur abnormal, screen other distributed nodes and replace Swap out existing abnormal one-level distributed node or the low level distributed node.
6. a kind of distributed file system, it is characterised in that the distributed file system is arranged at each of telecom carrier-class device On business device in node device, including:
Acquiring unit, for obtaining backup file to be stored;
One-level backup units, for the backup file automatic synchronization to be stored to one-level distributed node, the one-level When including that local business device starts at distributed node, automatically to being provided with the periphery business of the distributed file system Equipment carries out classification detection, determined by the periphery business device.
7. distributed file system according to claim 6, it is characterised in that also include:
Multi-level backup unit, for determining that the one-level distributed node is completed after storage, notifies true via the classification detection The distributed node of other fixed ranks, the synchronous backup file, is stored from the one-level distributed node.
8. the distributed file system according to claim 6 or 7, it is characterised in that the classification probe unit includes:
Detecting module, for the node centered on the business device, the periphery business to being provided with distributed file system sets It is standby to be detected;
Diversity module, for based on detection efficient, using periphery business device described in grade highest as one-level distributed node, Remaining described periphery business device is in turn divided into low level distributed node according to the detection efficient;Wherein, the detection Efficiency detects one or more determinations in file read-write speed and memory space, the detection efficient and institute by network state The grade for stating periphery business device is directly proportional.
9. distributed file system according to claim 6, it is characterised in that also include:
Rank confirmation unit, for determining the rank of the backup file, if the rank of the backup file is low level file, Perform local backup unit;If the rank of the backup file is high-level file, the one-level backup units are performed;It is described standby The rank of part file, is directly proportional to the impact produced by the backup file corruption or the safety for losing to the business device
The local backup unit, for the backup file to be carried out into secondary storage in local business device, while automatically The synchronous backup file is stored to one-level distributed node.
10. distributed file system according to claim 6, it is characterised in that also include:
Periodic detection unit, for periodically to the local node, the one-level distributed node and low level distribution Formula node is detected that the local node is local business device;When the local node occurs abnormal, judgement is performed Unit, when the one-level distributed node or the low level distributed node occur abnormal, performs replacement unit;
The judging unit, judges the consistent one-level distributed node of the backup file and the distributed section of the low level Whether the sum of point is not less than threshold value, if the consistent one-level distributed node of the backup file and low level distribution The sum of formula node is not less than threshold value, performs recovery unit, or remote acquisition unit;
The recovery unit, for the backup file to be obtained from the one-level distributed node on the local node File is recovered;
The remote acquisition unit, it is directly distributed from the one-level for when corresponding operating is carried out to the local node The backup file of the correspondence operation is obtained on node;
, there is abnormal one-level distributed node or described rudimentary for screening other distributed nodes and replacing in the replacement unit Other distributed node.
CN201610974296.6A 2016-11-04 2016-11-04 File backup method and distributed file system Pending CN106648970A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610974296.6A CN106648970A (en) 2016-11-04 2016-11-04 File backup method and distributed file system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610974296.6A CN106648970A (en) 2016-11-04 2016-11-04 File backup method and distributed file system

Publications (1)

Publication Number Publication Date
CN106648970A true CN106648970A (en) 2017-05-10

Family

ID=58820847

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610974296.6A Pending CN106648970A (en) 2016-11-04 2016-11-04 File backup method and distributed file system

Country Status (1)

Country Link
CN (1) CN106648970A (en)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108415794A (en) * 2018-01-30 2018-08-17 河南职业技术学院 File backup method and file backup device
CN110493027A (en) * 2018-09-27 2019-11-22 北京白山耘科技有限公司 A kind of distributed and management method
WO2020063424A1 (en) * 2018-09-27 2020-04-02 北京白山耘科技有限公司 Distributed data system, distributed data synchronization method, computer storage medium, and computer device
CN111083199A (en) * 2019-11-23 2020-04-28 上海畅星软件有限公司 High-concurrency, high-availability and service-extensible platform-based processing architecture
CN111866178A (en) * 2020-08-04 2020-10-30 蝉鸣科技(西安)有限公司 Distributed FTP/FTPS file transmission method and device and computer storage medium
CN112256490A (en) * 2020-11-17 2021-01-22 珠海大横琴科技发展有限公司 Data processing method and device

