CN108241469A - A kind of date storage method, apparatus and system - Google Patents

A kind of date storage method, apparatus and system Download PDF

Info

Publication number
CN108241469A
CN108241469A CN201611224468.4A CN201611224468A CN108241469A CN 108241469 A CN108241469 A CN 108241469A CN 201611224468 A CN201611224468 A CN 201611224468A CN 108241469 A CN108241469 A CN 108241469A
Authority
CN
China
Prior art keywords
node
storage management
storage
target
cluster
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.)
Granted
Application number
CN201611224468.4A
Other languages
Chinese (zh)
Other versions
CN108241469B (en
Inventor
赵龙
金辉
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 Kingsoft Cloud Network Technology Co Ltd
Beijing Kingsoft Cloud Technology Co Ltd
Original Assignee
Beijing Kingsoft Cloud Network Technology Co Ltd
Beijing Kingsoft Cloud Technology 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 Kingsoft Cloud Network Technology Co Ltd, Beijing Kingsoft Cloud Technology Co Ltd filed Critical Beijing Kingsoft Cloud Network Technology Co Ltd
Priority to CN201611224468.4A priority Critical patent/CN108241469B/en
Publication of CN108241469A publication Critical patent/CN108241469A/en
Application granted granted Critical
Publication of CN108241469B publication Critical patent/CN108241469B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input 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
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601—Interfaces specially adapted for storage systems
    • G06F3/0602—Interfaces specially adapted for storage systems specifically adapted to achieve a particular effect
    • G06F3/061—Improving I/O performance
    • G06F3/0611—Improving I/O performance in relation to response time
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input 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
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601—Interfaces specially adapted for storage systems
    • G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
    • G06F3/0629—Configuration or reconfiguration of storage systems
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input 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
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601—Interfaces specially adapted for storage systems
    • G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
    • G06F3/0653—Monitoring storage devices or systems
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input 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
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601—Interfaces specially adapted for storage systems
    • G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
    • G06F3/0655—Vertical data movement, i.e. input-output transfer; data movement between one or more hosts and one or more storage devices
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input 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
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601—Interfaces specially adapted for storage systems
    • G06F3/0668—Interfaces specially adapted for storage systems adopting a particular infrastructure
    • G06F3/067—Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS]

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

An embodiment of the present invention provides a kind of date storage method, apparatus and system, method is：Control node is after the storage request that client is sent is received；It is asked according to the storage, determines target storage management cluster；Then the status information of target storage cluster is obtained from the corresponding target storage cluster of target storage management cluster, further according to the status information of acquisition, determines to be in the storage management node of existing state in target storage management cluster；And target storage management node is determined from the storage management node in existing state；Request target storage management node determines the first memory node for storing target data according to storage request, so that client stores target data to the first memory node.Data storage is carried out using scheme provided in an embodiment of the present invention, control node can just obtain the corresponding status information of storage cluster after storage request is received, not need to be waited for, and save the time of data storage.

