WO2013035719A1 - データ配置システム、分散アクセスノード、データ配置方法およびプログラム - Google Patents
データ配置システム、分散アクセスノード、データ配置方法およびプログラム Download PDFInfo
- Publication number
- WO2013035719A1 WO2013035719A1 PCT/JP2012/072559 JP2012072559W WO2013035719A1 WO 2013035719 A1 WO2013035719 A1 WO 2013035719A1 JP 2012072559 W JP2012072559 W JP 2012072559W WO 2013035719 A1 WO2013035719 A1 WO 2013035719A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- node
- client
- data
- area
- information
- Prior art date
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
- H04L67/1004—Server selection for load balancing
- H04L67/1021—Server selection for load balancing based on client or server locations
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
- H04L67/1004—Server selection for load balancing
- H04L67/1008—Server selection for load balancing based on parameters of servers, e.g. available memory or workload
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1097—Protocols in which an application is distributed across nodes in the network for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]
-
- 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]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Definitions
- the present invention is based on the priority claim of Japanese patent application: Japanese Patent Application No. 2011-193634 (filed on September 6, 2011), the entire contents of which are incorporated herein by reference. Shall.
- the present invention relates to a data arrangement system, a distributed access node, a data arrangement method, and a program, and more particularly, to a data arrangement system, a distributed access node, a data arrangement method, and a program that store data transmitted from a client in a plurality of nodes.
- Patent Document 1 discloses a management program that can achieve uniform load on storage nodes belonging to a distributed storage system.
- a computer that executes this management program functions as a load information acquisition unit, an allocation unit, and an instruction unit.
- the load information acquisition unit stores a plurality of pieces of access information for each real storage area included in each storage node. Obtained from the storage node, generates load information of each real storage area based on the obtained access information, and stores the generated load information in the load information storage means.
- the assignment means is stored in the load information storage means. Based on the load information, the allocation relationship of the real storage area to the virtual storage area is changed so that the load of each storage node is uniform. The storage node is instructed to move data in the real storage area.
- a data center is used as a storage destination of data that is regularly and frequently transmitted from a device such as a mobile phone or a car that always has a position (hereinafter referred to as “client node”).
- client node normally move, and the power supply may be turned off or may be excluded from management targets.
- client node determines from the contents of data notified to the data center and specifies another client node in the vicinity or wants to notify the other client node of any kind.
- a data center in which data is efficiently arranged is necessary.
- the data is arranged so that the load is uniform based on the access information such as the frequency of access to the data.
- the access information such as the frequency of access to the data.
- all data is accessed almost equally and frequently. There is a problem that load distribution according to access frequency is ineffective.
- the present invention is a data placement system that can contribute to load distribution on the node side as a storage destination even for data that is frequently updated over a wide range, such as data transmitted from the client node described above.
- An object is to provide a distributed access node, a data arrangement method, and a program.
- a plurality of nodes serving as storage destinations of data transmitted from a client node at a predetermined frequency
- a distributed access node disposed between the plurality of nodes and the client node
- the distributed access node specifies a node area information holding unit for holding node area information in which one or more area information is associated with each node, and an area where the client node is located.
- a data arrangement system including an area specifying unit and a node specifying unit that selects a node associated with the specified area as a storage destination of transmission data from the client node.
- one or more pieces of area information are arranged between a plurality of nodes serving as storage destinations of data transmitted from a client node and the client node.
- a node area information holding unit that holds node area information associated with each other, an area specifying unit that specifies an area where the client node is located, and a storage location of transmission data from the client node, which is associated with the specified area
- a distributed access node comprising a node specifying unit that selects a selected node.
- a data placement method for determining a storage destination of data sent from a client node from among a plurality of nodes, the placement being performed between the plurality of nodes and the client node.
- the distributed access node is configured to identify the area where the client node is located, and to refer to node area information in which each node is associated with one or more area information, and transmit data from the client node. Selecting a node associated with the identified area as a storage location of the data.
- This method is associated with a specific machine called a distributed access node that is arranged between a plurality of nodes and a client node and selects a data storage destination.
- an area where a client node is located in a computer constituting a distributed access node arranged between a plurality of nodes serving as storage destinations of data sent from the client node and the client node Referring to node area information in which one or more area information is associated with each node, and as a storage destination of transmission data from the client node, the node is associated with the identified area. And a process for selecting the selected node is provided.
- This program can be recorded on a computer-readable storage medium. That is, the present invention can be embodied as a computer program product.
- FIG. 4 is a diagram in which client nodes are plotted in an area corresponding to node area information in FIG. 3. It is a flowchart showing operation
- a plurality of nodes serving as storage destinations for data transmitted from a client node, and a distributed access arranged between the plurality of nodes and the client node.
- the distributed access node refers to an area specifying unit that specifies an area where the client node is located, and node area information that associates each node with one or more pieces of area information, A node specifying unit that selects a node associated with the specified area as a storage destination of transmission data from the client node;
- the distributed access nodes are the nodes (51) to (54).
- the node associated with the area where the client node is located is selected as a data storage destination.
- FIG. 1 is a block diagram showing the overall configuration of the first embodiment of the present invention.
- a configuration including client nodes 11 to 14, a network 20, a distributed access node 30, a distributed system network 40, and nodes 51 to 54 configured by servers, storage, and the like.
- client nodes 11 to 14 there is shown a configuration including client nodes 11 to 14, a network 20, a distributed access node 30, a distributed system network 40, and nodes 51 to 54 configured by servers, storage, and the like.
