JP2008304982A - Information management method and information processor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an information management method, capable of maintaining appropriate redundancy for data holding in a network system in which data is distributed and shared among a plurality of nodes, to efficiently use distributed and shared data while preventing waste of resources by holding of data more than necessary and minimizing occurrence of a failure such that data cannot be acquired due to insufficient redundancy, and an information processor as the node. <P>SOLUTION: Held data is monitored to detect change in redundancy. The node holding the data is changed so that the redundancy is not less than or not more than a predetermined value. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an information management method in a network in which information is distributed and shared among a plurality of nodes, and an information processing apparatus as a node constituting the network.

  In recent years, a network having a communication form in which data is freely transmitted and received between arbitrary nodes constituting the network has been actively used.

  Conventionally, there is a server that plays the role of a host at the center, and each terminal as a client accesses the host server, and if it is necessary to exchange between terminals, the host server mediates A centralized network was the mainstream.

  In contrast, so-called distributed processing networks are gradually appearing. In order to realize a mechanism for distributing the information to be stored and for distributing the processing, the communication function must also be distributed. That is, data communication must be freely performed between the nodes constituting the network.

  A typical form is a form of a communication network called P2P (Peer to Peer). P2P is a network usage mode in which information is directly exchanged between an unspecified number of nodes, and there are two types, technically requiring a central server, and carrying data in a bucket relay manner. Even when a central server is required, the central server only provides a file search database and manages connection of nodes, and exchange of data itself is performed by direct connection between nodes.

  A technique for efficiently achieving such a distributed processing network form has been studied, and a system has been formed in which data is distributed and held among arbitrary nodes and data is transmitted and received between them. As a result, the degree of freedom as a form of use of the network system has improved, and the user has gained great convenience.

  For example, in the P2P network system described above, data can be used easily even if data is not held locally by sharing data among a number of nodes and transmitting / receiving data to / from each other.

  On the other hand, however, distributing and holding data may be burdensome in terms of efficiency compared to centralized data management. In addition, since data is distributed and held, reliability is required for each node.

  For example, if the data is distributed and held, the larger the network scale, the more widely the data can be distributed, and the amount of data that can be handled increases. In addition, reliability is required for all nodes to function normally.

  In order to collect necessary data from a wide network, the network and each node are burdened appropriately, and there is a possibility that a failure such as waiting or inability to receive may occur. In addition, if a failure such as a connection stop occurs in some nodes, there is a risk that necessary data cannot be acquired.

  How to reduce the risks associated with such distributed sharing of data will be an important issue in the future. In distributing and sharing data, in order to be able to acquire data even if a failure occurs in some nodes, the data is retained with redundancy, or the connection is efficiently organized according to the reliability of the node, Techniques have been developed to improve communication efficiency and reliability for data acquisition (see Patent Document 1).

  For example, Patent Document 1 proposes a technique for calculating reliability based on physical information such as CPU usage rate and memory usage rate of each node, and managing connection based on the reliability level of each node. The connection between the nodes is formed in a self-organized manner so that transmission / reception between the nodes is most efficiently performed.

  However, in this case as well, the connection management, that is, the topology of the network connection is only optimized according to the reliability of each node, and it is assumed that the actual data is held securely.

Otherwise, if a specific node temporarily fails or is suddenly disconnected from the network, it may be difficult to acquire data held by that node.
JP 2005-252596 A

  As described above, in order to ensure the retention and acquisition of data, there is a need for a network system that distributes and shares data so as to have redundancy among a plurality of nodes. The form can be assumed in various ways, but if there is too much redundancy in consideration of a failure etc., it means that excessive resources are wasted, and it is necessary to set an appropriate redundancy. .

  However, particularly in the network form such as the above-described P2P, the node connection form is easily changed, and the node is disconnected or restored, and it is difficult to maintain even if an appropriate redundancy is set.

  When the redundancy of data, such as a node disconnection, decreases, it may become impossible to access the data. On the other hand, if the redundancy is increased, resources will be used excessively, such as when the disconnected node is restored, the redundancy further increases.

  An object of the present invention is to solve the above-mentioned problems and maintain appropriate redundancy for data retention in a network system in which data is distributed and shared among a plurality of nodes, and therefore retains data more than necessary. As a node management method and information that makes it possible to efficiently use distributed and shared data without wasting resources and conversely failing to acquire data due to lack of redundancy. An information processing apparatus is provided.

  In order to solve the above problems, the present invention has the following features.

