CN106484321A - A kind of date storage method and data center - Google Patents

A kind of date storage method and data center Download PDF

Info

Publication number
CN106484321A
CN106484321A CN201610810735.XA CN201610810735A CN106484321A CN 106484321 A CN106484321 A CN 106484321A CN 201610810735 A CN201610810735 A CN 201610810735A CN 106484321 A CN106484321 A CN 106484321A
Authority
CN
China
Prior art keywords
queue
version number
data
version
storage
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
CN201610810735.XA
Other languages
Chinese (zh)
Inventor
白平昌
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.)
Huawei Digital Technologies Chengdu Co Ltd
Original Assignee
Huawei Symantec 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 Huawei Symantec Technologies Co Ltd filed Critical Huawei Symantec Technologies Co Ltd
Priority to CN201610810735.XA priority Critical patent/CN106484321A/en
Publication of CN106484321A publication Critical patent/CN106484321A/en
Priority to PCT/CN2017/081340 priority patent/WO2018045758A1/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
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/06Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601Interfaces specially adapted for storage systems
    • G06F3/0602Interfaces specially adapted for storage systems specifically adapted to achieve a particular effect
    • G06F3/0604Improving or facilitating administration, e.g. storage management
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/06Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601Interfaces specially adapted for storage systems
    • G06F3/0628Interfaces specially adapted for storage systems making use of a particular technique
    • G06F3/0655Vertical data movement, i.e. input-output transfer; data movement between one or more hosts and one or more storage devices
    • G06F3/0659Command handling arrangements, e.g. command buffers, queues, command scheduling
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/06Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601Interfaces specially adapted for storage systems
    • G06F3/0668Interfaces specially adapted for storage systems adopting a particular infrastructure
    • G06F3/067Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS]

Abstract

The embodiment of the invention discloses a kind of date storage method and data center, wherein, this date storage method includes:Data center reads the second edition historical queue that in local terminal, the already present version number of this object is constituted;Whether the version number judging the data of this object of this storage in first version historical queue is more than the version number in second edition historical queue;If more than the version number in second edition historical queue, data center's determination does not produce optimistic lock conflict, the data storage success of this object of this storage.It can be seen that, when the version number of the data of this object that this inventive embodiments can store at this is more than this object already present version number, the data storage success of this object that this is stored, and needs compared with each data center reaches the wheel method that stored of resolution, to greatly reduce network overhead.