- four client nodes and nodes are shown, but these numbers are selected to simplify the understanding of the present invention, and more client nodes, You may comprise with a node.
- the client nodes 11 to 14 that use the distributed system are connected to the distributed access node 30 via the network 20.
- the client nodes 11 to 14 transmit data obtained from sensors provided in the individual client nodes to the distributed access node 30.
- the client nodes 11 to 14 transmit information that can specify the area where the self-device is located in a form included in the data or attached to the data.
- the distributed access node 30 determines the data storage location uniquely by the distributed arrangement algorithm described later in detail based on the information included in the data transmitted from the client nodes 11 to 14 or attached, and distributes the distributed node. Via the system network 40, a request to write or update data received from the client nodes 11 to 14 is output to the nodes 51 to 54 as storage locations.
- FIG. 2 is a block diagram showing the configuration of the distributed access node 30 of the first embodiment shown in FIG.
- a storage state determination unit 3001 a node area information holding unit 3002, an area identification unit 3003, a node identification unit 3004, a client monitoring unit 3005, a node area information management unit 3006, and a CPU (Central A configuration of a distributed access node 30 including a processing unit (3070), a RAM (Random Access Memory) 3071, an HDD (Hard Disk Drive) 3072, and an input / output unit 3073 is shown.
- a processing unit (3070)
- RAM Random Access Memory
- HDD Hard Disk Drive
- the storage state determination unit 3001 exceeds the predetermined threshold information range when the transmission data from the new client node is stored in the node specified by the node specification unit 3004 among the nodes 51 to 54. It is determined whether or not.
- the threshold information is described assuming that the number of client nodes that can be managed by the nodes 51 to 54, that is, the upper limit value of the number of transmission data from the client nodes is set. Further, in the following description, transmission data from a client node is counted as one, and the number of client nodes and the number of transmission data are treated as being the same.
- the threshold information may be set for each area as shown in FIG. 3, or may be set and managed for each node as shown in FIG.
- the node area information holding unit 3002 holds information on an area obtained by dividing an area into a certain size together with its position information. Hereinafter, this area is referred to as “area”, and information indicating the area is referred to as “area information”. As an example of the unit of the area, there is a unit divided into arbitrary sizes such as a number km in length and a number km in width. Alternatively, as another example, if a mobile phone network is used as a communication network, the coverage area of the base station of the mobile phone network can be set as one area.
- the node area information holding unit 3002 holds information in which the configuration of each area is associated with information on the number of client nodes that can be stored in each area. This information is called “area configuration information”.
- the node area information holding unit 3002 holds information associated with one or more area information for each of the nodes 51 to 54. This information is called “node area information”.
- FIG. 3 is an example of area configuration information including the configuration of areas A to D as shown in FIG. 6 and threshold information thereof.
- the value of the threshold information field in FIG. 3 indicates the number of client nodes that can be stored for each area.
- the threshold information of areas A and B is 3, respectively. This indicates that the upper limit value of the number of transmission data that can be stored in areas A and B is 3.
- the threshold information is determined based on the processing capability of the node associated with the area. For example, when a storage area such as a RAM held by each node and a processing capability such as a CPU speed can be grasped, an optimum threshold value can be set for each node based on such information.
- the threshold information may be set in advance based on information obtained at the time of system construction. Further, when a new node is added to the nodes 51 to 54, the node may be notified of the processing capability and the like, and may be set based on the notification contents.
- FIG. 4 is an example of node area information indicating the correspondence between areas A to D and nodes 51 to 54 as shown in FIG.
- two areas A and B are associated with one node 51, as indicated by the numbers in parentheses at the corners of each area of FIG.
- Areas C and D are associated with the nodes 53 and 54, respectively. Further, no area is associated with the node 52.
- the threshold information (number of storable data) in areas A and B is half that in areas C and D because the node 51 is in charge of two areas.
- the area specifying unit 3003 specifies an area to which the client node belongs based on information notified from the client nodes 11 to 14.
- An example of such information is GPS (Global Positioning System) information.
- the area specifying unit 3003 holds a table for specifying area information from these pieces of information notified from the client nodes 11 to 14, and the client node is referred to by referring to this table. It is possible to adopt a configuration for specifying the area to which the device belongs.
- the node identifying unit 3004 refers to the node area information described above and selects a node that is a storage destination of transmission data from a client node in a certain area.
- the client monitoring unit 3005 grasps which client node is managed for each node based on information that can uniquely identify the client node.
- This management information is referred to as “client management information”.
- FIG. 5 is an example of a table for managing client management information. Further, as an example of information that can uniquely identify such a client node, there is a client node ID in data to be notified to the data center. As another example, position information by GPS may be used.
- the node area information management unit 3006 is activated when data from a new client node is stored in a certain node, and the node area information management unit 3006 refers to the node area information and the client management information.
- the number of areas handled by one node is set to one or more, and the node area information is updated so that the total number of client nodes existing in geographically adjacent areas falls within the number that can be handled by one server.
- the CPU 3070 controls each part of distributed access node 30.
- the RAM 3071 and the HDD 3072 are used as storage destinations of programs executed by the CPU 3070 and information used by each unit of the distributed access node.
- the input / output unit 3073 provides communication means and the like.