  1. A method for managing information in a network system in which information is distributed and shared among a plurality of nodes, and the change in redundancy of each information based on a monitoring target data list for monitoring the redundancy of each information that is distributed and shared In accordance with a change in the redundancy of each information detected by the redundancy detection step, the information is retained so that the redundancy of each information is equal to or greater than a predetermined lower limit value. A redundancy control step for changing a node, and an information management method.

  2. In the redundancy control step, according to a change in the redundancy of each information detected by the redundancy detection step, the information holding node is set so that the redundancy of each information is not more than a predetermined upper limit value. 2. The information management method according to 1, wherein the information is changed.

  3. In the redundancy detection step, the node that holds the monitoring target data list makes an inquiry to each node that holds each piece of information to be monitored based on the monitoring target data list, thereby changing the number of nodes holding the information. 3. The information management method according to 1 or 2, wherein a change in redundancy is detected.

  4). In the redundancy control step, the node that has detected a change in the redundancy of each information in the redundancy detection step changes the holding node according to the change in the redundancy of each information, The information management method according to any one of 1 to 3, wherein the monitoring target data list is updated according to a change.

  5. The monitoring target data list includes information held by a node holding the monitoring target data list, information held by a node within a broadcast reach range, information held by a predetermined node, Information management according to any one of 1 to 4, characterized in that at least one of information owned by a determined user and information requested to be monitored is described as monitoring target data Method.

  6). The monitoring target data list is held in a plurality of nodes, and in the redundancy control step, the updated contents of the monitoring target data list are notified to each other between the plurality of nodes, and the monitoring target held by each node 6. The information management method according to any one of 1 to 5, wherein each data list is updated.

  7. 7. The information management method according to any one of 1 to 6, wherein redundancy is provided by using a RAID technique for the distributed sharing of the information.

  8). An information processing apparatus as a node in a network system in which information is distributed and shared among a plurality of nodes, the storage means holding a monitoring target data list for monitoring the redundancy of each piece of information that is distributed and shared, and the monitoring target Based on the data list, redundancy detection means for detecting a change in redundancy of each information, and the redundancy of each information according to the change in redundancy of each information detected by the redundancy detection means An information processing apparatus comprising: redundancy control means for changing a node having the information so that the value of the information becomes equal to or greater than a predetermined lower limit value.

  9. The redundancy control means sets the information holding node so that the redundancy of each information becomes a predetermined upper limit value or less according to a change in the redundancy of each information detected by the redundancy detection means. 9. The information processing apparatus according to 8, wherein the information processing apparatus is changed.

  10. The redundancy detection means detects a change in redundancy from a change in the number of nodes holding the information by inquiring each node holding each information to be monitored based on the monitoring target data list. The information processing apparatus according to 8 or 9.

  11. The redundancy control means changes the holding node according to a change in redundancy of each information detected by the redundancy detection means, and updates the monitoring target data list according to the change of the holding node. 11. The information processing apparatus according to any one of 8 to 10, wherein:

  12 The monitoring target data list includes information held by a node holding the monitoring target data list, information held by a node within a broadcast reach range, information held by a predetermined node, 12. The information processing apparatus according to any one of 8 to 11, wherein at least one of information owned by a determined user and information requested to be monitored is described as monitoring target data .

  13. The redundancy control means notifies the update contents of the monitoring target data list to other nodes that hold the monitoring target data list, and shares the held monitoring target data list with each other. The information processing apparatus according to any one of 8 to 12.

  According to the information management method and the information processing apparatus as a node according to the present invention, in a network system in which information is distributed and shared among a plurality of nodes, the retained data is monitored to detect a change in redundancy. Also, the node that holds the data is changed so that the redundancy is greater than or less than a predetermined value.

  As a result, it is possible to maintain an appropriate redundancy for data retention. Accordingly, it is possible to efficiently use the distributed and shared data without causing unnecessary troubles such as not having unnecessary data and wasting resources, and conversely being unable to acquire data due to insufficient redundancy.

  Embodiments according to the present invention will be described below with reference to the drawings.

(Overall network configuration)
FIG. 1 is a diagram illustrating an example of an overall configuration of a network 1 configured by an information management method and an information processing apparatus according to the present embodiment. The overall configuration of the network 1 according to the embodiment of the present invention will be described with reference to FIG.

  As shown in FIG. 1, the network 1 includes a plurality of terminal devices 2 (21, 22,..., 2n), a switching hub 3, a router 4, and an authentication server 5 and other nodes (Local Area Network). ). These terminal devices 2 are connected to the switching hub 3 in a star shape by a twisted pair cable.