Description

A kind of date storage method, apparatus and system
Technical field
The present invention relates to resource management techniques field, more particularly to a kind of date storage method, apparatus and system.
Background technology
At present, common data-storage system generally includes：Control node, storage management cluster, storage cluster, wherein depositing Storage management cluster refers to the set of multiple storage management nodes, and storage cluster refers to the set of multiple memory nodes, storage tube Reason cluster is connected with storage cluster, it is generally the case that storage management node in storage management cluster can in storage cluster One or more memory nodes be connected.
Existing data storage procedure is substantially：Control node receives the storage for target data that client is sent please It asks, then asks one storage management node of distribution for the storage, above-mentioned target data is stored to the storage management section In the corresponding memory node of point.In the above process, when asking distribution storage management node for the storage, need to storage tube Reason cluster carries out spy work, finds the storage management node in existing state, the storage management node in existing state refers to It is capable of providing the node of mobile sms service.And existing spy mechanism living is typically periodically to visit work, i.e., control node is periodical Issued to storage management cluster and visit life and enable, storage management cluster receives requested database after above-mentioned spy life enables and carries out more Newly, after the completion of update, control node is determined to be in storage management cluster and be deposited further according to the state data memory information of database The storage management node of state living, then control node a storage tube is selected from the storage management node in existing state Node is managed, as target storage management node, by target storage management node for above-mentioned storage request one storage of distribution Node, so that client stores above-mentioned target data into selected memory node；That is control node only has The existing state of each storage management node can be determined after issuing the period start time for visiting life order and arriving, also could Enough for above-mentioned storage management node of the storage request distribution in existing state.
But if above-mentioned storage request is received between two adjacent periods initial times, then this is deposited Storage request is needed after waiting until that the latter period start time arrives, and control node can be in for storage request distribution and deposit Therefore the storage management node of state living, stores data, although being capable of complete paired data using above-mentioned date storage method Storage, but need to carry out after period start time arrives for storage request distribution storage management node, this The sample excessive time will be wasted in waiting so that the time of data storage is elongated.
Invention content
The embodiment of the present invention is designed to provide a kind of date storage method, apparatus and system, to save data storage Time.
In order to achieve the above objectives, the embodiment of the invention discloses a kind of date storage method, applied to data-storage system In control node, the method includes：
Receive the storage request for being used to store target data that client is sent；
It is asked according to the storage, target storage management collection is determined from the storage management cluster of the data-storage system Group；
The state of the target storage cluster is obtained from the corresponding target storage cluster of the target storage management cluster Information；
According to the status information obtained, determine to be in the storage management of existing state in the target storage management cluster Node；
Target storage management node is determined from the storage management node in existing state；
The target storage management node is asked to be determined according to the storage request for store the target data the One memory node, so that the client stores the target data to first memory node.
Preferably, it is described according to the status information obtained, it determines in the target storage management cluster in survival shape The storage management node of state, including：
According to the status information and preset correspondence obtained, determine to be in the target storage management cluster The storage management node of existing state, wherein, the correspondence is：Each storage management in the target storage management cluster Correspondence in node and the target storage cluster between each memory node.
Preferably, it is described that target storage management node is determined from the storage management node in existing state, including：
According to load balancing principle and/or the uniform principle of number of requests, from the storage management node in existing state Determine target storage management node.
Preferably, the target storage management node belongs to comprising at least two including the target storage management cluster During storage management cluster, first memory node is：In the target storage cluster with the target storage management node pair The memory node answered.
Preferably, the method further includes：
Described first is monitored during the client stores the target data to first memory node to deposit Whether storage node breaks down；
If it is, the target storage management node is asked to determine the second memory node, so that the client is by institute Target data is stated to store to second memory node, wherein, second memory node is：In the target storage cluster with Memory node in the corresponding memory node of the target storage management node in addition to first memory node.
Preferably, the method further includes：
The target during the client stores the target data to first memory node is monitored to deposit Whether storage management node asks upgrade maintenance；
If it is, be dead state by the status indication of the target storage management node, return perform it is described from The step of target storage management node is determined in the storage management node of existing state.
In order to achieve the above objectives, the embodiment of the invention also discloses a kind of data storage devices, and system is stored applied to data Control node in system, described device include：
Receiving module, for receiving being asked for storing the storage of target data for client transmission；
First determining module, for being asked according to the storage, from the storage management cluster of the data-storage system Determine target storage management cluster；
Module is obtained, is deposited for obtaining the target from the corresponding target storage cluster of the target storage management cluster The status information of accumulation；
Second determining module, for according to the status information obtained, determining to be in the target storage management cluster The storage management node of existing state；
Third determining module, for determining target storage management node from the storage management node in existing state；
First memory module, for the target storage management node to be asked to determine to store according to the storage request First memory node of the target data saves so that the client stores the target data to the described first storage Point.
Preferably, second determining module, is specifically used for：
According to the status information and preset correspondence obtained, determine to be in the target storage management cluster The storage management node of existing state, wherein, the correspondence is：Each storage management in the target storage management cluster Correspondence in node and the target storage cluster between each memory node.
Preferably, the third determining module, is specifically used for：
According to load balancing principle and/or the uniform principle of number of requests, from the storage management node in existing state Determine target storage management node.
Preferably, the target storage management node belongs to comprising at least two including the target storage management cluster During storage management cluster, first memory node is：In the target storage cluster with the target storage management node pair The memory node answered.
Preferably, described device further includes：
First monitoring modular stores the target data to first memory node for monitoring the client Whether first memory node breaks down in the process；
Second memory module, for the feelings to break down in first monitoring module monitors to first memory node Under condition, the target storage management node is asked to determine the second memory node, so that the client is by the target data It stores to second memory node, wherein, second memory node is：It is deposited in the target storage cluster with the target Store up the memory node in addition to first memory node in the corresponding memory node of management node.
Preferably, described device further includes：
Second monitoring modular stores the target data to first memory node for monitoring the client Whether the target storage management node asks upgrade maintenance in the process；
Trigger module, for asking upgrade maintenance in second monitoring module monitors to the target storage management node In the case of, it is dead state by the status indication of the target storage management node, triggers third determining module.
In order to achieve the above objectives, the embodiment of the invention also discloses a kind of data-storage system, the system comprises：Control Node, storage management cluster and storage cluster, it is described comprising at least one storage management node in the storage management cluster At least one memory node is included in storage cluster；
The control node, for receiving being asked for storing the storage of target data for client transmission；According to described Storage request, determines target storage management cluster from the storage management cluster；It is corresponded to from the target storage management cluster Target storage cluster in obtain the status information of the target storage cluster；According to the status information obtained, determine described The storage management node of existing state is in target storage management cluster；From the storage management node in existing state really The storage management that sets the goal node；The target storage management node is asked to determine to store the mesh according to the storage request Mark the first memory node of data；
The target storage management node, for determined according to the storage request for store the target data the One memory node；
First memory node, for receiving and storing the target data that the client is sent.
Therefore in scheme provided in an embodiment of the present invention, control node please in the storage for receiving client transmission After asking；It is asked according to the storage, determines target storage management cluster；Then it is deposited from the corresponding target of target storage management cluster The status information of target storage cluster is obtained in accumulation, further according to the status information of acquisition, determines target storage management cluster In be in existing state storage management node；And target storage management is determined from the storage management node in existing state Node；Request target storage management node determines the first memory node for storing target data according to storage request, with So that client stores target data to the first memory node.Data are carried out using scheme provided in an embodiment of the present invention to deposit Storage, control node can just obtain the corresponding status information of storage cluster after storage request is received, do not need to carry out etc. It treats, saves the time of data storage.
Description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, to embodiment or will show below There is attached drawing needed in technology description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this Some embodiments of invention, for those of ordinary skill in the art, without creative efforts, can be with Other attached drawings are obtained according to these attached drawings.
Fig. 1 is a kind of flow diagram of date storage method provided in an embodiment of the present invention；
Fig. 2 is a kind of schematic diagram of detailed process of data provided in an embodiment of the present invention storage；
Fig. 3 is the schematic diagram of another detailed process of data provided in an embodiment of the present invention storage；
Fig. 4 is the flow diagram of another date storage method provided in an embodiment of the present invention；
Fig. 5 is the flow diagram of another date storage method provided in an embodiment of the present invention；
Fig. 6 is a kind of structure diagram of data storage device provided in an embodiment of the present invention；
Fig. 7 is the structure diagram of another data storage device provided in an embodiment of the present invention；
Fig. 8 is the structure diagram of another data storage device provided in an embodiment of the present invention.
Specific embodiment
Below in conjunction with the attached drawing in the embodiment of the present invention, the technical solution in the embodiment of the present invention is carried out clear, complete Site preparation describes, it is clear that described embodiment is only part of the embodiment of the present invention, instead of all the embodiments.It is based on Embodiment in the present invention, those of ordinary skill in the art are obtained every other without making creative work Embodiment shall fall within the protection scope of the present invention.