- each unit (processing unit) of the distributed access node uses the RAM 3071, the HDD 3072, and the like described above for the CPU 3070. It can also be realized by a computer program that executes
- the node area information holding unit 3002 includes the area configuration information illustrated in FIG. It is assumed that the upper limit of the number of client nodes 54 that can be managed (the number of data that can be stored in the nodes 51 to 54) is set. In the following description, it is assumed that the number of client nodes that can be managed by the nodes 51 to 54 (the number of data that can be stored by the nodes 51 to 54) is six.
- FIG. 7 is a flowchart showing the operation (distributed placement algorithm) of the distributed access node according to the first embodiment of the present invention.
- the distributed access node 30 receives a data write request from a client node (step A1). This data write request is made by transmitting data including position information from the client node to the distributed access node 30.
- the distributed access node 30 identifies an area to which the client node belongs based on information included in or attached to data sent from the client node. Further, the distributed access node 30 specifies the node associated with the specified area using the node area information (step A2).
- the distributed access node 30 checks whether or not the data transmitted from the client node already exists in the identified node (step A3). Whether or not the confirmation includes an entry in the client management information (see FIG. 5) managed by the client monitoring unit 3005 having a combination of the client node ID of the client node and the identified node. Can be confirmed.
- the data transmitted from the client node already exists in the identified node No in Step A3
- the data on the node is overwritten and updated (Step A6).
- the distributed access node 30 added the transmission data from the client node to the identified node with reference to the threshold information In this case, it is determined whether or not the number of client nodes that can be managed by the node (the number of data that can be stored by the node) exceeds the upper limit (step A4). If it is determined that the number of transmission data does not exceed the upper limit value (Yes in step A4), the process proceeds to step A6, and new data is added to the node (step A6).
- step A4 if it is determined in step A4 that the number of client nodes that can be managed by the node (the number of data that can be stored by the node) exceeds the upper limit, the distributed access node 30 sends the client that has transmitted the data.
- the node is added, the relationship between the node and the area is reviewed, and the node area information is updated (step A5).
- a node that manages transmission data from the client node is determined, and new data is added to the node (step A6).
- the client monitoring unit 3005 adds an entry that associates the client node with the node in the client management information.
- nodes are allocated according to the area of the client node that has made a new data write request. For example, as shown in FIG. 6, in a state where the nodes 51 are associated with the areas A and B and the nodes 53 and 54 are associated with the areas C and D, respectively, data is newly written from the client node located in the area A.
- the node 52 is a node in an idle state (no allocation area) in the node area information of FIG.
- FIG. 8 is a diagram showing the number of information (number of client nodes) in each area in a tree structure.
- the numbers in parentheses indicate the nodes associated with the area.
- there are three client nodes in each of the areas A and B and a total of six client nodes exist in the area A + B one level above.
- there are 6 client nodes in each of the areas C and D and a total of 12 client nodes exist in the area C + D on the upper layer.
- the node 51 is associated with the area A + B, and the nodes C and D are associated with the nodes 53 and 54, respectively.
- the number of client nodes that can be managed by the node 51 associated with area A + B is exceeded. Therefore, processing for assigning the node 52 to the area A (or area B) is performed.
- the node 52 is associated with the area B, and FIG. 6 is rewritten as shown in FIG. Further, by adding the node 52, the threshold value for the area A (upper limit value for the number of client nodes) and the threshold value for the area B (upper limit value for the number of client nodes) are each rewritten to 6.
- data that is frequently updated is distributed evenly, and transmission data from neighboring client nodes can be autonomously arranged so that high-speed search is possible. .
- the reason is that the search within the nodes 51 to 54 is stopped, and the search across the nodes is eliminated as much as possible, so that the search through the network is reduced and the network bandwidth and the overhead of using the network are eliminated.
- data from client nodes located in close proximity is arranged in the same node as much as possible, there is an advantage that data of nearby client nodes can be searched at high speed.
- the distributed access node 30 can perform load distribution without acquiring the access information transmitted from the client node from each of the nodes 51 to 54. .
- the number of client nodes that can be managed by each of the nodes 51 to 54 has been described as six. However, the number of client nodes that can be managed by each of the nodes 51 to 54 has the same value. There is no need, and it can be set in detail based on the storage area such as RAM possessed by each of the nodes 51 to 54 and the processing capability such as CPU speed.
- the second embodiment of the present invention is configured to delete a client management information entry satisfying a predetermined deletion condition in the first embodiment of the present invention. Will be described in detail.
- FIG. 11 is a block diagram showing the configuration of the distributed access node 30A according to the second embodiment of this invention.
- the difference from the distributed access node 30 of the first embodiment shown in FIG. 2 is that a timer unit 3010 and a data deletion unit 3011 are added. Others are substantially the same as those in the first embodiment, and therefore the following description will be focused on the differences.
- the timer unit 3010 is means for issuing a notification to the data deleting unit 3011 at a preset time interval.
- the data deletion unit 3011 When the data deletion unit 3011 receives a notification from the timer unit 3010, the data deletion unit 3011 performs an operation of deleting an entry satisfying a predetermined deletion condition in the client management information held in the client monitoring unit.
- the client monitoring unit 3005a each time the client monitoring unit 3005a receives the transmission data from the client nodes 11 to 14 and writes it to the node, the client monitoring unit 3005a in this embodiment performs an operation of updating the latest writing time of the client management information as shown in FIG. .
- the basic operation of the distributed access node 30A is the same as that of the first embodiment of the present invention (see FIG. 7).
- the following operation is performed by adding the timer unit 3010 and the data deletion unit 3011 in parallel with the operation of FIG.