  A terminal device 2 as a node constituting a network is an information processing device according to the present invention, and executes data input / output processing with another device such as a personal computer, a workstation, or a printer. It is a device to do. Hereinafter, the term “node” simply refers to this terminal device, and a description will be given assuming that a personal computer as an information processing device is used.

  In the present embodiment, a communication network called P2P (Peer to Peer) is employed. P2P is a network usage mode in which information is directly exchanged between an unspecified number of nodes, and there are two types, technically requiring a central server, and carrying data in a bucket relay manner. Even when a central server is required, the central server only provides a file search database and manages connection of nodes, and exchange of data itself is performed by direct connection between nodes.

  In this embodiment, the central server is not used, and the connection topology of FIG. 3 will be described later. However, the nodes (information processing apparatuses) 2 associated in advance are directly connected and communicated. Other nodes are indirectly connected through directly connected nodes. The authentication server 5 is responsible only for management related to the certificate for authentication, and is not directly related to connection for communication. The router 4 is not directly involved in communication between nodes (information processing apparatuses).

  In P2P, since the nodes communicate directly with each other, the security of how to authenticate each other's validity and how to suppress the room for unauthorized entry is important. For this purpose, a digital certificate issued by the authentication server 5 is used.

  Hereinafter, from the above viewpoint, in the network according to the present embodiment, for the information distributed and shared among the nodes, the nodes 2 perform data communication between the nodes 2 while maintaining a certain redundancy, and efficiently use the information. A description will be given of a case where a change in the redundancy of each information is detected, the node of each information is changed according to the redundancy, and the redundancy of each information is controlled.

(Configuration of information processing device as a node)
FIG. 2 is a diagram illustrating an example of a hardware configuration of the node (information processing apparatus) 2.

  As shown in FIG. 2, the information processing apparatus 2 includes a CPU 20a, a RAM 20b, a ROM 20c, a hard disk 20d, a communication interface 20e, an image interface 20f, an input / output interface 20g, and other various circuits or devices.

  The communication interface 20e is, for example, a NIC (Network Interface Card), and is connected to one of the ports of the switching hub 3 via a twisted pair cable. The image interface 20f is connected to a monitor and sends a video signal for displaying a screen to the monitor.

  The input / output interface 20g is connected to an input device such as a keyboard or a mouse or an external storage device such as a CD-ROM drive. And the signal which shows the content of operation which the user performed with respect to the input device is input from an input device. Alternatively, data recorded on a recording medium such as a CD-ROM is read by an external storage device and input. Alternatively, data to be written to the recording medium is output to the external storage device.

  The hard disk 20d will be described later with reference to a functional block diagram (FIG. 5). However, the data holding unit 201, the monitoring target data list holding unit 202, the redundancy detecting unit 204, the redundancy operating unit 205, and the data operating unit 206 are described. A program and data for realizing functions such as a data reception unit 207, a data analysis unit 208, a data creation unit 209, a data transmission unit 210, other information holding unit 211, and other operation unit 212 are stored. These programs and data are read into the RAM 20b as necessary, and the programs are executed by the CPU 20a.

  Each node 2 is given a host name (machine name), an IP address, and a MAC address as identification information for identifying each other node 2. The host name can be freely assigned by the administrator of the network 1 or the like. The IP address is given according to the rules of the network 1. The MAC address is an address fixedly given to the communication interface 10e of the node 2. Note that another unique ID may be used instead of the MAC address.

  In this embodiment, it is assumed that host names such as “PC1”, “PC2”,... Are assigned to the nodes (information processing apparatuses) 21, 22,. Hereinafter, these nodes 2 may be described by host names.

(Node connection mode)
FIG. 3 is a diagram illustrating an example of a node connection form, that is, a logical topology of the information processing apparatus 2. A connection form of nodes (information processing apparatuses) will be described with reference to FIG.

  As shown in FIG. 3, the node 2 is assumed to be arranged in the virtual space. Then, as indicated by a dotted line, it is associated with at least one other node 2 in the vicinity in the virtual space. And, by these associations, all the nodes 2 are associated directly or indirectly with each other.

  Note that “directly related” means that they are connected by a single dotted line in FIG. 3 (for example, a relationship such as PC1 and PC2 or PC9 in FIG. 3), and “indirectly related”. Means that they are connected by two or more dotted lines and one or more nodes (for example, the relationship between PC1 and PC4 in FIG. 3). Node 2 sends data to other nodes 2 that are directly associated with it.