Below by specific embodiment, the present invention will be described in detail.
Fig. 1 is a kind of flow diagram of date storage method provided in an embodiment of the present invention, and the method comprising the steps of：
S101：Receive the storage request for being used to store target data that client is sent.
With universal and big data technology the rise of network, people are no longer limited to production, life, study, work The data generated in work are stored into the storage mediums such as local storage space or mobile hard disk, flash disk, but more and more will Above-mentioned data are stored into network storage space.
Due to the increase to data storage requirement of development and user of network, can be included in general data-storage system There are control node, multiple storage management clusters and multiple storage management clusters, can include in each storage management cluster more A storage management node, can include multiple memory nodes in each storage cluster, storage management cluster and storage cluster it Between often have correspondence, in order to facilitate data storage and the management of resource, a usual storage management cluster corresponds to one Storage cluster, the storage management node in storage management cluster can communicate with one or more of storage cluster memory node Connection.
For a user, user can by client that it is used send out store the storages of these above-mentioned data please It asks, in scheme provided in an embodiment of the present invention, the data stored are referred to as target data.Client mentioned here It is cloud client that can be, but not limited to, for example, Kingsoft enterprise cloud disk, illustratively, it is assumed that user is want a local data Size is that the video compress packet A of 50GB is stored into network storage space, then, user can pass through client " Kingsoft at this time Control node of the enterprise's cloud disk " into network end data storage system sends corresponding storage request, and control node is receiving After storage request, a process is can trigger to handle storage request, it is generally the case that a process can handle a storage Request.Storage request can include the attribute information of target data, such as：Title, size of data, data type of target data etc. Deng illustratively, storage request can be comprising the attribute information of target data：Video compress packet A, 50GB, ZIP compression text Part.
S102：It is asked according to the storage, target storage is determined from the storage management cluster of the data-storage system Manage cluster.
It will be appreciated to those of skill in the art that different storage management clusters can be realized to different types of data Resource is managed, for example, it is assumed that including 3 storage management clusters in data-storage system, respectively：cluster1、 Cluster2 and cluster3, wherein, cluster1 can be responsible for the data resource of document, image type, cluster2 Can be responsible for the data resource of compressed file type, cluster3 can be responsible for sound, video type data resource. Certainly, each storage management cluster might not be merely responsible for managing above-mentioned corresponding data resource, can also be in other storage tubes When managing cluster load excessive, to share the data resource of a part of the type.
As from the foregoing, control node can receive the storage for target data that user is sent by client Request includes the attribute information of target data in the storage request received due to control node, control node can The attribute information of target data included in being asked according to above-mentioned storage determines to manage from multiple storage management clusters State the target storage management node of storage request.For example, above-mentioned attribute information of the storage request comprising target data can be：Depending on Frequency compressed package A, 50GB, ZIP compressed file, then, control node can be determined that the storage is asked according to above-mentioned attribute information Corresponding target data type is compressed file, from the foregoing, it will be observed that above-mentioned 3 storage management clusters cluster1, cluster2 In cluster3, cluster2 can be responsible for the data resource of compressed file type, and therefore, control node is from above-mentioned 3 It is target storage management cluster that cluster2 is determined in a storage management cluster cluster1, cluster2 and cluster3.
S103：The target storage cluster is obtained from the corresponding target storage cluster of the target storage management cluster Status information.
From the foregoing, it will be observed that storage management cluster is corresponding with storage cluster, a usual storage management cluster corresponds to one and deposits Accumulation, control node can determine to store with target from multiple storage clusters after target storage management cluster is determined The corresponding storage cluster of cluster is managed, which is referred to as target storage cluster.
It will be appreciated to those of skill in the art that in data-storage system, storage cluster is the final of storage data Carrier, storage cluster can be understood as a very large memory space, and the total amount of general memory space can reach TB (Terabyte, terabyte or terabyte) order of magnitude, specific store tasks will be come specifically by its internal memory node Realize, memory node can there are many form, such as：Ceph, sheepdog, emc etc..Respectively depositing in same storage cluster It can be identical form to store up node, or different forms, if each memory node is same shape in storage cluster Formula, then such storage cluster is referred to as stand-alone mode, if each memory node is not same form in storage cluster, Such storage cluster is referred to as mixed insertion pattern.Certainly, the embodiment of the present invention is not to memory node each in storage cluster Concrete form, which is done, clearly to be limited.
As shown in Fig. 2, a kind of schematic diagram of detailed process for data provided in an embodiment of the present invention storage, shown in Fig. 2 Control node and cluster2 (upper dotted line frame) communication connection, it can be understood as control node is deposited with what is included in cluster2 Storage management node 1-4 is respectively provided with communication connection, and cluster2 is communicated to connect with storage cluster 11 (lower dotted line frame), it can be understood as Each storage management node in cluster2 communicates to connect or can be understood as each storage in cluster2 with storage cluster 11 Management node is communicated to connect with each memory node 1-6 in storage cluster 11.Control node shown in Fig. 2 receives storage please It asks, and asks to determine target storage management cluster as cluster2 according to storage, and cluster2 correspondences and storage cluster 11 are logical Letter connection, therefore, storage cluster 11 is referred to as target storage cluster.It lives at this point, control node can issue spy to storage cluster 11 Order, storage cluster 11 feed back its corresponding status information after receiving spy life and enabling to control node.In order to true The state of data storage is accurately obtained, in specific implementation provided in an embodiment of the present invention, control node can pass through Following two situations obtain the status information of target storage cluster.
Below by taking Fig. 2 as an example, the status information for obtaining target storage cluster is illustrated.
A kind of situation is：Control node can be directed to the storage received and ask to storage cluster 11 (target storage cluster) In each memory node issue one and visit life and enable, which, which enables, believes itself state for triggering each memory node Breath feeds back to control node；By the status information of each memory node, the state of target storage cluster can be accurately obtained Information.
Under normal conditions, each memory node in target storage cluster is received after the spy life that control node issues enables Can be immediately by the status information feedback of itself to control node, whole process is very short almost without time delay or time delay, but if mesh If certain memory node in mark storage cluster breaks down, then the memory node will not then feed back its correspondence to control node Status information, then control node will always be in wait state, can not obtain the status information of the target storage cluster.For The above-mentioned memory node is avoided to break down and cause the stand-by period of control node long, control node collects being stored to target Group in each memory node issue visit life enable after, it can be determined that it is anti-whether can to receive each memory node in preset time period The status information of feedback if the status information of certain memory node feedback is not received by within the preset period, is recognized For the moment memory node failure, in down state, control node according to the status information of the memory node received, It can obtain the corresponding status information of target storage cluster.
Another situation is：Assuming that it is provided with to record each memory node status information or deposit in storage cluster 11 The service node of accumulation status information asks to deposit cluster to target at this point, control node can also be directed to the storage received In service node issue and visit life and enable, received after above-mentioned spy life enables in service node, service node is recorded each The status information of memory node or the status information feedback of storage cluster 11 are to control node, the shape recorded according to service node State information particular content is different, and following two scenes can be divided to be introduced；Scene one：Service node is dedicated for record storage collection The status information of group 11, the status information of storage cluster 11 can include：The residual memory space total amount of storage cluster 11, storage Residual memory space is more than the quantity of memory node of 50GB, the total quantity of memory node and in writable shape in cluster 11 Quantity of memory node of state etc..Under the scene, the state letter for the storage cluster 11 that service node records its current time The status information of storage cluster 11 that breath feeds back to control node or the service node can record a record period thereon Feed back to control node.
Scene two, service node is dedicated for recording each memory node status information, then, service node receives above-mentioned Spy life enable after, service node by the status information feedback of each memory node that its current time records to control node or The status information feedback of each memory node that person's service node can record a record period thereon is to control node.Service During the concrete form of each memory node status information recorded in node can have very much, common a kind of form the simplest is table Form, for example, the corresponding status informations of memory node 1-6 recorded in service node are as shown in table 1 below, specially：
Table 1
Memory node title Whether writable state is in Memory space surplus Memory space total amount
Memory node 1 It is no 20GB 240GB
Memory node 2 It is 100GB 240GB
Memory node 3 It is 75GB 240GB
Memory node 4 \ \ \
Memory node 5 It is 80GB 240GB
Memory node 6 It is 128GB 240GB
It should be noted that table 1 is only a kind of specific example provided in an embodiment of the present invention, the application is not to storage The quantity of memory node and the particular state information of each memory node included in cluster 11 (target storage cluster) is done clearly It limits.Above-mentioned described record period is periodically can be to target storage cluster for the service node in target storage cluster The corresponding status information of each memory node is counted in corresponding status information or target storage cluster, such as set Record period is 0.5 second, that is that is, 0.5 second status information that will record a target storage cluster.In order not to increase number It is loaded, and according to the operation of storage system since the state data memory of short time memory accumulation varies less, then in the clothes After business node receives spy life order, it can not have in real time record the status information of target storage cluster, by this The status data of preceding record period record as current time status information feedback to control node.
Therefore no matter by which kind of above-mentioned mode, control node can be obtained from target storage cluster about The status information of target storage cluster, but under normal conditions, the corresponding status information of each memory node can more accurately bodies The now state data memory of entire target storage cluster, therefore, service node can save storage each in target storage cluster The corresponding status information feedback of point to control node, but the embodiment of the present invention the particular content of above-mentioned status information is not done it is bright Fidelity is determined.
S104：According to the status information obtained, depositing in existing state in the target storage management cluster is determined Store up management node.