- FIG. 13 is a flowchart showing an operation executed by the timer unit 3010 at a predetermined time interval in the distributed access node 30A according to the second embodiment of this invention.
- the timer unit 3010 issues a notification to the data deletion unit 3011 at every preset time interval (step B1).
- the data deletion unit 3011 reads one entry from the client management information held in the client monitoring unit 3005a, and the difference between the current time and the latest writing time exceeds the difference within a predetermined range. It is confirmed whether there is (step B2).
- the data deletion unit 3011 performs the above determination processing for all entries of the client management information held in the client monitoring unit 3005a (No in step B4).
- the series of processing ends (Yes in step B4).
- the client management information held in the client monitoring unit 3005a is not operating. It becomes possible to delete the entry for the node.
- the client management information aging process is performed on the distributed access node 30A side.
- a similar mechanism may be added to the data holding mechanism on the nodes 51 to 54 side. By doing so, it becomes possible to cause each of the nodes 51 to 54 to delete transmission data from unnecessary client nodes autonomously.
- the deletion determination time may be changed for each client node or area.
- FIG. 14 is a block diagram showing the configuration of the distributed access node 30B according to the third exemplary embodiment of the present invention.
- the difference from the distributed access node 30A of the second embodiment shown in FIG. 11 is that a node management unit 3020 is added.
- Others are substantially the same as those of the second embodiment, and therefore the following description will be focused on the differences.
- the node area information holding unit 3002a holds node area information to which a field indicating the power state of each node is added.
- the node management unit 3020 determines whether or not to perform power-off control or operation mode change (power saving mode, sleep mode, etc.) for a node whose number of managed client nodes has decreased, and Issues a power off command or a command to change the operation mode (power saving mode, sleep mode, etc.).
- the node management unit 3020 receives a response from the node that issued the power-off command, the node management unit 3020 adds that the power is off or sleeps to the node area information shown in FIG.
- the basic operation of the distributed access node 30B is the same as that of the first and second embodiments of the present invention (see FIGS. 7 and 13).
- the following operation is performed by adding the node management unit 3020.
- FIG. 16 is a flowchart showing an operation executed by the timer unit 3010 at predetermined time intervals in the distributed access node 30B according to the third exemplary embodiment of the present invention.
- Steps C1 to C4 in FIG. 16 correspond to steps B1 to B4 in FIG. 13 for explaining the operation of the second embodiment.
- the node area information management unit 3006 updates the client management information (step C5). Specifically, the correspondence between the nodes and the areas is reviewed so as to minimize the number of nodes covering the areas A to D based on the result of the entry deletion in step C3.
- the above review can be performed in the following procedure.
- the number of areas handled by one node is one or more. If the number of areas is 0, it is determined that the node is unnecessary. Also. If the total number of client nodes existing in geographically adjacent areas is less than or equal to the number of client nodes that can be managed by one node (here, 6), these geographically adjacent areas A group is managed by one node. Adjacent areas can be grouped up until the number of client nodes located in these areas reaches the upper limit of the number of client nodes that can be managed by one node. For example, if the total number of client nodes located in the areas A to D in FIG. 6 is 6, one area can cover the areas A to D, and the other nodes are unnecessary.
- step C6 As a result of the update of the client management information in step C5 in FIG. 16, it is confirmed whether there is an unnecessary node (a node in which the number of associated client nodes is 0) (step C6). For example, power-off control of the unnecessary node or change of the operation mode (power saving mode, sleep mode, etc.) is performed (step C7).
- the node management unit 3020 may again turn on the power or change the operation mode for a node that has been subjected to power-off control or operation mode change (power saving mode, sleep mode, etc.). .
- the node area information management unit 3006 may determine whether or not there is a surplus in the number of data that can be stored in each node. For example, as shown in FIG. 6, when all three nodes can use all the data that can be stored, the node area information management unit 3006 causes the node management unit 3020 to issue a power-on request, For example, the power of the node 52 may be turned on.
- the present invention is not limited to the above-described embodiments, and further modifications, replacements, and replacements may be made without departing from the basic technical idea of the present invention. Adjustments can be made.
- the number of data that each node can store has been described as six.
- the threshold information is not limited to the above example. For example, you may change according to the performance of each node, and you may change dynamically according to a time slot
- the number of client nodes and the number of transmission data are the same. However, in each node, when it is necessary to hold past data transmitted from the client node, The number of transmission data (upper limit value) is larger than the number of client nodes.
- the upper limit value (threshold value information) is determined based on the number of client nodes and the number of transmission data.
- the upper limit value may be determined based on the data amount such as the number of bytes.
- the upper limit value (threshold value information) is set and the node is added when the upper limit value is exceeded, but when the lower limit value (threshold information) is set and falls below this, You may make it accommodate the client node managed by the node below the said lower limit (threshold value information) in the node with surplus power among adjacent nodes.
- the comparison with the threshold information is performed when data is received from the new client node.
- the comparison with the threshold information is performed when data is received from the new client node. Instead, it may be confirmed whether transmission data from the client node is within the range of the threshold information at a predetermined trigger such as a predetermined time interval.
- the distributed access node may have a function of detecting that the client node has moved to the adjacent area, and the client management information may be updated each time. In this way, the data from the client node can be stored in the moved node without waiting for the reception of the data from the client node.
- the distributed access node may be provided with a function of detecting a change in the operating state of the client node, and the entry of the client node that has become inactive may be deleted from the client management information. In this way, the number of client management information entries can be reduced and the number of nodes assigned to these nodes can be reduced.