Description

A kind of date storage method and data center
Technical field
The present invention relates to communication technical field is and in particular to a kind of date storage method and data center.
Background technology
At present, cloud storage is one with data storage and manages cloud computing system as core, user can be any When, any place, access data is connected in cloud by web-enabled terminal.Wherein, object storage is the core of cloud storage, Also referred to as object-based storage, the maximum feature of object storage system is object oriented is a domain name addresses, once this domain Name is arranged to open, and everyone can access this object.Ask to solve disaster tolerance and availability of object storage system etc. Topic, can adopt the scheme of multiple data centers.
In the object storage system of multiple data centers, each data center can process the service request of client, that is, Object can be uploaded to any one data center it is also possible to download to same object from any one data center.Wherein, right As uploading to after any one data center it is ensured that the data of this object of the final storage of each data center is same number According to could make, when any one data center downloads this object, identical data can be downloaded to.Therefore, when object stores Have in system upload object data when, can start one wheel resolution process, if being accepted by most of data centers, in each data The heart can preserve the data of this object of this upload, thus so that when any one data center downloads this object, can download To identical data.Needing each data center to carry out multiple network interaction however, reaching a wheel resolution, leading to network overhead Greatly.
Content of the invention
Embodiments provide a kind of date storage method and data center, network overhead can be reduced.
Embodiment of the present invention first aspect provides a kind of date storage method, is applied to the object including multiple data centers Storage system, including:During the data of data center's storage object, read second of object already present version number composition in local terminal Vncsion history queue;Data center judges whether the version number of the data of object of this storage in first version historical queue is big Version number in second edition historical queue;If more than the version number in second edition historical queue, data center determines Do not produce optimistic lock conflict, the data storage success of the object of this storage.
Wherein, when version number includes the reception of storage request receiving the data for storage object of client transmission Between and general unique identifier UUID;Posterior version number was more than reception time version number formerly, for reception the reception time Time identical version number, the big version number of UUID is more than the little version number of UUID.The different storage request corresponding objects of object Different pieces of information, the different version numbers of the data correspondence of different object, therefore, it can distinguish each storage by version number Ask the data of corresponding object, and it is possible to determine that the different numbers of corresponding object are asked in different storage according to this version number According to the new and old relationship.Wherein, UUID is general unique identifier (Universally Unique Identifier, UUID), should UUID in a distributed system, is unique in each data center, by the current date, the time, clock sequence and Globally unique IEEE machine recognition number is constituted.Wherein, first part of UUID is relevant with the time, for example, same data After being centrally generated a UUID, generate a UUID within excessively several seconds again, then first partly different, remaining is identical.Even if different Data center's two of data storage requests receiving this object times identical, but general unique identifier is different, because This, the reception time in version number can be for receiving the local zone time of the data center that storage is asked it is not necessary to be the full time The accurate time service that the quasi synchronous atomic clock of essence determines.This embodiment adopts the full time due to not being strict with each data center The quasi synchronous atomic clock of essence ensures time consistency, thus reducing the difficulty of realizing of date storage method, saves data storage Cost.
It can be seen that, the version number of the data of this object that this inventive embodiments can store at this exists more than this object Version number when, the data storage success of this object that this is stored, and need to reach the method that a wheel resolution is stored Compare, greatly reduce network overhead.In addition, each storage request corresponds to a version number, this version number can be used for difference not Ask the new and old relationship of the data of corresponding object with storage, in object storage system, each data center is in being somebody's turn to do that this stores The version number of the data of object could store successfully so that each data center is each when being more than the already present version number of this object The latest data of this object that the data of this object of storage receives for object storage system is it is ensured that in each data center this is right The data of elephant consistent.
In the first possible implementation of first aspect, data center reads the already present version of object in local terminal Number constitute second edition historical queue before, data center can receive other data centers transmission duplicate requests;From multiple The data of extracting object and the first version historical queue of object in system request;Response duplicate requests, the data of storage object. In the minds of the data of the object that other data centers can be received by this embodiment is also synchronized in notebook data.
In conjunction with the implementation of first aspect, in the possible implementation of the second of first aspect, data center reads Before taking the second edition historical queue that the already present version number of object in local terminal is constituted, receive the storage request of object;Data Center is directed to the version number that storage request generates the data of this object to be stored;Version number is added to object by data center Vncsion history queue tail of the queue, constitute object first version historical queue;Data center's response storage request, storage object Data, and send duplicate requests to other data centers, duplicate requests carry the first version historical queue of object.This enforcement The data storage of the object in the storage request that notebook data center not only can be received by mode, can also be to other numbers Send duplicate requests according to center so that other data can also synchronously store the data of this object according to duplicate requests.
In conjunction with the implementation of first aspect, or the possible implementation of the first or second of first aspect, In the third possible implementation of first aspect, data center's this storage in judging first version historical queue The version number that the version number of the data of object is not more than in second edition historical queue is it may be determined that produce optimistic lock conflict, originally The data storage failure of the object of secondary storage;Then, data center can compare first version historical queue and go through with the second edition Version number in history queue, merges and arranges out this historical queue of the third edition;Storage this historical queue of the third edition is to update local terminal pair The vncsion history queue of elephant.Wherein, can optimistic lock conflict in data center, determines that the data of the object of this storage is Stale data, but in order to ensure that notebook data center upgrades in time the vncsion history queue of this object, still can be by first version The second edition historical queue that the already present version number of this object of historical queue and local terminal is constituted merges, and obtains this history of the third edition Queue, to update the vncsion history queue of this object in local terminal.
Correspondingly, either trigger data central store data or duplicate requests trigger data central store are asked in storage Data, if data center determines does not produce optimistic lock conflict then it represents that first version historical queue is current up-to-date version Historical queue, data center can store this first version historical queue as the vncsion history queue of this object, in order under Use during the data of secondary this object of renewal.
In conjunction with the third possible implementation of first aspect, in the 4th kind of possible implementation of first aspect In, data center compares the version number in first version historical queue and second edition historical queue, merges and arranges out the third edition This historical queue, can be specially:Data center determines first version historical queue and the second edition from tail of the queue to correct order Whether there is first identical version number in this historical queue;If there is first identical version number, data center will First version historical queue is carried out with all version numbers of first identical version number to tail of the queue in second edition historical queue Sequence, merges this historical queue of the third edition;If there is not first identical version number, first version is gone through by data center History queue is ranked up with all version numbers in second edition historical queue, merges this historical queue of the third edition.
Embodiment of the present invention second aspect provides a kind of data center, and this data center has and realizes above-mentioned first aspect The behavioral function of the method providing, described function can be realized by hardware it is also possible to execute corresponding software in fact by hardware Existing.Described hardware or software include one or more modules corresponding with above-mentioned functions.
Brief description
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 institute in embodiment Need use accompanying drawing be briefly described it should be apparent that, drawings in the following description be only the present invention some enforcement Example, for those of ordinary skill in the art, on the premise of not paying creative work, can also obtain according to these accompanying drawings Obtain other accompanying drawings.
Fig. 1 is a kind of schematic diagram of object storage system disclosed in the embodiment of the present invention;