It will be appreciated to those of skill in the art that each storage management node is for managing storage in storage management cluster Corresponding data resource in cluster, therefore, the storage state of storage cluster will directly affect each storage tube in storage management cluster The existing state of node is managed, so can directly judge each storage tube in storage management cluster according to the status information of storage cluster Manage the corresponding existing state of node.
With reference to the foregoing embodiments, the status information of target storage cluster that control node receives can include target and store The quantity of memory node of the residual memory space more than 50GB in the residual memory space total amount of cluster, target storage cluster is deposited Store up the total quantity of node and the quantity of the memory node in writable state, it is assumed that the status information that control node receives It is 45GB for storage cluster 11 (target storage cluster) remaining memory space total amount, the storage section of the big 50GB of residual memory space Point have 0, the total quantity of memory node be 6, the memory node in writable state has 4, target data is：Data are big The small video compress packet A for 50GB, based on above-mentioned it is found that 11 remaining memory space total amount of storage cluster is only 45GB, although There are the memory node that 4 are in writable state in storage cluster 11, but since the size of data of target data is 50GB, because This, the above-mentioned memory node in writable state can not store target data, it is possible to judgement and storage cluster All storage management nodes in 11 corresponding cluster2 (target storage management cluster) can not provide management service, i.e., Each storage management node is in dead state in the target storage management cluster.
From the foregoing, it will be observed that between the storage management node in cluster2 and each memory node in storage cluster 11, there are phases The relationship answered, it is generally the case that the storage management node in cluster2 can be with multiple memory nodes in storage cluster 11 Communication connection, which can be managed the data resource of memory node connected to it, in order to further The existing state to storage management node judge, in a kind of specific implementation provided in an embodiment of the present invention, institute The storage management node for according to the status information obtained, determining that existing state is in the target storage management cluster is stated, It can be with：
According to the status information and preset correspondence obtained, determine to be in the target storage management cluster The storage management node of existing state, wherein, the correspondence is：Each storage management in the target storage management cluster Correspondence in node and the target storage cluster between each memory node.
As shown in Table 1 above, 6 memory nodes are shared in the target storage cluster shown in table 1, wherein, memory node 1 In can not write state, the corresponding memory space total amounts of memory node 1-6 are 240GB, and still, memory node 4 is right in table 1 The status information answered is not recorded, and memory node 4 probably breaks down this explanation at this time, and memory node 2 is in writeable Enter state, memory space surplus be 100GB, memory node 3 be in writable state, memory space surplus be 75GB, deposit Storage node 5 is in writable state, memory space surplus is 80GB, and memory node 6 is in writable state, memory space remains Surplus is 128GB；Assuming that sharing 4 storage management nodes in cluster2, storage management node 1-4 is respectively designated as, is preset Correspondence can be：
The correspondence of storage management node 1 is communicated to connect with the memory node 1 in storage cluster 11 and memory node 4；
The correspondence of storage management node 2 communicates with memory node 2, memory node 3 and the memory node 6 in storage cluster 11 Connection；
Storage management node 3 corresponds to and 5 communication link of memory node 1, memory node 2 and memory node in storage cluster It connects；
The correspondence of storage management node 4 is communicated to connect with the memory node 5 in storage cluster.
By the status information and preset correspondence of each memory node in above-mentioned table 1 it is found that with storage management node 1 The memory node 1 of communication connection be in can not write state, and the memory node 4 communicated to connect with storage management and node 1 is sent out Raw failure, hence, it can be determined that storage management node 1 is now in dead state；
It is in memory node 2, memory node 3 and memory node 6 that storage management node 2 communicates to connect writable State, and remaining memory space is all higher than 50GB, hence, it can be determined that storage management node 2 is now in existing state；
The memory node communicated to connect with storage management node 3 for memory node 1, memory node 2 and memory node 5, Wherein, memory node 1 be in can not write state, but memory node 2 and memory node 5 are in writable state and residue is deposited Storage space is all higher than 50GB, then, at this time if storage management node 3 receives storage request, storage request can be directed to Target data is stored, hence, it can be determined that storage management node 3 is now in existing state；
The memory node communicated to connect with storage management node 4 only has memory node 5, and memory node 5 is in writable State and remaining memory space are more than 50GB, hence, it can be determined that storage management node 4 is also at existing state at this time.
It can to sum up determine the storage management node in storage state in cluster2 (target storage management cluster) Have：Storage management node 2, storage management node 3 and storage management node 4.It should be noted that preset corresponding pass in above-mentioned System is only a kind of specific example provided in an embodiment of the present invention, and the application does not do the concrete form of correspondence and clearly limits It is fixed.
S105：Target storage management node is determined from the storage management node in existing state.
Control node is in cluster2 is determined after the storage management node of existing state, can be from survival One is determined in the storage management node of state and is used as target storage management node.Such as random manner can be taken from all One is randomly selected in cluster2 in all storage management nodes in existing state and is used as target storage management node, but Choosing target storage management node using random manner may cause certain storage management nodes often to be hit, and other Storage management node is then idle for a long time, this will cause resource allocation uneven, it is thus possible to the phenomenon that causing network congestion.
For it is above-mentioned the phenomenon that, it is above-mentioned from survival in a kind of specific implementation provided in an embodiment of the present invention Target storage management node is determined in the storage management node of state, it can be with：
According to load balancing principle and/or the uniform principle of number of requests, from the storage management node in existing state Determine target storage management node.
In order to the operation for enabling entire data-storage system uniform and stable, usual control node is receiving visitor After the storage request that family end is sent, target storage management node can be determined according to load balancing principle.Illustratively, it is false If above-mentioned have determined that the storage management node in existing state has storage management node 2, storage management section in cluster2 Point 3 and storage management node 4, wherein 2 corresponding memory node 2 of storage management node, memory node 3 and memory node 6 are distinguished Corresponding residual memory space is 100GB, 75GB and 128GB, then, 2 corresponding residual memory space of storage management node is total It measures as 303GB；3 corresponding memory node 1 of storage management node, memory node 2 and the corresponding remaining storage of memory node 5 Space is 20GB, 100GB and 80GB, but be in can not write state for memory node 1, then 1 corresponding residue of memory node Memory space will not consider that therefore, 3 corresponding residual memory space total amount of storage management node is 180GB；Storage management node 4 corresponding 5 remaining memory spaces of memory node are 80GB, then 4 corresponding residual memory space total amount of storage management node For 80GB；Under normal conditions, the residual memory space total amount the big, illustrates that the data that can be stored are more, i.e., load capacity is got over By force, to sum up, 2 corresponding residual memory space total amount of storage management node is maximum, therefore, can be by storage management node 2 It is determined as target storage management node.
Target storage management node can also be determined according to the uniform principle of number of requests, those skilled in the art can manage Solution, in target storage management cluster each storage management node except name it is different in addition to, in structure be it is reciprocity, therefore, can To determine target according to the number of requests that respectively the storage management node in existing state receives in target storage management cluster Storage management node, such as will can be in the storage management node of existing state to receive in target storage management cluster and ask That for seeking minimum number is determined as target storage management node；Control node can also count received in a period of time please Quantity is sought, it is which is a to determine the storage received request, for example, the storage request that control node receives is the 91st, and this When cluster2 in the storage management node in existing state have 3,91/3 remainder is 1, then will determine first in 3 It is a, that is, storage management node 2 is target storage management node.
Or can also determine target storage management node according to load balancing principle and the uniform principle of number of requests, just It is finally to determine in the storage management node in existing state from target storage management cluster with reference to above two principle Go out target storage management node.
It should be noted that above-mentioned determine target storage management node from the storage management node in existing state Method is only several specific examples provided in an embodiment of the present invention, does not form the tool used to determining target storage management node The restriction of body mode, any modification, equivalent replacement, improvement and so on, include all within the spirits and principles of the present invention Within the scope of the present invention.
S106：The target storage management node is asked to determine to store the target data according to the storage request The first memory node so that the client stores the target data to first memory node.
As from the foregoing, the storage management node in cluster2 can be deposited with one or more of storage cluster 11 Store up node communication connection, it is assumed that the above-mentioned target storage management node determined is storage management node 2, then, it controls at this time Node will ask storage management node 2 to distribute specific memory space for target data, and above-mentioned storage management node 2 is corresponding Memory node is memory node 2, memory node 3 and memory node 6, and storage management node determines to deposit in which way First memory node the application of storage target data does not clearly state it, and concrete mode can be deposited according to specific data Storage structure determines, for example, distributed storage architecture, more copy storage organizations etc. may be used in data store organisation.
Illustratively, it is assumed that storage management node 2 is corresponding surplus according to memory node 2, memory node 3 and memory node 6 Complementary space size determines the first memory node, it may be determined that memory node 6 is first memory node.Determining memory node 6 Can to be stored after the first memory node to target data, can at least have to the storing process of target data following several Kind mode：
Mode 1, storage management node 2 determine memory node 6 for after the first memory node, memory node 6 is corresponded to The information such as network address feed back to control node, control node can after the network address for receiving above-mentioned memory node 6 To send a response instruction to client, while the network address of memory node 6 is sent to client, such client energy Enough by 6 corresponding network address of memory node, memory node 6 is directly accessed and target data is stored to memory node 6 In, so as to complete the storage to target data.
Mode 2, storage management node 2 determine memory node 6 as after the first memory node, are the by memory node 6 The message feedback of one memory node is to control node, after control node determines 6 corresponding network address of memory node, to visitor Family end sends a response instruction, while the network address of memory node 6 is sent to client, and such client is receiving Memory node 6 can directly be accessed after above-mentioned network address and store target data into memory node 6, complete to mesh Mark the storage of data.
When memory node 6 is determined as the first memory node by storage management node 2, one is issued to memory node 6 for mode 3 A active request store instruction, memory node 6 is connected according to the active request store instruction to be established with client, and then is realized Storage to target data.