  FIG. 4 is a diagram showing an example of the connection table TL of the nodes 2 associated as shown in FIG. For each node 2, a list of information for connection with other nodes 2 capable of direct data transmission, that is, “directly related” is stored in a table.

  For example, connection tables TL1, TL2, TL6, TL7, TL8, and TL9 as shown in FIG. 4 are stored in PC1, PC2, PC6, PC7, PC8, and PC9 in FIG. 3, respectively.

(Processing functions of each part of the information processing device)
FIG. 5 is a block diagram illustrating an example of a functional configuration of the node (information processing apparatus) 2. The processing function of each unit of the node (information processing apparatus) 2 will be described with reference to FIG.

  First, processing functions other than processing related to redundancy control of distributed and shared data will be described.

  The other information holding unit 211 is attribute data indicating attributes of the node 2 or the user, digital certificate of the node 2 itself, data used by the operating system (OS) or application software, and the user created by the application software Data and other various data are stored as files.

  The digital certificate is issued by the authentication server 5 at the request of the node 2, held by the node 2, and used to authenticate each other during communication between the nodes 2.

  In addition, the other information holding unit 211 stores a connection table TL that shows a list of attributes such as the host name, IP address, and MAC address of another node 2 directly associated with the node 2 itself. . For example, FIG. 4 shows an example in which the connection tables TL1, TL2, TL6, TL7, TL8, and TL9 are stored in the connection table holding unit 201 of PC1, PC2, PC6, PC7, PC8, and PC9 in FIG. As stated above. The contents of these connection tables TL are created in advance by the administrator based on the association of each node 2.

  The other operation unit 212 manages the connection table TL held in the other information holding unit 211. In addition, processing such as storing data in the other information holding unit 211 or updating the stored data is performed. For example, the attribute data is updated every time the environment or setting content of the node 2 changes. Further, the authentication process of the other node 2 is performed based on the digital certificate transmitted from the other node 2.

  The other operation unit 212 performs data communication with the other nodes 2 of the network 1 via the data reception unit 207 and the data transmission unit 210 as necessary, and refers to the data of the other information holding unit 211 as necessary. Or update.

  The data receiving unit 207 performs control processing for performing data communication with other nodes 2. The data receiving unit 207 receives a packet necessary for the node 2 among the packets flowing through the network 1.

  The data analyzing unit 208 extracts necessary information from the received data received by the data receiving unit 207 and analyzes the contents thereof, thereby determining the type of the received data.

  The data creation unit 209 creates transmission data for transmission to another node 2 based on an instruction from the other operation unit 212.

  The data transmission unit 210 transmits the packetized transmission data generated by the transmission data creation unit 209 to the other nodes 2.

<Processing functions related to information redundancy detection and redundancy control>
Next, processing functions related to processing for detecting a change in redundancy and processing for controlling redundancy when sharing and sharing information will be described with reference to FIG.

  In the following description, all of the information distributed and shared by each node is referred to as data and described.

  The data operation unit 206 stores data (information) in the data holding unit 201, refers to it, and performs writing, reading, various update processes, and the like according to requests. It has processing functions related to the handling of distributed and shared data.

  Also, when using data that is divided and distributed and shared, a process of restoring the divided data is also performed.

  The data holding unit 201 holds data to be processed by the data operation unit 206, that is, distributed and shared data.

  The monitoring target data list holding unit 202 includes a monitoring target data list (hereinafter, abbreviated monitoring data list for short) of a list of monitoring target data whose redundancy is to be controlled and a list of nodes having the data among distributed and shared data. Also called). That is, it functions as a storage means. A specific example of the contents of the monitoring data list will be described later.

  The redundancy detection unit 204 monitors the monitoring target data according to the monitoring target data list, checks the possession node of the data, and detects a change in redundancy. That is, it functions as a redundancy detection means.

  In addition, the redundancy detection unit 204 performs an update process as necessary in order to reflect the detected change in redundancy, that is, the change in the owned node, in the monitoring target data list of the monitoring target data list holding unit 202. Details of the redundancy detection process will be described later.

  The redundancy operation unit 205 controls the redundancy within a predetermined upper limit or lower limit range based on the change in the redundancy detected by the redundancy detection unit 204 with respect to the monitoring target data. Perform the change process. That is, it functions as redundancy control means.

  Further, the redundancy operation unit 205 performs update processing as necessary to reflect the result of the control of redundancy, that is, the change of the owned node, in the monitoring target data list of the monitoring target data list holding unit 202. Details of the redundancy control process will be described later.