Fig. 2 is a kind of schematic flow sheet of date storage method disclosed in the embodiment of the present invention;
Fig. 3 is a kind of structural representation of data center disclosed in the embodiment of the present invention;
Fig. 4 is another structural representation of data center disclosed in the embodiment of the present invention.
Specific embodiment
For the ease of understanding the embodiment of the present invention, first the application scenarios of the embodiment of the present invention are described below.Please join Read Fig. 1, Fig. 1 is a kind of schematic diagram of object storage system disclosed in the embodiment of the present invention.As shown in figure 1, the embodiment of the present invention In, object storage system can include multiple data centers, the embodiment of the present invention taking Liang Ge data center DC1, DC2 as a example, often Individual data center can include load balancing module (Load Balancing, LB), object storage subsystem and KV (Key- Value) subsystem, wherein, load balancing subsystem is used for dividing the storage request of client Client A, Client B respectively Pairing each object storage subsystem, it is to avoid the load excessive of certain object storage subsystem;Object storage subsystem is responsible for Multiple version numbers of object are simultaneously ranked up by the data distribution version number of object;KV subsystem responsible is by the data storage of object Or be written in disk.As shown in figure 1, each data center can receive the storage request of client upload, no matter store Request uploads from which data center, and object storage system Zhong Ge data center can adopt number provided in an embodiment of the present invention According to storage method, store the data of this object and reduce network overhead.It is described in detail individually below.
Refer to Fig. 2, Fig. 2 is a kind of date storage method disclosed in the embodiment of the present invention, as shown in Fig. 2 this data is deposited Method for storing comprises the following steps:
When S101, the data of data center's storage object, read that in local terminal, the already present version number of this object is constituted the Two vncsion history queues;
Whether S102, data center judge in first version historical queue the version number of the data of this object of this storage More than the version number in second edition historical queue;If more than the version number in second edition historical queue, execution step S103;If the version number no more than in second edition historical queue, process ends.
S103, data center's determination do not produce optimistic lock conflict, the data storage success of this object of this storage.
In the embodiment of the present invention, the executive agent of step S101 to S103 can also be the object storage subsystem of data center Server of system or data center etc., the embodiment of the present invention does not limit.
In the embodiment of the present invention, the storage that version number includes receiving the data for storage object of client transmission please The reception time asked and general unique identifier UUID;The reception time, posterior version number was more than reception time version formerly Number, for the time of reception identical version number, the big version number of UUID is more than the little version number of UUID.Wherein, the difference of object The different pieces of information of storage request corresponding objects, the corresponding different version number of different pieces of information of object, therefore, it can by version number Distinguish the data that corresponding object is asked in each storage, and it is possible to determine that different storage requests correspond to according to this version number The different pieces of information of object the new and old relationship.Wherein, UUID is general unique identifier (Universally Unique Identifier, UUID), this UUID in a distributed system, is unique in each data center, by current day Phase, time, clock sequence and globally unique IEEE machine recognition number are constituted.Wherein, first part of UUID and the time Relevant, for example, after same data center generates a UUID, generate a UUID within excessively several seconds, then first part be not again With remaining is identical.Even if the time that different data centers receives two storage requests of the data of this object is identical, but logical Different with unique identifier, therefore, the reception time in version number can be for receiving the local of the data center that storage is asked Time is it is not necessary to the accurate time service that determines of the full time of the being quasi synchronous atomic clock of essence.This embodiment is each due to not being strict with Data center adopts the quasi synchronous atomic clock of full time essence to ensure time consistency, thus reducing the realization of date storage method Difficulty, saves the cost of data storage.
Alternatively, data center read local terminal in the already present version number of object constitute second edition historical queue it Before, can also comprise the following steps:
11) data center receives the duplicate requests of other data centers transmission;
12) the first version historical queue of the data center's data of extracting object and this object from this duplicate requests;
13) data center responds this duplicate requests, stores the data of this object.
In this embodiment, the data of storage is derived from the synchronous data of other data centers, thus ensureing each data center Data consistent.
Alternatively, the data of data center's storage object can also be asked to the storage that notebook data center sends for client The data of corresponding object, may comprise steps of:
21) data center receives the storage request of object;
22) data center is directed to the version number that this storage request generates this object;
23) this version number is added to the tail of the queue of the vncsion history queue of this object by data center, constitutes the of this object One vncsion history queue;
24) data center responds this storage request, stores the data of this object, and please to the transmission duplication of other data centers Ask.
Wherein, the data syn-chronization that this duplicate requests is used for for storage asking corresponding object is written to other data centers, This duplicate requests carries the first version historical queue of object.Wherein, version number includes the corresponding storage request of this version number Reception time and UUID;This reception time is the reception of the storage request corresponding to the data of notebook data center object to be stored Time.
In the embodiment of the present invention, data center is in the data of storage object, if not this object, the version of this object Historical queue is sky, NULL;If there is this object, the vncsion history queue of this object is not empty.For example, data The version number that center is directed to this storage request generation is V2, if vncsion history queue before this storage request for the data center is NULL, then first version historical queue is NULL->V2;If vncsion history queue before this storage request for the data center For NULL->V1, then first version historical queue is NULL->V1->V2.In addition, being located near correct in vncsion history queue Version number be less than be located near tail of the queue version number, newly-generated version number can be directly appended to vncsion history to row team Tail, constitutes first version historical queue.
In the embodiment of the present invention, for data center when receiving storage request, generation is directed to first version historical queue After the version number of the data of object to be stored, by this version number be added to the storage of this data center or already present should The tail of the queue of vncsion history queue of object and constitute;Second edition historical queue is that data center each stores or existed This object vncsion history queue.For example, when data center DC1 receives storage request, can generate please for this storage Ask the version number of the data of object to be stored, and this version number is added to the team of the vncsion history queue of this object in DC1 Tail, constitutes first version historical queue;When data center DC2 receives the duplicate requests of DC1 transmission, the first edition can be extracted This historical queue, and read the second edition historical queue that the version number of this object already present in DC2 is constituted, DC2 is according to this First version historical queue and second edition historical queue store the data that this object corresponding is asked in this storage.
In the embodiment of the present invention, if data center judges the data of the object of this storage in first version historical queue Version number be not more than version number in second edition historical queue, this date storage method can also comprise the following steps:
31) data center determines the optimistic lock conflict of generation, the data storage failure of the described object of this storage described;
32) data center compares the version number in first version historical queue and second edition historical queue, merges arrangement Go out this historical queue of the third edition;
33) data center's storage this historical queue of the third edition is to update the vncsion history queue of this object in local terminal.
In this embodiment, data center stores this this historical queue of third edition, in order to the number of this object of next update According to when, using the already present vncsion history queue in this data center as this object of this this historical queue of third edition.
For example, scene 1 is:Data center DC1 receives the storage request of object, and the version number of generation is V1;DC1 The vncsion history queue of this object having existed is NULL, then first version historical queue is NULL->V1.
For scene 1, the second edition historical queue that DC1 reads is NULL, and DC1 judges first version history in row Version number V1 is more than the NULL in second edition historical queue, and therefore, DC1 determines and do not produce optimistic lock conflict, therefore DC1 can be by The data storage success of version number V1 this object corresponding.
For example, scene 2 is:Data center DC1 receives the duplicate requests of data center DC2 transmission, and this duplication please Ask and ask the data syn-chronization of corresponding object to be written to DC1 for the storage receiving DC2, this duplicate requests carries this object First version historical queue is NULL->The vncsion history queue of this object that V1, DC1 have existed is NULL, wherein, version Number V1 is directed to, for DC2, the data genaration that corresponding object is asked in storage that it receives.