It should be noted that above-mentioned only list several frequently seen data storage method, not the embodiment of the present invention is to depositing The exhaustion of storage mode, therefore, the above do not store target data to the concrete mode in the first memory node to form and limit.
It is noted that during above-mentioned only by taking storage management node individually belongs to a certain storage management cluster as an example It illustrates, but in order to save the cost that data-storage system is built, the ratio that the quantity of storage management node is often set Less, this will will appear the phenomenon that same storage management node belongs to multiple storage management clusters.
For above-mentioned situation, in a kind of specific implementation provided in an embodiment of the present invention, the target storage tube When reason node is belonged to comprising at least two storage management cluster including the target storage management cluster, the first storage section It puts and is：Memory node corresponding with the target storage management node in the target storage cluster.
In the case where target storage management node belongs to multiple storage management clusters simultaneously, it is above-mentioned determining for storing During the first memory node of target data, storage management cluster corresponding from aforementioned fixed target storage management cluster is needed In select in memory node corresponding with target storage management node.
As shown in figure 3, the schematic diagram of another detailed process for data provided in an embodiment of the present invention storage, by Fig. 3 As it can be seen that control node and cluster2 and cluster3 are communicated to connect, it can be understood as control node can with cluster2 and Each storage management node communication connection in cluster3, for control node, communicates with the storage management collection of connection Group be in form it is reciprocity, as seen from Figure 3, with cluster2 communication connection be storage cluster 11, communicate with cluster3 Connection is storage cluster 21, then, storage management node 2 had not only belonged to cluster2 but also had belonged to cluster3, if control section When point asks to determine cluster2 as target storage management cluster according to storage, this also means that the corresponding storage of target data please Asking will be managed by the storage management node 2 in cluster2, and therefore, storage management node 2 is determining the first storage section Point when, also will be in cluster2 corresponds to storage cluster 11 with 2 corresponding memory node 2 of storage management node, memory node 3 It is determined in memory node 6.Also if control node is according to the determining target storage management cluster of storage request Cluster3, then, target data it is corresponding storage request by by cluster3 storage management node 2 (assuming that Storage management node 2 in cluster2 is also named as storage management node 2 in cluster3) it is managed, therefore, deposit Store up management node 2 when determining the first memory node, can in storage cluster 21 corresponding with cluster3 with storage management It is determined in 2 corresponding memory node of node, since the process is similar to the process of determining first memory node in cluster2, Therefore it will not be described in great detail at this.
Under normal conditions, in order to further save the cost of building of data-storage system, control node usually has parallel The high characteristic of degree, that is, control node being capable of parallel multiple processes simultaneously.By taking Fig. 3 as an example, it is assumed that control node connects simultaneously Two storage requests are received, wherein, storage request 1 is asked for the storage for target data, and storage request 2 is for another number According to storage ask, at this time control node will simultaneously parallel two processes, process 1 be used for handle storage request 1, process 2 is used for Processing stores request 2, it is assumed that control node determines that storage 1 corresponding target storage management cluster of request is cluster2, stores 2 corresponding target storage clusters are asked as cluster3, due to be between process 1 and process 2 it is parallel, the fortune of process 1 Row process and the operational process of process 2 are independent of each other.
Illustratively, it is assumed that the objective management node for storage request 1 determined in process 1 is depositing in cluster2 Management node 2 is stored up, the target storage management node for storage request 2 determined in process 2 is the storage tube in cluster3 Node 2 (assuming that the storage management node 2 in cluster2 is also named as storage management node 2 in cluster3) is managed, for For process 1, the corresponding storage request 1 of target data is managed by the storage management node 2 in cluster2, that , storage management node 2 when determining the first memory node, also will be in cluster2 corresponds to storage cluster 11 with storage tube It is determined in 2 corresponding memory node 2 of reason node, memory node 3 and memory node 6.For process 2, another data Corresponding storage request 2 is (assuming that the storage management node 2 in cluster2 exists by the storage management node 2 in cluster3 Also storage management node 2 is named as in cluster3) it is managed, therefore, storage management node 2 is determining the first storage section During point, it can be determined in storage cluster 21 corresponding with cluster3 with 2 corresponding memory node of storage management node.
Due to being run parallel between process 1 and process 2, the operating status of process 1 and the operating status of process 2 are mutual It does not influence, for example, process 2 breaks down before the storage to storage 2 other corresponding data of request is completed, leads to process 2 Corresponding storage management node 2 loses data managing capacity, i.e., in dead state, due to the operating status of process 1 and process 2 Operating status be independent of each other, even if storage management node 2 simultaneously belong to cluster2 and cluster3, then, belong to Storage management node 2 in cluster3, which is in dead state and nor affects on, belongs to depositing for storage management node 2 in cluster2 State living；Likewise, if the storage management node 2 for assuming to belong in cluster2 is in dead state, will not influence The existing state of storage management node 2 in cluster3.
It should be noted that it is above-mentioned only using the storage management node 2 in cluster2 belong to two storage management clusters as Example illustrates, but the embodiment of the present invention does not belong to storage management node the particular number of storage management cluster and limits It is fixed, for example, above-mentioned storage management node 2 can belong to cluster1, cluster2 and cluster3 simultaneously, meanwhile, target Storage management node may be storage management node 3, storage management node 4, and therefore, the embodiment of the present invention does not also deposit target The particular content of storage management node does further restriction.
To sum up, application drawing 1 provides embodiment, and control node is after the storage request that client is sent is received；Root It is asked according to the storage, determines target storage management cluster；Then from the corresponding target storage cluster of target storage management cluster The status information of target storage cluster is obtained, further according to the status information of acquisition, determines to be in target storage management cluster and deposit The storage management node of state living；And target storage management node is determined from the storage management node in existing state；Please Target storage management node is asked to determine the first memory node for storing target data according to storage request, so that client End stores target data to the first memory node.Data storage, control section are carried out using scheme provided in an embodiment of the present invention Point can just obtain the corresponding status information of storage cluster after storage request is received, and do not need to be waited for, save The time of data storage.Simultaneously as above-mentioned is that directly the storage cluster of bottom is carried out to visit work, therefore more accurate in real time The state data memory for reflecting data-storage system, disposably accept for storage request and provide guarantee, determining target Load Balanced principle and number of requests uniform principle are employed during storage management node, enables to uniformly dividing for data resource Cloth, the phenomenon that alleviating network congestion to a certain extent.
Fig. 4 is the flow diagram of another date storage method provided in an embodiment of the present invention, real except being provided including Fig. 1 It applies outside the step S101-S106 in example, this method further includes step：
S107：It monitors described during the client stores the target data to first memory node Whether one memory node breaks down；
S108：In the case where above-mentioned steps S107 monitors that first memory node breaks down, the mesh is asked Mark storage management node determines the second memory node, is deposited so that the client stores the target data to described second Node is stored up, wherein, second memory node is：It is corresponding with the target storage management node in the target storage cluster Memory node in memory node in addition to first memory node.
It can be there are many mode from the foregoing, it will be observed that being stored to target data, it is contemplated that target data is being stored to first It is possible that the situation of failure during memory node, in scheme provided in an embodiment of the present invention, control node can be with The storing process of target data is monitored when target data is stored to the first memory node and started, if monitoring first Memory node before target data storage does not complete if breaking down, then, control node can be stored from above-mentioned target With determining second memory node of conduct in other memory nodes of target storage management node communication connection in cluster, lay equal stress on Newly target data will be completed as previously described to store to the process of the second memory node.
Illustratively, it is assumed that in above-mentioned cluster2 storage management node 2 in storage cluster 11 memory node 2, deposit Storage node 3 and memory node 6 communicate to connect, and wherein memory node 6 is the first memory node, stores to storage and saves in target data During point 6, memory node 6 breaks down, and control node will exist again after monitoring that memory node 6 breaks down Determine a memory node as the second memory node in memory node 2 and memory node 3, it is assumed that it is the to determine memory node 2 Two memory nodes, since memory node 2 and memory node 6 are complete equity in structure, target data is stored to depositing Process in storage node 2 stores target data the same to the process in memory node 6 with aforementioned, and details are not described herein again.
As it can be seen that application drawing 4 provides embodiment, control node stores target data to the process progress of the first memory node Monitoring, if monitoring that the first memory node breaks down in the above process, control node will be deposited from above-mentioned target again With determining second memory node of conduct in other memory nodes of target storage management node communication connection in accumulation, and Target data is stored again into the second memory node, it will not be because of a certain memory node in this way in target data storing process Failure and the phenomenon that lead to target data storage failure, improve the stability of data-storage system.
Fig. 5 is the flow diagram of another storage method provided in an embodiment of the present invention, except including Fig. 1 offer embodiments In step S101-S106 outside, this method further includes step：
S109：Monitor the mesh during the client stores the target data to first memory node Whether mark storage management node asks upgrade maintenance；
S110：It, will in the case where above-mentioned steps S109 monitors the target storage management node request upgrade maintenance The status indication of the target storage management node is dead state, returns and performs step S105.
It will be appreciated to those of skill in the art that the priority ratio of upgrade maintenance is higher under normal conditions, normal conditions Under, as long as target storage management node asks upgrade maintenance, control node will interrupt to be run on the target storage management node All tasks, so that target control node can complete the upgrade maintenance task of itself, it is contemplated that target data is stored to Target storage management node carries out the situation of upgrade maintenance during one memory node, in scheme provided in an embodiment of the present invention In, control node can target data store to the first memory node start when, monitoring objective storage management node whether please Upgrade maintenance is sought, if monitoring target storage management node request upgrade maintenance, control node if stores the target The status indication of management node is dead state, and return to step S105 redefines a storage management node to take over the mesh The data administration tasks on storage management node are marked, to realize the free upgrade maintenance of target storage management node.