- the position of each client node is specified based on data included in or attached to data transmitted from the client node. It is also possible to adopt a configuration for acquiring For example, when the client node is a mobile phone terminal or the like, the position information can be received from the position management server or the like.
- the area specifying unit specifies an area where the client node is located based on information included in data transmitted from a client node or information attached to the data.
- the distributed access node further includes: Storage state for confirming whether transmission data from the client node stored in the individual node falls within a threshold information range determined for each node or each area at a predetermined opportunity A determination unit; A node area information management unit that updates the node area information so that transmission data from the client node stored in each node falls within the threshold information range based on the confirmation result of the storage state determination unit; , A data placement system comprising: [Fourth form] In the third data placement system, A data arrangement system in which a value corresponding to the processing capability of each node is set as the threshold information.
- the threshold information is a data placement system calculated based on a storage area or CPU speed held by each node.
- the threshold information is a data arrangement system calculated based on information included in a notification at the time of network connection from the individual nodes at the time of system construction.
- the distributed access node further includes: A client monitoring unit that holds client management information in which the identified node is associated with the client node; A data placement system in which confirmation by the threshold information is omitted when a data transmission source is a client node registered in the client management information.
- the distributed access node further includes a function of detecting that the client node has moved to an adjacent area, and updating the client management information according to the detection result, A data arrangement system for changing a storage destination of transmission data from the client node without waiting for reception of transmission data from the client node.
- the distributed access node further includes data having a function of detecting a change in the operating state of the client node and deleting an entry of the client node that has become inactive from the client management information according to the detection result. Placement system.
- the distributed access node further includes: A data arrangement system comprising a data deletion unit that deletes an entry satisfying a predetermined deletion condition among the entries of the client management information at a predetermined time interval.
- a data arrangement system comprising a data deletion unit that deletes an entry satisfying a predetermined deletion condition among the entries of the client management information at a predetermined time interval.
- the node area information management unit is a data arrangement system that associates a plurality of areas with one node within a range that falls within the range of the threshold information.
- the distributed access node further includes: A data arrangement system comprising a node management unit that transitions a node that does not store transmission data from the client node to a power-off state or a hibernation state.
- the node area information management unit determines whether or not each node has a capacity to store data when the node area information is updated, and when the remaining capacity is insufficient, Request to add a node