For scene 2, the second edition historical queue that DC1 reads is NULL, and DC1 judges first version history in row Version number V1 is more than the NULL in second edition historical queue, and therefore, DC1 determination does not produce optimistic lock conflict, can be by version The data storage success of number V1 this object corresponding.
For example, scene 3 is:Data center DC1 receives two storage requests of same object, for this two The version number that storage request generates is respectively V1, V2, and wherein, the reception time of version number V1 corresponding storage request is prior to version The reception time of number V2 corresponding storage request, therefore version number V1 is less than version number V2, the version of this object that DC1 has existed This historical queue is NULL.
For this scene 3, the data of DC1 elder generation storage version V1 this object corresponding, determine first version historical queue For NULL->V1, the second edition historical queue of reading is NULL, and DC1 judges that first version history is big to Lie Zhong version number V1 NULL in second edition historical queue, therefore, DC1 determination does not produce optimistic lock conflict, and version number V1 can be corresponded to by DC1 This object data storage success;Secondly, the data of DC1 storage version V2 this object corresponding, determines that first version is gone through History queue is NULL->V1->V2, the second edition historical queue of reading is NULL->V1, DC1 judge first version history pair Lie Zhong version number V2 is more than the version number V1 in second edition historical queue, and therefore, DC1 determination does not produce optimistic lock conflict, DC1 Can be by the data storage success of this object corresponding for version number V2.
For example, scene 4 is:Data center DC1 receives the storage request of object, and the version number of generation is V3, DC1 The vncsion history queue of this object having existed is NULL->V1->V2, then first version historical queue is NULL->V1-> V2->V3.
For scene 4, the second edition historical queue that DC1 reads is NULL->V1->V2, DC1 judge that first version is gone through History is more than each version number in second edition historical queue to the version number V3 in row, and therefore, DC1 determines that not producing optimism locks punching Prominent, DC1 can be by the data storage success of this object corresponding for version number V3.
For example, scene 5 is:Data center DC1 receives the duplicate requests of data center DC2 transmission, and this duplication please Ask and ask the data syn-chronization of corresponding object to be written to DC1 for the storage receiving DC2, this duplicate requests carries this object First version historical queue is NULL->V1->The vncsion history queue of this object that V2, DC1 have existed is NULL->V1, its In, the data genaration of corresponding object is asked in the storage that version number V2 is directed to its reception for DC2.
For scene 5, the second edition historical queue that DC1 reads is NULL->V1, DC1 judge first version history pair The version number V2 of the tail of the queue of row is more than the V1 of tail of the queue in second edition historical queue, and therefore, DC1 determines that not producing optimism locks punching Prominent, DC1 can be by the data storage success of this object corresponding for version number V2.
For example, scene 6 is:Data center DC1 receives the storage request of object, and the version number of generation is V1;With When, data center DC2 receives the storage request of same target, and the version number of generation is V2, has existed in DC1 and DC2 The vncsion history queue of this object is NULL, and version number V1 is more than version number V2 (that is, in data center DC1 data Although heart DC2 is simultaneously received storage request, in the version number that both generate, UUID is different, and the UUID of V1 is more than V2's UUID).
For above-mentioned scene 6, first, DC1 is in the data of the corresponding object of storage version V1, first version history team It is classified as NULL->V1, second edition historical queue is NULL, according to version number's comparison rule, judges first version historical queue Middle V1 is more than the NULL in second edition historical queue, determines and does not produce optimistic lock conflict, and DC1 is by corresponding for version number V1 object Data storage success;Secondly, DC1 receives the duplicate requests of DC2 transmission, in the data of the corresponding object of storage version V2 When, first version historical queue is NULL->V2, now second edition historical queue is NULL->(i.e. DC1 stores successfully version to V1 During the data of the corresponding object of this number V1, the vncsion history queue of this object is just the data of the corresponding object of storage version V1 When first version historical queue NULL->V1), due in first version historical queue V2 be less than second edition historical queue in V1, determine and produce optimistic lock conflict, by the data storage failure of corresponding for version number V2 object, version number V2 is corresponding for DC1 Data is stale data, and therefore, DC1 determines first version historical queue and second edition history from tail of the queue to correct order There is not identical version number in queue;First version historical queue is entered with all version numbers in second edition historical queue Row sequence, sorts to version number V2 and version number V1;Version number V1 is more than version number V2, therefore this historical queue of the third edition is NULL->V2->V1, DC1 store this this historical queue of third edition and the vncsion history queue of this object of local terminal are updated to NULL-> V2->V1.
For example, scene 7 is:Data center DC1 receives the storage request of object, and the version number of generation is V1;With When, data center DC2 receives the storage request of same target, and the version number of generation is V2, has existed in DC1 and DC2 The vncsion history queue of this object is NULL, and version number V1 is less than version number V2 (i.e. although data center's DC1 sum It is simultaneously received storage request according to center DC2, but in the version number of both generations, UUID is different, the UUID of V1 is less than V2's UUID).
For above-mentioned scene 7, first, DC1 is in the data of the corresponding object of storage version V1, first version history team It is classified as NULL->V1, second edition historical queue is NULL, and in first version historical queue, V1 is more than second edition historical queue In NULL, determine and do not produce optimistic lock conflict, DC1 is by the success of the data storage of corresponding for version number V1 object;Secondly, DC1 Receive the duplicate requests of DC2 transmission, in the data of the corresponding object of storage version V2, first version historical queue is NULL->V2, now second edition historical queue is NULL->(i.e. DC1 is storing the number of the successfully corresponding object of version number V1 to V1 According to when, the vncsion history queue of this object is just the first version historical queue during data of the corresponding object of storage version V1 NULL->V1);In first version historical queue, V2 is more than the V1 in second edition historical queue, determines and does not produce optimistic lock punching Prominent, DC1 is successful by the data storage of corresponding for version number V2 object.
It should be noted that the DC1 storage of either the DC1 data storage of storage request triggering or duplicate requests triggering Data, if DC1 determines does not produce optimistic lock conflict then it represents that first version historical queue is current up-to-date vncsion history team Row, DC1 can store this first version historical queue as the vncsion history queue of this object, in order to this object of next update Data when use.
Date storage method shown in Fig. 2, in the data of storage object, can read this object in local terminal already present The second edition historical queue that version number is constituted;Data center judges this object of this storage in first version historical queue Whether the version number of data is more than the version number in second edition historical queue;If more than the version in second edition historical queue Number, then data center's determination does not produce optimistic lock conflict, the data storage success of this object of this storage;It can be seen that, this invention When the version number of the data of this object that embodiment can store at this is more than this object already present version number, this is deposited The data storage success of this object of storage, compared with needing to reach the method that a wheel resolution is stored, greatly reduces network Expense.In addition, each stores the corresponding version number of request, this version number can be used for distinguishing the different storage corresponding objects of request Data the new and old relationship, in object storage system, the version number of the data of this object that each data center stores at this is big Could store when this object already present version number successfully so that the data of this object of each storage of each data center is Object storage system receive this object latest data it is ensured that in each data center the data of this object consistent.
The embodiment of the present invention, is described in detail to date storage method taking the object storage system in Fig. 1 as a example, permissible Comprise the following steps:
Client Client A sends storage request to data center DC1, and the number for storage object Obj is asked in this storage According to 1;
DC1 receives storage request, and in DC1, this storage is requested assignment to object storage in DC1 by load balancing module 1 System 11 is processed;
Object storage subsystem 11 is directed to the version number V1 that this storage request generates the data 1 of object Obj;
The vncsion history queue that object storage subsystem 11 reads object Obj in DC1 is NULL;
Version number V1 is added to the tail of the queue of the vncsion history queue of object Obj by object storage subsystem 11, constitutes object The first version historical queue of Obj is NULL->V1;
Object storage subsystem 11 calls the data 1 of KV subsystem 1 storage object Obj in DC1, and to data center DC2 Send duplicate requests, in this duplicate requests, carry first version historical queue NULL->The data 1 of V1 and object Obj;
When object storage subsystem 11 calls the data 1 of KV subsystem 1 storage object Obj in DC1, do not receive KV subsystem The optimistic lock conflict of 1 return, determines that the data 1 of object Obj stores successfully, and now in DC1 object Obj vncsion history team It is classified as NULL->V1;