To sum up, the embodiment that application drawing 5 provides, can realize in target data storing process freely to target The upgrade maintenance of storage management node facilitates the later stage to work the repair and maintenance of data-storage system, further improves number According to the stability of storage system.
Fig. 6 is a kind of structure diagram of data storage device provided in an embodiment of the present invention, which includes：Receive mould Block 201, the first determining module 202 obtain module 203, the second determining module 204, the storage mould of third determining module 205 and first Block 206.
Receiving module 201, for receiving being asked for storing the storage of target data for client transmission；
First determining module 202, for being asked according to the storage, from the storage management cluster of the data-storage system In determine target storage management cluster；
Module 203 is obtained, for obtaining the mesh from the corresponding target storage cluster of the target storage management cluster Mark the status information of storage cluster；
Second determining module 204, for according to the status information obtained, determining in the target storage management cluster In the storage management node of existing state；
Further, the second above-mentioned determining module 204, is specifically used for：
According to the status information and preset correspondence obtained, determine to be in the target storage management cluster The storage management node of existing state, wherein, the correspondence is：Each storage management in the target storage management cluster Correspondence in node and the target storage cluster between each memory node.
Third determining module 205, for determining target storage management section from the storage management node in existing state Point；
Further, above-mentioned third determining module 205, is specifically used for：
According to load balancing principle and/or the uniform principle of number of requests, from the storage management node in existing state Determine target storage management node.
First memory module 206, for the target storage management node to be asked to determine to be used for according to the storage request The first memory node of the target data is stored, is deposited so that the client stores the target data to described first Store up node.
Further, including the target storage management node belongs to comprising the target storage management cluster at least During two storage management clusters, first memory node is：In the target storage cluster with the target storage management section The corresponding memory node of point.
Application drawing 6 provides embodiment, and control node is after the storage request that client is sent is received；It please according to the storage It asks, determines target storage management cluster；Then target is obtained from the corresponding target storage cluster of target storage management cluster to deposit The status information of accumulation further according to the status information of acquisition, determines depositing in existing state in target storage management cluster Store up management node；And target storage management node is determined from the storage management node in existing state；Request target stores Management node determines the first memory node for storing target data according to storage request, so that client is by number of targets According to storing to the first memory node.Data storage is carried out using scheme provided in an embodiment of the present invention, control node is receiving The corresponding status information of storage cluster can be just obtained after storage request, does not need to be waited for, saves data storage Time.Simultaneously as above-mentioned is that directly the storage cluster of bottom is carried out to visit work, therefore more accurately reflect number in real time According to the state data memory of storage system, disposably accepted for storage request and provide guarantee, determining target storage management section Load Balanced principle and the uniform principle of number of requests are employed during point, enables to being uniformly distributed for data resource, certain journey The phenomenon that network congestion is alleviated on degree.
Fig. 7 is the structure diagram of another data storage device provided in an embodiment of the present invention, and the device is in addition to including The receiving module 201 of Fig. 4 embodiments offer, the first determining module 202, acquisition module 203, the second determining module 204, third are true Outside 205 and first memory module 206 of cover half block, which further includes：First monitoring modular 207 and the second memory module 208.
First monitoring modular 207 stores the target data to the described first storage section for monitoring the client Whether first memory node breaks down during point；
Second memory module 208, for monitoring that event occurs in first memory node in first monitoring modular 207 In the case of barrier, the target storage management node is asked to determine the second memory node, so that the client is by the mesh Mark data are stored to second memory node, wherein, second memory node is：In the target storage cluster with it is described Memory node in the corresponding memory node of target storage management node in addition to first memory node.
Application drawing 7 provides embodiment, and control node stores target data to the process of the first memory node and is monitored, If monitoring that the first memory node breaks down in the above process, control node will be again from above-mentioned target storage cluster In with determining second memory node of conduct, and by target in other memory nodes of target storage management node communication connection Data are stored again into the second memory node, will not be due to a certain memory node failure in this way in target data storing process The phenomenon that leading to target data storage failure, improves the stability of data-storage system.
Fig. 8 is the structure diagram of another data storage device provided in an embodiment of the present invention, and the device is in addition to including The receiving module 201 of Fig. 4 embodiments offer, the first determining module 202, acquisition module 203, the second determining module 204, third are true Outside 205 and first memory module 206 of cover half block, which further includes：Second monitoring modular 209 and trigger module 210.
Second monitoring modular 209 stores the target data to the described first storage section for monitoring the client Whether the target storage management node asks upgrade maintenance during point；
Trigger module 210, for monitoring that the target storage management node request rises in second monitoring modular 209 In the case that grade is safeguarded, it is dead state by the status indication of the target storage management node, triggers third determining module.
The embodiment that application drawing 8 provides, can realize in target data storing process freely to target storage management The upgrade maintenance of node facilitates the later stage to work the repair and maintenance of data-storage system, further improves data storage system The stability of system.
The embodiment of the present invention additionally provides a kind of data-storage system, which includes：Control node, storage management cluster And storage cluster, comprising at least one storage management node in the storage management cluster, comprising extremely in the storage cluster A few memory node；
The control node, for receiving being asked for storing the storage of target data for client transmission；According to described Storage request, determines target storage management cluster from the storage management cluster；It is corresponded to from the target storage management cluster Target storage cluster in obtain the status information of the target storage cluster；According to the status information obtained, determine described The storage management node of existing state is in target storage management cluster；From the storage management node in existing state really The storage management that sets the goal node；The target storage management node is asked to determine to store the mesh according to the storage request Mark the first memory node of data；
The target storage management node, for determined according to the storage request for store the target data the One memory node；
First memory node, for receiving and storing the target data that the client is sent.
The control node, is specifically used for：
According to the status information and preset correspondence obtained, determine to be in the target storage management cluster The storage management node of existing state, wherein, the correspondence is：Each storage management in the target storage management cluster Correspondence in node and the target storage cluster between each memory node.
Further, the control node, is specifically used for：
According to load balancing principle and/or the uniform principle of number of requests, from the storage management node in existing state Determine target storage management node.
It should be noted that the target storage management node belong to comprising including the target storage management cluster extremely During few two storage management clusters, first memory node is：In the target storage cluster with the target storage management The corresponding memory node of node.
Data storage is carried out using data-storage system provided in an embodiment of the present invention, storage request can received The corresponding status information of storage cluster can be just obtained afterwards, does not need to be waited for, and save the time of data storage.Meanwhile Directly the storage cluster of bottom is carried out to visit work, more can accurately reflect the data storage of data-storage system in real time State disposably accepts for storage request and provides guarantee.
In another data-storage system provided in an embodiment of the present invention, control node is additionally operable to：
Described first is monitored during the client stores the target data to first memory node to deposit Whether storage node breaks down；
If it is, the target storage management node is asked to determine the second memory node, so that the client is by institute Target data is stated to store to second memory node, wherein, second memory node is：In the target storage cluster with Memory node in the corresponding memory node of the target storage management node in addition to first memory node.
In data-storage system provided in this embodiment, control node stores target data to the first memory node Process is monitored, and if monitoring that the first memory node breaks down in the above process, control node will be again from upper It states and determines that a conduct second is deposited in other memory nodes communicated to connect in target storage cluster with target storage management node Node is stored up, and target data is stored again into the second memory node, it will not be because of certain in this way in target data storing process One memory node failure and the phenomenon that lead to target data storage failure, improve the stability of data-storage system.
In another data-storage system provided in an embodiment of the present invention, control node is additionally operable to：
The target during the client stores the target data to first memory node is monitored to deposit Whether storage management node asks upgrade maintenance；
If it is, be dead state by the status indication of the target storage management node, return perform it is described from The step of target storage management node is determined in the storage management node of existing state.
In data-storage system provided in this embodiment, it can realize in target data storing process freely to mesh The upgrade maintenance of storage management node is marked, the later stage is facilitated to work the repair and maintenance of data-storage system, is further improved The stability of data-storage system.
Each embodiment in this specification is described using relevant mode, identical similar portion between each embodiment Point just to refer each other, and the highlights of each of the examples are difference from other examples.For above-mentioned device For embodiment and system embodiment, since it is substantially similar to embodiment of the method, so description is fairly simple, correlation Place illustrates referring to the part of embodiment of the method.
It should be noted that herein, relational terms such as first and second and the like are used merely to a reality Body or operation are distinguished with another entity or operation, are deposited without necessarily requiring or implying between these entities or operation In any this practical relationship or sequence.Moreover, term " comprising ", "comprising" or its any other variant are intended to Non-exclusive inclusion, so that process, method, article or equipment including a series of elements not only will including those Element, but also including other elements that are not explicitly listed or further include as this process, method, article or equipment Intrinsic element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that Also there are other identical elements in process, method, article or equipment including the element.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the scope of the present invention.It is all Any modification, equivalent replacement, improvement and so within the spirit and principles in the present invention, are all contained in protection scope of the present invention It is interior.