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102025756A (en) * 2009-09-09 2011-04-20 中兴通讯股份有限公司 Distributed system and data synchronization method thereof
CN104142871A (en) * 2013-05-10 2014-11-12 中国电信股份有限公司 Data backup method and device and distributed file system
CN104410683A (en) * 2014-11-21 2015-03-11 四川神琥科技有限公司 Data backup method
CN105120003A (en) * 2015-09-10 2015-12-02 浪潮(北京)电子信息产业有限公司 Method for realizing data backup in cloud environment

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102025756A (en) * 2009-09-09 2011-04-20 中兴通讯股份有限公司 Distributed system and data synchronization method thereof
CN104142871A (en) * 2013-05-10 2014-11-12 中国电信股份有限公司 Data backup method and device and distributed file system
CN104410683A (en) * 2014-11-21 2015-03-11 四川神琥科技有限公司 Data backup method
CN105120003A (en) * 2015-09-10 2015-12-02 浪潮(北京)电子信息产业有限公司 Method for realizing data backup in cloud environment

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
MICROSOFT著: "《操作系统的安装、配置和管理 Windows XP专业版的安装、配置和管理》", 31 July 2003 *

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108415794A (en) * 2018-01-30 2018-08-17 河南职业技术学院 File backup method and file backup device
CN110493027A (en) * 2018-09-27 2019-11-22 北京白山耘科技有限公司 A kind of distributed and management method
WO2020063424A1 (en) * 2018-09-27 2020-04-02 北京白山耘科技有限公司 Distributed data system, distributed data synchronization method, computer storage medium, and computer device
WO2020063650A1 (en) * 2018-09-27 2020-04-02 北京白山耘科技有限公司 Distributed data management system and management method, computer storage medium and computer device
CN111083199A (en) * 2019-11-23 2020-04-28 上海畅星软件有限公司 High-concurrency, high-availability and service-extensible platform-based processing architecture
CN111866178A (en) * 2020-08-04 2020-10-30 蝉鸣科技(西安)有限公司 Distributed FTP/FTPS file transmission method and device and computer storage medium
CN112256490A (en) * 2020-11-17 2021-01-22 珠海大横琴科技发展有限公司 Data processing method and device

Similar Documents

Publication Publication Date Title
CN106648970A (en) File backup method and distributed file system
CN111031096B (en) Distributed storage system construction method based on mimicry defense
CN101334825B (en) Application program management and operation system and method
CN103853634B (en) Disaster recovery system and disaster recovery method
CN103019614B (en) Distributed memory system management devices and method
CN103064763B (en) A kind of data back up method and relevant device, system
CN104166606A (en) File backup method and main storage device
CN102317910A (en) Methods, devices and system for virtual data backup and reintegration
CN100449501C (en) Computer network datarecovering method
CN110784331B (en) Consensus process recovery method and related nodes
CN106708653B (en) Mixed tax big data security protection method based on erasure code and multiple copies
CN105426427A (en) MPP database cluster replica realization method based on RAID 0 storage
CN103049400A (en) Disk reconfiguration method and disk reconfiguration device
CN110807064A (en) Data recovery device in RAC distributed database cluster system
CN104794028A (en) Disaster tolerance processing method and device, main data center and backup data center
CN102170460A (en) Cluster storage system and data storage method thereof
CN105404565B (en) A kind of dual-active data guard method and device
CN105022678A (en) Data backup method and apparatus for virtual machine
KR100922584B1 (en) Distributed object-sharing system and method thereof
CN101937378B (en) Method for carrying out back-up protection on data of storage equipment and computer system
CN103092765A (en) Solid state storage system, solid state storage device and data writing method
CN113051428B (en) Method and device for back-up storage at front end of camera
CN102520880B (en) System raid management module design method
CN104407806A (en) Method and device for revising hard disk information of redundant array group of independent disk (RAID)
CN105703941A (en) Method and device for processing configuration transaction

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
WD01 Invention patent application deemed withdrawn after publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20170510