- the node management unit is a data arrangement system in which a node that does not store transmission data from the client node is changed to a power-on state or an operation state in response to the request.
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Information Transfer Between Computers (AREA)
- Computer And Data Communications (AREA)
Abstract
Description
本発明は、日本国特許出願:特願2011-193634号(2011年 9月 6日出願)の優先権主張に基づくものであり、同出願の全記載内容は引用をもって本書に組み込み記載されているものとする。
本発明は、データ配置システム、分散アクセスノード、データ配置方法およびプログラムに関し、特に、クライアントから送信されたデータを複数のノードに格納するデータ配置システム、分散アクセスノード、データ配置方法およびプログラムに関する。
続いて、本発明の第1の実施形態について図面を参照して詳細に説明する。図1は、本発明の第1の実施形態の全体構成を示すブロック図である。図1を参照すると、クライアントノード11~14と、ネットワーク20と、分散アクセスノード30と、分散システムネットワーク40と、サーバ、ストレージ等によって構成されたノード51~54とを含む構成が示されている。なお、図1の例では、それぞれ4つのクライアントノードとノードを示しているが、これらの数は本発明の理解を簡単にするために選択されたものであり、これ以上の数のクライアントノード、ノードで構成してもよい。
続いて、次に、本発明の第1の実施形態におけるクライアント管理情報のエントリのうち、所定の削除条件を満たしたものを削除するようにした本発明の第2の実施形態について図面を参照して詳細に説明する。
続いて、次に、本発明の第2の実施形態の分散アクセスノード30Aに、各ノードの動作状態(電源オンオフ、動作モードの変更等)の制御機能を追加した本発明の第3の実施形態について図面を参照して詳細に説明する。
[第1の形態]
(上記第1の視点によるデータ配置システム参照)
[第2の形態]
第1の形態のデータ配置システムにおいて、
前記エリア特定部は、クライアントノードから送信されたデータに含まれる情報または前記データに添付された情報に基づいて前記クライアントノードの位置するエリアを特定するデータ配置システム。
[第3の形態]
第1または第2の形態のデータ配置システムにおいて、
前記分散アクセスノードは、さらに、
所定の契機で、前記個々のノードにおいて格納されている前記クライアントノードからの送信データが、個々のノードまたは個々のエリア毎に定められた閾値情報の範囲に収まっているか否かを確認する格納状態判定部と、
前記格納状態判定部の確認結果に基づいて、個々のノードにおいて格納されている前記クライアントノードからの送信データが前記閾値情報の範囲に収まるよう、前記ノードエリア情報を更新するノードエリア情報管理部と、
を備えるデータ配置システム。
[第4の形態]
第3のデータ配置システムにおいて、
前記閾値情報として、個々のノードの処理能力に応じた値が設定されるデータ配置システム。
[第5の形態]
第4の形態のデータ配置システムにおいて、
前記閾値情報は、個々のノードが保有する記憶領域またはCPU速度に基づいて算出されるデータ配置システム。
[第6の形態]
第4または第5の形態のデータ配置システムにおいて、
前記閾値情報は、システム構築時における前記個々のノードからのネットワーク接続時の通知に含まれる情報に基づいて算出されるデータ配置システム。
[第7の形態]
第3から第6いずれか一の形態のデータ配置システムにおいて、
前記分散アクセスノードは、さらに、
前記特定されたノードと、前記クライアントノードとを対応付けたクライアント管理情報を保持するクライアント監視部を備え、
データの送信元が、前記クライアント管理情報に登録されているクライアントノードである場合、前記閾値情報による確認を省略するデータ配置システム。
[第8の形態]
第7の形態のデータ配置システムにおいて、
前記分散アクセスノードは、さらに、前記クライアントノードが隣のエリアに移動したことを検出し、前記検出結果に応じて、前記クライアント管理情報を更新する機能を備え、
前記クライアントノードからの送信データの受信を待たずに、前記クライアントノードからの送信データの格納先を変更するデータ配置システム。
[第9の形態]
第7または第8の形態のデータ配置システムにおいて、
前記分散アクセスノードは、さらに、前記クライアントノードの動作状態の変化を検出し、前記検出結果に応じて、前記クライアント管理情報から、非動作状態となったクライアントノードのエントリを削除する機能を備えるデータ配置システム。
[第10の形態]
第7から第9いずれか一の形態のデータ配置システムにおいて、
前記分散アクセスノードは、さらに、
所定の時間間隔で、前記クライアント管理情報のエントリのうち、所定の削除条件を満たしたエントリを削除するデータ削除部を備えるデータ配置システム。
[第11の形態]
第3から第10いずれか一の形態のデータ配置システムにおいて、
前記ノードエリア情報管理部は、前記閾値情報の範囲に収まる範囲で、一のノードに複数のエリアを関連付けるデータ配置システム。
[第12の形態]
第1から第10いずれか一の形態のデータ配置システムにおいて、
前記分散アクセスノードは、さらに、
前記クライアントノードからの送信データを格納していないノードを、電源オフ状態または休止状態に遷移させるノード管理部を備えるデータ配置システム。
[第13の形態]
第12の形態のデータ配置システムにおいて、
前記ノードエリア情報管理部は、前記ノードエリア情報の更新の際に、前記各ノードにデータを格納する余力があるか否かを判断し、前記余力が不足する場合、前記ノード管理部に対し、ノードの追加を要求し、
前記ノード管理部は、前記要求に応じて、前記クライアントノードからの送信データを格納していないノードを、電源オン状態または稼動状態に遷移させるデータ配置システム。
[第14の形態]
(上記第2の視点による分散アクセスノード参照)
[第15の形態]
(上記第3の視点によるデータ配置方法参照)
[第16の形態]
(上記第4の視点によるプログラム参照)
なお、上記した第14~第16の形態は、第2~第13の形態と同様に、派生する形態に展開することが可能である。
なお、前述の特許文献の開示を、本書に引用をもって繰り込むものとする。
本発明の全開示(請求の範囲および図面を含む)の枠内において、さらにその基本的技術思想に基づいて、実施形態ないし実施例の変更・調整が可能である。また、本発明の請求の範囲および図面の枠内において種々の開示要素(各請求項の各要素、各実施例の各要素、各図面の各要素等を含む)の多様な組み合わせないし選択が可能である。すなわち、本発明は、請求の範囲を含む全開示、技術的思想にしたがって当業者であればなし得るであろう各種変形、修正を含むことは勿論である。特に、本書に記載した数値範囲については、当該範囲内に含まれる任意の数値ないし小範囲が、別段の記載のない場合でも具体的に記載されているものと解釈されるべきである。
20 ネットワーク
30、30A、30B 分散アクセスノード
40 分散システムネットワーク
51、52、53、54 ノード
3001 格納状態判定部
3002、3002a ノードエリア情報保持部
3003 エリア特定部
3004 ノード特定部
3005、3005a クライアント監視部
3006 ノードエリア情報管理部
3010 タイマー部
3011 データ削除部
3020 ノード管理部
3070 CPU
3071 RAM
3072 HDD
3073 入出力部
Claims (10)
- クライアントノードから所定の頻度で送信されるデータの格納先となる複数のノードと、
前記複数のノードと前記クライアントノードとの間に配置された分散アクセスノードと、を含み、
前記分散アクセスノードは、
前記各ノードに対し、それぞれ1つ以上のエリア情報を関連付けたノードエリア情報を保持するノードエリア情報保持部と、