Data center DC2 receives the duplicate requests of object storage subsystem 11 transmission, and in DC2, load balancing module 1 will This duplicate requests is distributed to object storage subsystem 21 in DC2 and is processed;
Object storage subsystem 21 extracts the data 1 of object Obj and the first version of object Obj from duplicate requests Historical queue is NULL->V1;
Object storage subsystem 21 calls KV subsystem 2 response duplicate requests, the data 1 of storage object Obj in DC2;
During the data 1 of KV subsystem 2 storage object Obj, read that in DC2, the already present version number of object Obj is constituted the Two vncsion history queues are NULL;
KV subsystem 2 judges that the version number V1 of the data 1 of object Obj in first version historical queue is more than empty second Vncsion history queue, determines and does not produce optimistic lock conflict;
Object storage subsystem 21 does not receive the optimistic lock conflict of KV subsystem 2 return, determines that the data 1 of object Obj is deposited Store up successfully, and now the vncsion history queue of object Obj is NULL- in DC2>V1;
Now, client Client B sends storage request to data center DC2, and this storage is asked for storage object The data 2 of Obj;Meanwhile, client Client A sends storage request to data center DC1, and this storage is asked right for storing Data 3 as Obj;
DC2 receives storage request, and in DC2, this storage is requested assignment to object storage in DC2 by load balancing module 2 System 21 is processed;DC1 receives storage request, and in DC1, this storage is requested assignment to object in DC1 and deposits by load balancing module 1 Storage subsystem 11 is processed;
Object storage subsystem 21 is directed to the version number V2 that this storage request generates the data 2 of object Obj;Object storage System 11 is directed to the version number V3 that this storage request generates the data 3 of object Obj;Wherein, the reception time in version number V2 with The reception time in version number V3 is identical, but the UUID in version number V2 is less than the UUID in version number V3;
The vncsion history queue that object storage subsystem 21 reads object Obj in DC2 is NULL->V1;Object stores subsystem Version number V2 is added to the tail of the queue of the vncsion history queue of object Obj by system 21, constitutes the first version historical queue of object Obj For NULL->V1->V2;The vncsion history queue that object storage subsystem 11 reads object Obj in DC1 is NULL->V1;Object Version number V3 is added to the tail of the queue of the vncsion history queue of object Obj by storage subsystem 11, constitutes the first version of object Obj Historical queue is NULL->V1->V3;
Object storage subsystem 21 calls the data 2 of KV subsystem 2 storage object Obj in DC2, and to data center DC1 Send duplicate requests, in this duplicate requests, carry first version historical queue NULL->V1->The data 2 of V2 and object Obj; Object storage subsystem 11 calls the data 3 of KV subsystem 1 storage object Obj in DC1, and sends duplication to data center DC2 Request, carries first version historical queue NULL- in this duplicate requests>V1->The data 3 of V3 and object Obj;
When object storage subsystem 21 calls the data 2 of KV subsystem 2 storage object Obj in DC2, do not receive KV subsystem The optimistic lock conflicts of 2 returns, determine that the data 2 of object Obj stores successfully, and now in DC2 object Obj vncsion history team It is classified as NULL->V1->V2;When object storage subsystem 11 calls the data 3 of KV subsystem 1 storage object Obj in DC1, do not receive The optimistic lock conflict returning to KV subsystem 1, determines that the data 3 of object Obj stores successfully, and now object Obj in DC1 Vncsion history queue is NULL->V1->V3;
Data center DC2 receives the duplicate requests of object storage subsystem 11 transmission, and in DC2, load balancing module 1 will This duplicate requests is distributed to object storage subsystem 21 in DC2 and is processed;Data center DC1 receives object storage subsystem 21 The duplicate requests sent, in DC1, this duplicate requests is distributed to object storage subsystem 11 in DC1 and is processed by load balancing module 1;
Object storage subsystem 21 extracts the data 3 of object Obj and the first version of object Obj from duplicate requests Historical queue is NULL->V1->V3;Object storage subsystem 11 extracts the data 2 of object Obj and right from duplicate requests As the first version historical queue of Obj is NULL->V1->V2;
Object storage subsystem 21 calls KV subsystem 2 response duplicate requests, the data 3 of storage object Obj in DC2;Right As storage subsystem 11 calls KV subsystem 1 response duplicate requests, the data 2 of storage object Obj in DC2;
During the data 3 of KV subsystem 2 storage object Obj, read that in DC2, the already present version number of object Obj is constituted the Two vncsion history queues are NULL->V1->V2;During the data 2 of KV subsystem 1 storage object Obj, read DC1 in object Obj The second edition historical queue that the version number existing is constituted is NULL->V1->V3;
KV subsystem 2 judges that the version number V3 of the data 3 of object Obj in first version historical queue is more than its second edition This historical queue NULL->V1->Version number in V2, determines and does not produce optimistic lock conflict;KV subsystem 1 judges first version In historical queue, the version number V2 of the data 2 of object Obj is not more than its second edition historical queue ULL->V1->V3 in V3, Determine and produce optimistic lock conflict;
Object storage subsystem 21 does not receive the optimistic lock conflict of KV subsystem 2 return, determines that the data 3 of object Obj is deposited Store up successfully, and now the vncsion history queue of object Obj is NULL- in DC2>V1->V2->V3;Object storage subsystem 11 Receive the optimistic lock conflict of KV subsystem 1 return, determine that the data 2 of object Obj stores unsuccessfully, now, object storage subsystem 11 compare first version historical queue NULL->V1->V2 and second edition historical queue NULL->V1->Version number in V3, closes List the third edition this historical queue NULL- side by side>V1->V2->V3;Object storage subsystem 11 stores this this history of third edition team Row vncsion history queue in DC1 as object Obj.
It can be seen that, in DC1 and DC2, the data of the object Obj of final storage is data 3;This object Obj's of final storage Vncsion history queue is NULL->V1->V2->V3.This date storage method need not be carried out between data center DC1 and DC2 Network is resolved, and the data that each data center just can realize stored object is consistent, reduces network overhead.In addition, this is real Apply store in version number in mode request the reception time be receive storage request data center local zone time, without It is the accurate time service realizing the quasi synchronous atomic clock of full time essence by GPS, thus, reduce enforcement difficulty and the one-tenth of scheme This.
Refer to Fig. 3, Fig. 3 is a kind of structural representation of data center, this data center disclosed in the embodiment of the present invention Can include with lower module:
Read module 310, is constituted in the data of storage object, reading the already present version number of this object in local terminal Second edition historical queue;
Judge module 320, for judging the version of the data of the described object of this storage in first version historical queue Number whether more than the version number in second edition historical queue;
Determining module 330, for judging this object of this storage in first version historical queue in judge module 320 Data version number be more than second edition historical queue in version number when, determine do not produce optimistic lock conflict, this storage Described object data storage success;
In the embodiment of the present invention, the storage that version number includes receiving the data for storage object of client transmission please The reception time asked and general unique identifier UUID;The reception time, posterior version number was more than reception time version formerly Number, for the time of reception identical version number, the big version number of UUID is more than the little version number of UUID.Wherein, the difference of object The different pieces of information of storage request corresponding objects, also corresponding different version number is asked in different storages, therefore, it can by version number Distinguish the data that corresponding object is asked in each storage, and it is possible to determine that different storage requests correspond to according to this version number The different pieces of information of object the new and old relationship.Wherein, UUID is general unique identifier (Universally Unique Identifier, UUID), this UUID in a distributed system, is unique in each data center, by current day Phase, time, clock sequence and globally unique IEEE machine recognition number are constituted.Wherein, first part of UUID and the time Relevant, for example, after same data center generates a UUID, generate a UUID within excessively several seconds, then first part be not again With remaining is identical.Even if the time that different data centers receives two storage requests of the data of this object is identical, but logical Different with unique identifier, therefore, the reception time in version number can be for receiving the local of the data center that storage is asked Time is it is not necessary to the accurate time service that determines of the full time of the being quasi synchronous atomic clock of essence.This embodiment is each due to not being strict with Data center adopts the quasi synchronous atomic clock of full time essence to ensure time consistency, thus reducing the realization of date storage method Difficulty, saves the cost of data storage.
In the embodiment of the present invention, the data center shown in Fig. 3 also includes:
First receiver module 340, for read that in local terminal, the already present version number of object is constituted in read module 310 the Before two vncsion history queues, receive the duplicate requests of other data centers transmission;
Extraction module 350, the first version history team of the data for extracting object from duplicate requests and this object Row;
First memory module 360, for responding this duplicate requests, stores the data of this object.
In the embodiment of the present invention, the data center shown in Fig. 3 also includes:
Second receiver module 370, for read that in local terminal, the already present version number of object is constituted in read module 310 the Before two vncsion history queues, receive the storage request of object;
Generation module 380, for generating the version number of the data of this described object to be stored for storage request;