Claims (13)

  1. A kind of 1. date storage method, applied to the control node in data-storage system, which is characterized in that the method packet It includes：
    Receive the storage request for being used to store target data that client is sent；
    It is asked according to the storage, target storage management cluster is determined from the storage management cluster of the data-storage system；
    The status information of the target storage cluster is obtained from the corresponding target storage cluster of the target storage management cluster；
    According to the status information obtained, determine to be in the storage management section of existing state in the target storage management cluster Point；
    Target storage management node is determined from the storage management node in existing state；
    The target storage management node is asked to determine to deposit for storing the first of the target data according to the storage request Node is stored up, so that the client stores the target data to first memory node.
  2. 2. according to the method described in claim 1, it is characterized in that, described according to the status information obtained, the mesh is determined The storage management node that existing state is in storage management cluster is marked, including：
    According to the status information and preset correspondence obtained, determine in the target storage management cluster in survival The storage management node of state, wherein, the correspondence is：Each storage management node in the target storage management cluster With the correspondence between each memory node in the target storage cluster.
  3. 3. method according to claim 1 or 2, which is characterized in that described from the storage management node in existing state In determine target storage management node, including：
    According to load balancing principle and/or the uniform principle of number of requests, determined from the storage management node in existing state Target storage management node.
  4. 4. method according to claim 1 or 2, which is characterized in that
    The target storage management node belongs to comprising at least two storage management collection including the target storage management cluster During group, first memory node is：Storage section corresponding with the target storage management node in the target storage cluster Point.
  5. 5. method according to claim 1 or 2, which is characterized in that the method further includes：
    Monitor the first storage section during the client stores the target data to first memory node Whether point breaks down；
    If it is, the target storage management node is asked to determine the second memory node, so that the client is by the mesh Mark data are stored to second memory node, wherein, second memory node is：In the target storage cluster with it is described Memory node in the corresponding memory node of target storage management node in addition to first memory node.
  6. 6. according to the method described in claim 1, it is characterized in that, the method further includes：
    Monitor the target storage tube during the client stores the target data to first memory node Whether reason node asks upgrade maintenance；
    If it is, being dead state by the status indication of the target storage management node, return described in performing from survival The step of target storage management node is determined in the storage management node of state.
  7. A kind of 7. data storage device, applied to the control node in data-storage system, which is characterized in that described device packet It includes：
    Receiving module, for receiving being asked for storing the storage of target data for client transmission；
    First determining module for being asked according to the storage, is determined from the storage management cluster of the data-storage system Target storage management cluster；
    Module is obtained, for obtaining the target storage collection from the corresponding target storage cluster of the target storage management cluster The status information of group；
    Second determining module, for according to the status information obtained, determining in the target storage management cluster in survival The storage management node of state；
    Third determining module, for determining target storage management node from the storage management node in existing state；
    First memory module, for the target storage management node to be asked to be determined according to the storage request for described in storing First memory node of target data, so that the client stores the target data to first memory node.
  8. 8. device according to claim 7, which is characterized in that second determining module is specifically used for：
    According to the status information and preset correspondence obtained, determine in the target storage management cluster in survival The storage management node of state, wherein, the correspondence is：Each storage management node in the target storage management cluster With the correspondence between each memory node in the target storage cluster.
  9. 9. device according to claim 7 or 8, which is characterized in that the third determining module is specifically used for：
    According to load balancing principle and/or the uniform principle of number of requests, determined from the storage management node in existing state Target storage management node.
  10. 10. device according to claim 7 or 8, which is characterized in that
    The target storage management node belongs to comprising at least two storage management collection including the target storage management cluster During group, first memory node is：Storage section corresponding with the target storage management node in the target storage cluster Point.
  11. 11. device according to claim 7 or 8, which is characterized in that described device further includes：
    First monitoring modular stores the target data to the process of first memory node for monitoring the client Described in the first memory node whether break down；
    Second memory module, for situation about breaking down in first monitoring module monitors to first memory node Under, the target storage management node is asked to determine the second memory node, so that the client deposits the target data It stores up to second memory node, wherein, second memory node is：It is stored in the target storage cluster with the target Memory node in the corresponding memory node of management node in addition to first memory node.
  12. 12. device according to claim 7, which is characterized in that described device further includes：
    Second monitoring modular stores the target data to the process of first memory node for monitoring the client Described in target storage management node whether ask upgrade maintenance；
    Trigger module, for the feelings in second monitoring module monitors to target storage management node request upgrade maintenance Under condition, it is dead state by the status indication of the target storage management node, triggers third determining module.
  13. 13. a kind of data-storage system, which is characterized in that the system comprises：Control node, storage management cluster and storage Cluster comprising at least one storage management node in the storage management cluster, is deposited in the storage cluster comprising at least one Store up node；
    The control node, for receiving being asked for storing the storage of target data for client transmission；According to the storage Request determines target storage management cluster from the storage management cluster；From the corresponding mesh of the target storage management cluster The status information of the target storage cluster is obtained in mark storage cluster；According to the status information obtained, the target is determined The storage management node of existing state is in storage management cluster；Mesh is determined from the storage management node in existing state Mark storage management node；The target storage management node is asked to determine to store the number of targets according to the storage request According to the first memory node；
    The target storage management node, for determining to deposit for storing the first of the target data according to the storage request Store up node；
    First memory node, for receiving and storing the target data that the client is sent.
CN201611224468.4A 2016-12-27 2016-12-27 Data storage method, device and system Active CN108241469B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201611224468.4A CN108241469B (en) 2016-12-27 2016-12-27 Data storage method, device and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201611224468.4A CN108241469B (en) 2016-12-27 2016-12-27 Data storage method, device and system