  The data operation unit 206, the redundancy operation unit 205, and the redundancy detection unit 204 perform data communication with other nodes 2 of the network 1 via the data reception unit 207 and the data transmission unit 210 as necessary, and also as necessary. Accordingly, the data in the data holding unit 201 and the monitoring target data list holding unit 202 is referred to or updated.

(Data sharing processing with redundancy)
In the network 1 according to the present embodiment, as described above, mutual authentication is performed between “associated” nodes (information processing apparatuses) defined by the connection table TL, and data is transmitted / received to / from each other.

  Based on this communication, the data is distributed and shared to each node, and the data shared and distributed as needed is retrieved and acquired. It can be used.

  However, such distributed storage of data may be burdensome in terms of efficiency compared to centralized data management. In addition, since data is distributed and held, reliability is required for each node.

  For example, in order to collect necessary data from a wide network, an appropriate burden is imposed on the network and each node, and there is a possibility that a failure such as waiting or inability to receive may occur. In addition, if a failure such as a connection stop occurs in some nodes, there is a risk that necessary data cannot be acquired.

  In the network 1 according to the present embodiment, in order to cope with such a risk, distributed sharing of data having redundancy such that the same data is held by a plurality of nodes is performed.

  Various known techniques have also been proposed for the method of retaining data for providing redundancy, but here, the RAID (Redundant Arrays of Independent Disks) technique is adopted. This is a mirroring technique known as RAID1.

<About redundancy in mirroring>
Simply put, the same data is shared by multiple nodes in the system, but how many nodes have the same data, that is, the nodes that can access and retrieve that data. The number is redundancy.

  For example, when two nodes respectively hold certain data, the redundancy is “2”. When three nodes are respectively held, the redundancy is “3”. Thereafter, the redundancy increases as the number of nodes held increases.

  If only one node is in possession and cannot be obtained other than accessing that one, the redundancy is “1”, that is, “not redundant”.

  The unit to be shared with redundancy may be a file unit or a bit unit.

  The redundancy in the case of dividing and sharing data will be described. A state in which two pieces of divided data are distributed and shared among a plurality of nodes and divided data corresponding to two pieces of data before division as a whole is distributed and shared is defined as redundancy “2”. Redundancy is “3” for three pieces of data before division. Thereafter, the redundancy increases as the number of nodes holding each divided data increases.

  Further, when the divided data is held in only one node, that is, only one piece of data before the division is held, the redundancy is “1” = “not redundant”.

  The above redundancy indicates the number of different nodes that can be accessed to acquire certain data. For example, if the redundancy of certain data is “2”, the node that cannot be accessed among the owned nodes This means that even if there is one, data can be acquired by accessing the other owned node.

  In the present embodiment, as will be described later, when distributing and sharing each data, an appropriate redundancy is set and distributed and shared, and the redundancy is monitored. Accordingly, control is performed so as to be within an appropriate fixed redundancy range.

  In other words, in the present embodiment, predetermined monitoring target data is described in the monitoring target data list together with the holding node and held. That is, based on the monitoring target data list, the holding node is checked for each piece of monitoring target data, and a change in the holding node is detected.

  If there is data for which a change in redundancy is detected, the data holding node is changed in order to control the data. That is, if the redundancy is too low, a holding node that newly holds the data is determined. On the other hand, if the redundancy is too high, the holding node from which the data is deleted is determined.

  The monitoring target data list is updated in a timely manner in accordance with the change in the redundancy and the change in the holding node.

  In the following, the detection of redundancy at the time of data sharing and the control processing for redundancy will be described.

(Redundancy detection and redundancy control processing)
FIG. 6 is a flowchart showing a typical processing flow of redundancy detection and redundancy control during data sharing. An example of redundancy detection processing and redundancy control processing as a whole will be described with reference to FIG.

<Redundancy detection processing>
First, in step S11, the redundancy detection unit 204 of the node that monitors the monitoring target data refers to the monitoring data list of the monitoring target data list holding unit 202 and starts monitoring.

  Nodes that monitor data are set in advance, and each has a monitoring data list. The number of nodes that monitor data is arbitrary, and may be one or all nodes.

  Considering the possibility of failure and the speed of change detection, it is desirable that the number of nodes to be monitored is larger. However, the detection load increases if there are too many nodes. The network size (number of nodes), the number of data to be monitored, the communication load, etc. may be appropriately determined.

  In the monitoring data list, data to be monitored and a list of owned nodes are described. In addition, it is desirable that a threshold value used as a criterion for redundancy, the current redundancy, information on owned nodes, and the like are also described.