Add module 390, for version number is added to the tail of the queue of the vncsion history queue of object, constitutes the of this object One vncsion history queue;
Second memory module 390a, for response storage request, the data of storage object, and sends to other data centers Duplicate requests, duplicate requests carry the first version historical queue of described object.
In the embodiment of the present invention, the first receiver module 340 and the second receiver module 370 can be same receiver module, Execute different operations.
In the embodiment of the present invention, determining module 330, it is additionally operable to judge first version historical queue in judge module 320 In this storage the data of described object version number be not more than second edition historical queue in version number when, determine produce Optimistic lock conflict, the data storage failure of the described object of this storage;
And then, the data center shown in Fig. 3 also includes:
Comparison module 390b, for comparing the version number in first version historical queue and second edition historical queue, closes List this historical queue of the third edition side by side;
3rd memory module 390c, is additionally operable to storage this historical queue of the third edition and is gone through with the version updating this object in local terminal History queue.
In the embodiment of the present invention, the first memory module 360, the second memory module 390a and the 3rd memory module 390c are permissible For same memory module, execute different operations respectively.
In the embodiment of the present invention, comparison module 390b can be included with lower unit:
Determining unit, for correct order, determining first version historical queue and second edition history team from tail of the queue Whether there is first identical version number in row;
Combining unit, for having the in determining first version historical queue and second edition historical queue in determining unit During one identical version number, first version historical queue is arrived with first identical version number in second edition historical queue All version numbers of tail of the queue are ranked up, and merge this historical queue of the third edition;Combining unit, is additionally operable to determine in determining unit There is not first identical version number, by first version history team in first version historical queue and second edition historical queue Row are ranked up with all version numbers in second edition historical queue, merge this historical queue of the third edition.
It should be noted that the number of either data center's data storage of storage request triggering or duplicate requests triggering According to central store data, if data center determines does not produce optimistic lock conflict then it represents that first version historical queue is current Up-to-date vncsion history queue, data center can store this first version historical queue as the vncsion history team of this object Row, in order to this object of next update data when use.
In data center shown in Fig. 3, read module can in the data of storage object, read local terminal in this object The second edition historical queue that the version number existing is constituted;Judge module is it can be determined that this is deposited in first version historical queue Whether the version number of the data of this object of storage is more than the version number in second edition historical queue;Determining module, for sentencing Disconnected module judges that the version number of the data of this object of this storage in first version historical queue is more than second edition history During version number in queue, determine and do not produce optimistic lock conflict, the data storage success of this object of this storage;It can be seen that, should When the version number of the data of this object that inventive embodiments can store at this is more than this object already present version number, incite somebody to action this The data storage success of this object of secondary storage, compared with needing to reach the method that a wheel resolution is stored, greatly reduces Network overhead.In addition, each corresponding version number of storage request, this version number can be used for distinguishing different storage requests corresponding The new and old relationship of the data of object, the version of the data of this object that each data center stores at this in object storage system Could store during number more than this object already present version number successfully so that the number of each this object storing of each data center According to the latest data of this object receiving for object storage system, thus ensure that of the data of this object in each data center Cause.
The embodiment of the present invention and described embodiment of the method are based on same design, the technology effect that embodiment therein brings Fruit is also identical, and detailed process can refer to the description of said method embodiment.
Refer to Fig. 5, Fig. 5 is another structural representation of data center disclosed in the embodiment of the present invention, real in the present invention Apply in example, data center includes processor 410, memorizer 420 and transceiver 430.Transceiver 430 is used for and external equipment between Transceiving data.The quantity of the processor 410 in data center can be one or more.In some embodiments of the present invention, place Reason device 410, memorizer 420 and transceiver 430 can be connected by bus system or other modes.Data center can be used for executing Method shown in Fig. 2.The implication of the term being related to regard to the present embodiment and citing, may be referred to above-mentioned corresponding method and implement Example.Here is omitted.
Wherein, store program codes in memorizer 420, processor 410 is used for calling the program generation of storage in memorizer 420 Code, for executing following operation:
During the data of storage object, read the second edition history team that described in local terminal, the already present version number of object is constituted Row;
Whether the version number of the data of described object judging in first version historical queue this storage is more than described the Version number in two vncsion history queues;
If more than the version number in described second edition historical queue it is determined that not producing optimistic lock conflict, described this The data storage success of the described object of storage;
When described version number includes the reception of storage request receiving the data for storage object of client transmission Between and general unique identifier UUID;Posterior version number was more than reception time version number formerly, for reception the reception time Time identical version number, the big version number of UUID is more than the little version number of UUID.
In the embodiment of the present invention, processor 410 is used for calling the program code of storage in memorizer 420, and data center reads Before taking the second edition historical queue that the already present version number of object described in local terminal is constituted, can also carry out following operation:
The duplicate requests that other data centers send are received by transceiver 430;
The first version historical queue of the data of extracting object and described object from described duplicate requests;
Respond described duplicate requests, store the data of described object.
In the embodiment of the present invention, processor 410 is used for calling the program code of storage in memorizer 420, reads in local terminal Before the second edition historical queue of described object already present version number composition, can also carry out following operation:
Receive the storage request of object;
Generate the version number of the data of this described object to be stored for described storage request;
Described version number is added to the tail of the queue of the vncsion history queue of described object, constitutes the first version of described object Historical queue;
Response described storage request, is stored the data of described object, and is sent to other data centers by transceiver 430 Duplicate requests, described duplicate requests carry the first version historical queue of described object.
In the embodiment of the present invention, processor 410 is used for calling in memorizer 420 program code of storage, is judging the In one vncsion history queue, the version number of the data of described object of this storage is not more than in described second edition historical queue Version number, can also carry out following operation:
Determine and produce optimistic lock conflict, the data storage failure of the described object of this storage described;
Version number relatively in described first version historical queue and described second edition historical queue, merges and arranges out the Three vncsion history queues;
Store described this historical queue of the third edition to update the vncsion history queue of object described in local terminal.
In the embodiment of the present invention, processor 410 is used for calling in memorizer 420 program code of storage, and relatively more described the Version number in one vncsion history queue and described second edition historical queue, merges and arranges out this historical queue of the third edition, can To be specially:
From tail of the queue to correct order, determining in described first version historical queue and described second edition historical queue is First identical version number of no presence;
If there is first identical version number, by described first version historical queue and described second edition history team Described in row, all version numbers of first identical version number to tail of the queue are ranked up, and merge this historical queue of the third edition;
If there is not first identical version number, by described first version historical queue and described second edition history All version numbers in queue are ranked up, and merge this historical queue of the third edition.
Above date storage method a kind of disclosed in the embodiment of the present invention and data center are described in detail, herein In apply specific case the principle of the present invention and embodiment be set forth, the explanation of above example is only intended to help Assistant's solution method of the present invention and its core concept;Simultaneously for one of ordinary skill in the art, according to the think of of the present invention Think, all will change in specific embodiments and applications, in sum, it is right that this specification content should not be construed as The restriction of the present invention.