前記クライアントノードの位置するエリアを特定するエリア特定部と、
前記クライアントノードからの送信データの格納先として、前記特定されたエリアに関連付けられたノードを選択するノード特定部と、
を備えること、を特徴とするデータ配置システム。 - 前記分散アクセスノードは、さらに、
所定の契機で、前記個々のノードにおいて格納されている前記クライアントノードからの送信データが、個々のノードまたは個々のエリア毎に定められた閾値情報が示す閾値の範囲に収まっているか否かを確認する格納状態判定部と、
前記格納状態判定部の確認結果に基づいて、個々のノードにおいて格納されている前記クライアントノードからの送信データが前記閾値の範囲に収まるよう、前記ノードエリア情報を更新するノードエリア情報管理部と、
を備える請求項1のデータ配置システム。 - 前記分散アクセスノードは、さらに、
前記特定されたノードと、前記クライアントノードとを対応付けたクライアント管理情報を保持するクライアント監視部を備え、
データの送信元が、前記クライアント管理情報に登録されているクライアントノードである場合、前記閾値による確認を省略する請求項2のデータ配置システム。 - 前記分散アクセスノードは、さらに、
所定の時間間隔で、前記クライアント管理情報のエントリのうち、所定の削除条件を満たしたエントリを削除するデータ削除部を備える請求項3のデータ配置システム。 - 前記ノードエリア情報管理部は、前記閾値の範囲に収まる範囲で、一のノードに複数のエリアを関連付ける請求項3または4のデータ配置システム。
- 前記分散アクセスノードは、さらに、
前記クライアントノードからの送信データを格納していないノードを、電源オフ状態または休止状態に遷移させるノード管理部を備える請求項1から5いずれか一のデータ配置システム。 - クライアントノードから送信されたデータの格納先となる複数のノードと前記クライアントノードとの間に配置され、
前記各ノードに対し、それぞれ1つ以上のエリア情報を関連付けたノードエリア情報を保持するノードエリア情報保持部と、
前記クライアントノードの位置するエリアを特定するエリア特定部と、
前記クライアントノードからの送信データの格納先として、前記特定されたエリアに関連付けられたノードを選択するノード特定部と、
を備える分散アクセスノード。 - 前記ノードエリア情報の各エントリには、個々のノード又は個々のエリアが、前記クライアントノードからの送信データを格納可能な閾値情報が含まれており、
所定の契機で、前記個々のノードにおいて格納されている前記クライアントノードからの送信データが前記閾値情報が示す閾値の範囲に収まっているか否かを確認する格納状態判定部と、
前記格納状態判定部の確認結果に基づいて、個々のノードにおいて格納されている前記クライアントノードからの送信データが前記閾値の範囲に収まるよう、前記ノードエリア情報を更新するノードエリア情報管理部と、
を備える請求項7の分散アクセスノード。 - クライアントノードから送られるデータの格納先を、複数のノードの中から決定するデータ配置方法であって、
前記複数のノードと前記クライアントノードとの間に配置された分散アクセスノードが、
クライアントノードの位置するエリアを特定するステップと、
前記各ノードに対し、それぞれ1つ以上のエリア情報を関連付けたノードエリア情報を参照して、前記クライアントノードからの送信データの格納先として、前記特定されたエリアに関連付けられたノードを選択するステップと、
を含むデータ配置方法。 - 前記ノードエリア情報保持部には、個々のノードが、前記クライアントノードからの送信データを格納可能な閾値情報が含まれており、
前記分散アクセスノードが、所定の契機で、前記個々のノードにおいて格納されている前記クライアントノードからの送信データが前記閾値情報が示す閾値の範囲に収まっているか否かを確認するステップと、
前記確認結果に基づいて、個々のノードにおいて格納されている前記クライアントノードからの送信データが前記閾値の範囲に収まるよう、前記ノードエリア情報を更新するステップと、
を含む請求項9のデータ配置方法。
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US14/343,037 US20140201268A1 (en) | 2011-09-06 | 2012-09-05 | Data placement system, distributed access node, data placement method and program |
JP2013532619A JP6015660B2 (ja) | 2011-09-06 | 2012-09-05 | データ配置システム、分散アクセスノード、データ配置方法およびプログラム |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2011193634 | 2011-09-06 | ||
JP2011-193634 | 2011-09-06 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2013035719A1 true WO2013035719A1 (ja) | 2013-03-14 |
Family
ID=47832164
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2012/072559 WO2013035719A1 (ja) | 2011-09-06 | 2012-09-05 | データ配置システム、分散アクセスノード、データ配置方法およびプログラム |
Country Status (3)
Country | Link |
---|---|
US (1) | US20140201268A1 (ja) |
JP (1) | JP6015660B2 (ja) |
WO (1) | WO2013035719A1 (ja) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2015210534A (ja) * | 2014-04-23 | 2015-11-24 | 富士通フロンテック株式会社 | メンテナンスサーバ、テーブル更新プログラム及びテーブル更新方法 |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP6155861B2 (ja) * | 2013-06-06 | 2017-07-05 | 富士通株式会社 | データ管理方法、データ管理プログラム、データ管理システム及びデータ管理装置 |
US10983991B1 (en) * | 2020-09-29 | 2021-04-20 | Atlassian Pty Ltd. | Project management system data storage |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2004054316A (ja) * | 2002-07-16 | 2004-02-19 | Matsushita Electric Ind Co Ltd | サーバ装置および情報出力装置 |
JP2004072265A (ja) * | 2002-08-02 | 2004-03-04 | Kddi Corp | 機能分散システムおよび負荷分散方法、コンピュータプログラム |
JP2006301749A (ja) * | 2005-04-18 | 2006-11-02 | Hitachi Information Technology Co Ltd | サーバ装置 |
JP2009536402A (ja) * | 2006-05-05 | 2009-10-08 | マイクロソフト コーポレーション | デプロイメントユニットによる多数のユーザのグローバルプロビジョニング |
JP2009237935A (ja) * | 2008-03-27 | 2009-10-15 | Nec Corp | 負荷分散装置及び方法とプログラム |
JP2011086066A (ja) * | 2009-10-14 | 2011-04-28 | Hitachi Automotive Systems Ltd | 振り分け処理装置、計算機システム及びリクエスト振り分け方法 |