Publications (2)

Publication Number Publication Date
CN108241469A true CN108241469A (en) 2018-07-03
CN108241469B CN108241469B (en) 2021-04-30

Family

ID=62701577

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201611224468.4A Active CN108241469B (en) 2016-12-27 2016-12-27 Data storage method, device and system

Country Status (1)

Country Link
CN (1) CN108241469B (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110267060A (en) * 2019-07-24 2019-09-20 北京视界云天科技有限公司 Video file stores method for implanting, device, computer equipment and storage medium
CN114553747A (en) * 2022-02-22 2022-05-27 度小满科技(北京)有限公司 Method, device, terminal and storage medium for detecting abnormality of redis cluster
WO2022143065A1 (en) * 2020-12-28 2022-07-07 网联清算有限公司 Database probing method and apparatus
WO2023185071A1 (en) * 2022-03-31 2023-10-05 北京沃东天骏信息技术有限公司 Data query method, data writing method, related apparatus and system

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103139302A (en) * 2013-02-07 2013-06-05 浙江大学 Real-time copy scheduling method considering load balancing
US20150154128A1 (en) * 2013-06-12 2015-06-04 International Business Machines Corporation Processing input/output requests using proxy and owner storage systems
CN105205160A (en) * 2015-09-29 2015-12-30 浙江宇视科技有限公司 Data write-in method and device

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103139302A (en) * 2013-02-07 2013-06-05 浙江大学 Real-time copy scheduling method considering load balancing
US20150154128A1 (en) * 2013-06-12 2015-06-04 International Business Machines Corporation Processing input/output requests using proxy and owner storage systems
CN105205160A (en) * 2015-09-29 2015-12-30 浙江宇视科技有限公司 Data write-in method and device

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110267060A (en) * 2019-07-24 2019-09-20 北京视界云天科技有限公司 Video file stores method for implanting, device, computer equipment and storage medium
CN110267060B (en) * 2019-07-24 2021-03-26 北京视界云天科技有限公司 Video file storage injection method and device, computer equipment and storage medium
WO2022143065A1 (en) * 2020-12-28 2022-07-07 网联清算有限公司 Database probing method and apparatus
CN114553747A (en) * 2022-02-22 2022-05-27 度小满科技(北京)有限公司 Method, device, terminal and storage medium for detecting abnormality of redis cluster
WO2023185071A1 (en) * 2022-03-31 2023-10-05 北京沃东天骏信息技术有限公司 Data query method, data writing method, related apparatus and system

Also Published As

Publication number Publication date
CN108241469B (en) 2021-04-30

Similar Documents

Publication Publication Date Title
CN105739924B (en) Caching method and system based on cache cluster
CN108241469A (en) A kind of date storage method, apparatus and system
CN105549905B (en) A kind of method that multi-dummy machine accesses distributed objects storage system
CN107111596B (en) The method of lock management, lock server and client in a kind of cluster
US6895485B1 (en) Configuring and monitoring data volumes in a consolidated storage array using one storage array to configure the other storage arrays
CN107566531A (en) A kind of Elasticsearch cluster expansion methods for supporting balanced resource
US5504888A (en) File updating system employing the temporary connection and disconnection of buffer storage to extended storage
CN104035925B (en) Date storage method, device and storage system
CN103856393A (en) Distributed message-oriented middleware system based on database and operating method thereof
CN108829352A (en) A kind of user's quota method and system of distributed memory system
CN103095805A (en) Cloud storage system of data intelligent and decentralized management
CN101443761A (en) QOS-enabled lifecycle management for file systems
CN110321301A (en) A kind of method and device of data processing
CN106936623A (en) The management method of distributed cache system and cache cluster
JP2007102762A (en) Resource management method in logically partition divisioned storage system
CN105700908A (en) Management system and control method for management system
CN108900626A (en) Date storage method, apparatus and system under a kind of cloud environment
CN109792393A (en) Virtualize the software upgrading in off-line accounting system
CN106383668A (en) Information storage method, storage management device and client
CN110083649A (en) It is a kind of based on cold, warm, dsc data data management system and data managing method
CN107085546A (en) Data managing method and device based on failure field technique
CN109407975A (en) Data writing method and calculate node and distributed memory system
CN101599978A (en) Resource allocation methods and equipment
CN101344882A (en) Data query method, insertion method and deletion method
CN105450679B (en) Carry out the method and system of data cloud storage

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
GR01 Patent grant
GR01 Patent grant