Claims (10)

1. a kind of date storage method, is applied to the object storage system including multiple data centers it is characterised in that including:
During the data of data center's storage object, read the second edition that described in local terminal, the already present version number of object is constituted and go through History queue;
Described data center judges whether the version number of the data of described object of this storage in first version historical queue is big Version number in described second edition historical queue;
If more than the version number in described second edition historical queue, the determination of described data center does not produce optimistic lock conflict, The data storage success of the described object of this storage described;
Described version number include receiving the data for storage object of client transmission the reception time of storage request and General unique identifier UUID;The reception time, posterior version number was more than reception time version number formerly, for the time of reception Identical version number, the big version number of UUID is more than the little version number of UUID.
2. method according to claim 1 is it is characterised in that object described in described data center reading local terminal exists Version number constitute second edition historical queue before, methods described also includes:
Described data center receives the duplicate requests of other data centers transmission;
The first version historical queue of the data of extracting object and described object from described duplicate requests;
Respond described duplicate requests, store the data of described object.
3. method according to claim 1 is it is characterised in that object described in described data center reading local terminal exists Version number constitute second edition historical queue before, methods described also includes:
Data center receives the storage request of object;
Described data center is directed to the version number that described storage request generates this data of described object to be stored;
Described version number is added to the tail of the queue of the vncsion history queue of described object by described data center, constitutes described object First version historical queue;
Described data center response described storage request, stores the data of described object, and replicates to the transmission of other data centers Request, described duplicate requests carry the first version historical queue of described object.
If 4. the method according to any one of claims 1 to 3 is it is characterised in that described data center judges the first edition In this historical queue, the version number of the data of described object of this storage is not more than the version in described second edition historical queue This number, methods described also includes:
Described data center determines the optimistic lock conflict of generation, the data storage failure of the described object of this storage described;
Version number in the more described first version historical queue of described data center and described second edition historical queue, merges Arrange out this historical queue of the third edition;
Described data center stores described this historical queue of the third edition to update the vncsion history queue of object described in local terminal.
5. method according to claim 4 is it is characterised in that described data center more described first version historical queue With the version number in described second edition historical queue, merge and arrange out this historical queue of the third edition, including:
Described data center determines described first version historical queue and described second edition history from tail of the queue to correct order Whether there is first identical version number in queue;
If there is first identical version number, described data center is by described first version historical queue and the described second edition Described in this historical queue, all version numbers of first identical version number to tail of the queue are ranked up, and merge the third edition and originally go through History queue;
If there is not first identical version number, described data center is by described first version historical queue and described second All version numbers in vncsion history queue are ranked up, and merge this historical queue of the third edition.
6. a kind of data center is it is characterised in that include:
Read module, in the data of storage object, read that the already present version number of object described in local terminal constitutes the Two vncsion history queues;
Judge module, whether the version number for judging the data of the described object of this storage in first version historical queue is big Version number in described second edition historical queue;
Determining module, for judging the number of the described object of this storage in first version historical queue in described judge module According to version number be more than the version number in the described second edition historical queue when, determine and do not produce optimistic lock conflict, described this The data storage success of the described object of storage;
Described version number include receiving the data for storage object of client transmission the reception time of storage request and General unique identifier UUID;The reception time, posterior version number was more than reception time version number formerly, for the time of reception Identical version number, the big version number of UUID is more than the little version number of UUID.
7. data center according to claim 6 is it is characterised in that also include:
First receiver module, for reading second of object already present version number composition described in local terminal in described read module Before vncsion history queue, receive the duplicate requests of other data centers transmission;
Extraction module, for the first version history team of the data of extracting object and described object from described duplicate requests Row;
First memory module, for responding described duplicate requests, stores the data of described object.
8. data center according to claim 6 is it is characterised in that also include:
Second receiver module, for reading second of object already present version number composition described in local terminal in described read module Before vncsion history queue, receive the storage request of object;
Generation module, for generating the version number of the data of this described object to be stored for described storage request;
Add module, for described version number is added to the tail of the queue of the vncsion history queue of described object, constitutes described object First version historical queue;
Second memory module, for responding described storage request, stores the data of described object, and sends to other data centers Duplicate requests, described duplicate requests carry the first version historical queue of described object.
9. the data center according to any one of claim 6 to 8 it is characterised in that
Described determining module, is additionally operable to judge in first version historical queue the described right of this storage in described judge module When the version number of the data of elephant is not more than the version number in described second edition historical queue, determines and produce optimistic lock conflict, institute State the data storage failure of the described object of this storage;
Described data center also includes:
Comparison module, for the version number in relatively described first version historical queue and described second edition historical queue, closes List this historical queue of the third edition side by side;
3rd memory module, for storing described this historical queue of the third edition to update the vncsion history team of object described in local terminal Row.
10. data center according to claim 9 is it is characterised in that described comparison module includes:Determining unit, is used for From tail of the queue to correct order, determine in described first version historical queue and described second edition historical queue with the presence or absence of the One identical version number;
Combining unit, for determining described first version historical queue and described second edition historical queue in described determining unit During first identical version number of middle presence, described in described first version historical queue and described second edition historical queue All version numbers of first identical version number to tail of the queue are ranked up, and merge this historical queue of the third edition;
Described combining unit, is additionally operable to determine that described first version historical queue and the described second edition are gone through in described determining unit There is not first identical version number, by described first version historical queue and described second edition historical queue in history queue In all version numbers be ranked up, merge this historical queue of the third edition.
CN201610810735.XA 2016-09-08 2016-09-08 A kind of date storage method and data center Pending CN106484321A (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201610810735.XA CN106484321A (en) 2016-09-08 2016-09-08 A kind of date storage method and data center
PCT/CN2017/081340 WO2018045758A1 (en) 2016-09-08 2017-04-21 Data storage method and data centre