Family Cites Families (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2004126716A (ja) * | 2002-09-30 | 2004-04-22 | Fujitsu Ltd | 広域分散ストレージシステムを利用したデータ格納方法、その方法をコンピュータに実現させるプログラム、記録媒体、及び広域分散ストレージシステムにおける制御装置 |
US20060272031A1 (en) * | 2005-05-24 | 2006-11-30 | Napster Llc | System and method for unlimited licensing to a fixed number of devices |
JP4707525B2 (ja) * | 2005-10-11 | 2011-06-22 | 株式会社リコー | 分散情報管理システム、情報収集サーバ、情報取得サーバ、分散情報管理方法及び分散情報管理プログラム |
JP2008112293A (ja) * | 2006-10-30 | 2008-05-15 | Hitachi Ltd | 管理計算機及び電源制御方法並びに計算機システム |
JP2009009194A (ja) * | 2007-06-26 | 2009-01-15 | Hitachi Ltd | 消費電力を削減する機能を備えたストレージシステム |
WO2009032710A2 (en) * | 2007-08-29 | 2009-03-12 | Nirvanix, Inc. | Filing system and method for data files stored in a distributed communications network |
JP2009080674A (ja) * | 2007-09-26 | 2009-04-16 | Hitachi Ltd | 制御装置、アクセス制御方法、及びストレージノード |
CN101771715B (zh) * | 2008-12-26 | 2014-04-16 | 华为技术有限公司 | 分布式网络构建存储的方法、装置和系统 |
CN106407044B (zh) * | 2010-12-14 | 2019-09-24 | 株式会社日立制作所 | 信息处理系统中的故障恢复方法和信息处理系统 |
-
2012
- 2012-09-05 JP JP2013532619A patent/JP6015660B2/ja active Active
- 2012-09-05 WO PCT/JP2012/072559 patent/WO2013035719A1/ja active Application Filing
- 2012-09-05 US US14/343,037 patent/US20140201268A1/en not_active Abandoned
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2004054316A (ja) * | 2002-07-16 | 2004-02-19 | Matsushita Electric Ind Co Ltd | サーバ装置および情報出力装置 |
JP2004072265A (ja) * | 2002-08-02 | 2004-03-04 | Kddi Corp | 機能分散システムおよび負荷分散方法、コンピュータプログラム |
JP2006301749A (ja) * | 2005-04-18 | 2006-11-02 | Hitachi Information Technology Co Ltd | サーバ装置 |
JP2009536402A (ja) * | 2006-05-05 | 2009-10-08 | マイクロソフト コーポレーション | デプロイメントユニットによる多数のユーザのグローバルプロビジョニング |
JP2009237935A (ja) * | 2008-03-27 | 2009-10-15 | Nec Corp | 負荷分散装置及び方法とプログラム |
JP2011086066A (ja) * | 2009-10-14 | 2011-04-28 | Hitachi Automotive Systems Ltd | 振り分け処理装置、計算機システム及びリクエスト振り分け方法 |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2015210534A (ja) * | 2014-04-23 | 2015-11-24 | 富士通フロンテック株式会社 | メンテナンスサーバ、テーブル更新プログラム及びテーブル更新方法 |
Also Published As
Publication number | Publication date |
---|---|
JPWO2013035719A1 (ja) | 2015-03-23 |
US20140201268A1 (en) | 2014-07-17 |
JP6015660B2 (ja) | 2016-10-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP4265245B2 (ja) | 計算機システム | |
US10254998B2 (en) | Coordinated garbage collection of flash devices in a distributed storage system | |
EP3334123B1 (en) | Content distribution method and system | |
KR101502896B1 (ko) | 맵 리듀스를 이용한 분산 메모리 클러스터 제어 장치 및 방법 | |
KR101589801B1 (ko) | 피어 투 피어 데이터 마이그레이션 | |
CN102724314B (zh) | 一种基于元数据管理的分布式缓存客户端 | |
CN103366022B (zh) | 信息处理系统及其处理方法 | |
CN104536702A (zh) | 一种存储阵列系统及数据写请求处理方法 | |
CN110427386A (zh) | 数据处理方法、装置及计算机存储介质 | |
CN104506462A (zh) | 一种分布式交换机中mac地址管理方法及设备 | |
JP2007226646A (ja) | ストレージシステム | |
JP6015660B2 (ja) | データ配置システム、分散アクセスノード、データ配置方法およびプログラム | |
CN105095495A (zh) | 一种分布式文件系统缓存管理方法和系统 | |
CN112019577B (zh) | 专属云存储实现方法、装置、计算设备及计算机存储介质 | |
CN107704507A (zh) | 数据库处理方法和设备 | |
JP2003296153A (ja) | ストレージシステムおよびそのためのプログラム | |
WO2012136828A1 (en) | Multi-user cache system | |
JP5351839B2 (ja) | オーダ処理方法、プログラムおよびネットワークシステム | |
CN103078926B (zh) | 分布式存储系统的文件访问方法和装置以及系统 | |
CN104468674B (zh) | 数据迁移方法及装置 | |
KR101771178B1 (ko) | 인메모리 캐시를 관리하는 방법 | |
KR101771183B1 (ko) | 인메모리 캐시를 관리하는 방법 | |
CN102170476A (zh) | 一种基于云节点自主学习的云计算方法和装置 | |
WO2012056835A1 (ja) | データ記憶制御システム、データ記憶制御方法およびデータ記憶制御プログラム | |
JP5530878B2 (ja) | 分散システムにおけるデータレプリケーション管理方法 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 12829529 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2013532619 Country of ref document: JP Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 14343037 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 12829529 Country of ref document: EP Kind code of ref document: A1 |