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610810735.XA CN106484321A (en) 2016-09-08 2016-09-08 A kind of date storage method and data center

Publications (1)

Publication Number Publication Date
CN106484321A true CN106484321A (en) 2017-03-08

Family

ID=58273617

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610810735.XA Pending CN106484321A (en) 2016-09-08 2016-09-08 A kind of date storage method and data center

Country Status (2)

Country Link
CN (1) CN106484321A (en)
WO (1) WO2018045758A1 (en)

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107153912A (en) * 2017-04-11 2017-09-12 广州市食蚁兽网络技术有限公司 One kind growth intelligent data analysis system
WO2018045758A1 (en) * 2016-09-08 2018-03-15 华为技术有限公司 Data storage method and data centre
CN108769172A (en) * 2018-05-21 2018-11-06 杭州有赞科技有限公司 A kind of method of data synchronization and system
CN109032527A (en) * 2018-07-27 2018-12-18 深圳华大北斗科技有限公司 Data processing method, storage medium and computer equipment
CN109088913A (en) * 2018-06-29 2018-12-25 华为技术有限公司 The method and load-balanced server of request data
CN109634526A (en) * 2018-12-11 2019-04-16 浪潮(北京)电子信息产业有限公司 A kind of data manipulation method and relevant apparatus based on object storage
CN110597673A (en) * 2019-09-25 2019-12-20 腾讯科技(深圳)有限公司 Disaster recovery method, device and equipment of storage system and computer readable storage medium
CN111400283A (en) * 2020-03-19 2020-07-10 中国建设银行股份有限公司 Data processing method, system, electronic equipment and storage medium
CN114598559A (en) * 2021-07-22 2022-06-07 湖南亚信软件有限公司 Data processing method and device, electronic equipment and computer readable storage medium

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111901420B (en) * 2020-07-28 2023-06-16 深圳市康冠科技股份有限公司 Data synchronization method, device and system

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030101321A1 (en) * 2001-11-29 2003-05-29 Ohran Richard S. Preserving a snapshot of selected data of a mass storage system
CN102622284A (en) * 2012-02-21 2012-08-01 上海交通大学 Data asynchronous replication method directing to mass storage system
CN103220336A (en) * 2013-03-21 2013-07-24 中国科学院计算技术研究所 Vector clock achieving method and system in file synchronization
CN103561095A (en) * 2013-11-04 2014-02-05 金蝶软件(中国)有限公司 Data synchronous method and node and storage service cluster
CN104487951A (en) * 2012-05-15 2015-04-01 日本电气株式会社 Distributed data management device and distributed data operation device
CN104753987A (en) * 2013-12-26 2015-07-01 北京东方通科技股份有限公司 Distributed session management method and system

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102279857B (en) * 2010-06-11 2015-03-04 阿里巴巴集团控股有限公司 Method and system for realizing data reproduction
US20160140194A1 (en) * 2014-11-13 2016-05-19 Bigtera Limited Data migration system and data migration method thereof
CN106484321A (en) * 2016-09-08 2017-03-08 华为数字技术(成都)有限公司 A kind of date storage method and data center

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030101321A1 (en) * 2001-11-29 2003-05-29 Ohran Richard S. Preserving a snapshot of selected data of a mass storage system
CN102622284A (en) * 2012-02-21 2012-08-01 上海交通大学 Data asynchronous replication method directing to mass storage system
CN104487951A (en) * 2012-05-15 2015-04-01 日本电气株式会社 Distributed data management device and distributed data operation device
CN103220336A (en) * 2013-03-21 2013-07-24 中国科学院计算技术研究所 Vector clock achieving method and system in file synchronization
CN103561095A (en) * 2013-11-04 2014-02-05 金蝶软件(中国)有限公司 Data synchronous method and node and storage service cluster
CN104753987A (en) * 2013-12-26 2015-07-01 北京东方通科技股份有限公司 Distributed session management method and system

Cited By (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018045758A1 (en) * 2016-09-08 2018-03-15 华为技术有限公司 Data storage method and data centre
CN107153912A (en) * 2017-04-11 2017-09-12 广州市食蚁兽网络技术有限公司 One kind growth intelligent data analysis system
CN108769172A (en) * 2018-05-21 2018-11-06 杭州有赞科技有限公司 A kind of method of data synchronization and system
CN109088913B (en) * 2018-06-29 2021-05-11 华为技术有限公司 Method for requesting data and load balancing server
CN109088913A (en) * 2018-06-29 2018-12-25 华为技术有限公司 The method and load-balanced server of request data
CN109032527A (en) * 2018-07-27 2018-12-18 深圳华大北斗科技有限公司 Data processing method, storage medium and computer equipment
CN109032527B (en) * 2018-07-27 2021-07-27 深圳华大北斗科技有限公司 Data processing method, storage medium and computer device
CN109634526A (en) * 2018-12-11 2019-04-16 浪潮(北京)电子信息产业有限公司 A kind of data manipulation method and relevant apparatus based on object storage
CN109634526B (en) * 2018-12-11 2022-04-22 浪潮(北京)电子信息产业有限公司 Data operation method based on object storage and related device
CN110597673A (en) * 2019-09-25 2019-12-20 腾讯科技(深圳)有限公司 Disaster recovery method, device and equipment of storage system and computer readable storage medium
CN110597673B (en) * 2019-09-25 2021-04-30 腾讯科技(深圳)有限公司 Disaster recovery method, device and equipment of storage system and computer readable storage medium
CN111400283A (en) * 2020-03-19 2020-07-10 中国建设银行股份有限公司 Data processing method, system, electronic equipment and storage medium
CN111400283B (en) * 2020-03-19 2024-02-06 中国建设银行股份有限公司 Data processing method, system, electronic equipment and storage medium
CN114598559A (en) * 2021-07-22 2022-06-07 湖南亚信软件有限公司 Data processing method and device, electronic equipment and computer readable storage medium

Also Published As

Publication number Publication date
WO2018045758A1 (en) 2018-03-15

Similar Documents

Publication Publication Date Title
CN106484321A (en) A kind of date storage method and data center
CN111091429A (en) Electronic bill identification distribution method and device and electronic bill generation system
CN106899648B (en) Data processing method and equipment
JPH10187519A (en) Method for preventing contention of distribution system
US10282457B1 (en) Distributed transactions across multiple consensus groups
CN111258976A (en) Distributed lock implementation method, system, device and storage medium
CN109144748B (en) Server, distributed server cluster and state driving method thereof
CN103731465A (en) Distributed system and transaction treatment method thereof
CN111158949A (en) Configuration method, switching method and device of disaster recovery architecture, equipment and storage medium
CN113010549A (en) Data processing method based on remote multi-active system, related equipment and storage medium
US8996484B2 (en) Recursive lock-and-propagate operation
CN103532870A (en) Method and device for updating Web application cluster
CN113315754B (en) Intelligent linkage method, device, equipment and medium for firewall of container visit
CN105323271B (en) Cloud computing system and processing method and device thereof
CN107967265B (en) File access method, data server and file access system
CN106951443B (en) Method, equipment and system for synchronizing copies based on distributed system
CN112527901A (en) Data storage system, method, computing device and computer storage medium
CN115390939B (en) Service processing method and system
CN110659303A (en) Read-write control method and device for database nodes
CN113032477B (en) Long-distance data synchronization method and device based on GTID and computing equipment
CN113515574B (en) Data synchronization method and device
CN117478504B (en) Information transmission method, device, terminal equipment and storage medium
CN114442947B (en) Cross-domain bucket deleting method, system, terminal and storage medium
CN117472617B (en) Message processing method and device, electronic equipment and readable storage medium
CN116166300B (en) Upgrade management method and device for intellectual property system

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20170308