  The monitoring data list is held by a node that monitors data, and is updated as appropriate. Specific examples of the contents of the monitoring data list are shown in FIGS.

Data to be monitored may be set in advance. For example, the following setting method can be considered.
1. Data held by the node that holds the monitored data list.
2. Data held by nodes within the broadcast range.
3. Data held by other predetermined nodes.
4). Data owned by a predetermined user.
5. Data requested to be monitored by the user.

  The above data is listed as a monitoring target in the monitoring data list. Alternatively, any combination of the above data is described in the monitoring data list. As a result, the degree of freedom of setting according to the importance of the data, the position state in the network, the owner, the user's request, etc. can be secured.

  As a method for performing monitoring, a method in which a node that performs monitoring polls each node at an appropriate timing and confirms the holding state of each data can be used.

  Appropriate timing for performing polling may be set to be performed periodically at a predetermined time or every period. That way, the polling interval will not be too wide and the frequency will not increase too much.

  Further, every time data is accessed, it may be detected and polling may be performed at that timing. Then, the polling frequency can be changed according to the frequency of data use.

  Here, polling is performed periodically. In step S11, the node that performs arbitrary monitoring refers to the monitoring data list, polls each node, and inquires about the holding status of the monitoring target data.

  In the next step S12, the redundancy detection unit 204 of the node that monitors the monitoring target data detects a change in redundancy for each monitoring target data.

  By confirming the holding status of each monitoring target data in each node in the previous step, it is possible to compare the description of the monitoring data list of the monitoring target data list holding unit 202 with the current state. For each data, the confirmed current redundancy, that is, the number of owned nodes is obtained, and it is detected whether it has changed from the description in the monitoring data list, that is, whether the number of retained nodes has increased or decreased.

  At this time, the description of the monitoring data list of the monitoring target data list holding unit 202 may be updated according to the current data holding status.

  In the next step S13, as a result of detection by the redundancy detection unit 204, it is determined whether or not there is a change in redundancy for each data, and the next processing is distributed.

  When there is a change in redundancy (step S13: YES), for example, the number of owned nodes has changed for some reason, such as when a specific node stops or returns, and the next step S14. Proceeding thereafter, the redundancy control process is executed.

  When there is no change in the redundancy (step S13: NO), it is a case where the number of owned nodes has not changed, the process returns to step S11, and the redundancy detection process is repeated from the next periodic polling execution.

  Accordingly, the above steps S11 to S13 function as a redundancy detection step.

<Redundancy control processing>
After step S14, it functions as a redundancy control process. In other words, the redundancy operation unit 205 changes the possession node so that the redundancy is within the range of the reference redundancy (upper limit, lower limit, or both) according to the change in the redundancy of each data. Do.

  If the redundancy is higher than the reference upper limit value, resources may be used unnecessarily for data retention, and access time may increase.

  In step S14, it is determined whether the redundancy is below a predetermined lower limit value S1. When the redundancy is lower than the lower limit S1 (step S14: YES), step S16a is executed to increase the redundancy that has been reduced too much. If it is equal to or greater than the lower limit value S1 (step S14: NO), step S15 is executed.

  In step S15, it is determined whether or not the redundancy exceeds a predetermined upper limit value S2. When the data redundancy exceeds the upper limit value S2 (step S15: YES), step S17a is executed to reduce the redundancy that has increased too much.

  When the data redundancy is equal to or lower than the upper limit value S2 (step S15: NO), this means that the redundancy is within the predetermined upper and lower limits, and the redundancy control process is not performed, and the process proceeds to step S11. Return and repeat the redundancy detection process from the next periodic polling execution.

  Here, whether or not it is within the range between the upper limit and the lower limit is determined by the above step S14 and step S15, but only one of the upper limit or the lower limit is executed by executing only one of step S14 and step S15. It doesn't matter. Further, the upper limit and the lower limit may coincide (S1 = S2).

  In step S16b subsequent to step S16a, processing is performed to increase the redundancy that has been reduced too much, that is, to change the number of nodes holding the data. To change the possession node, for example, the following procedure may be executed.

  An increase number Su of owned nodes necessary for setting the current redundancy to the lower limit S1 or more is calculated. An inquiry is made to a node other than the holding node as to whether or not the node can be held (step S16a).

  The Su nodes are determined as additional possession nodes with priority given to the nodes that responded quickly. An instruction is issued to the current holding node to send a copy of the data to the additional holding node (step S16b).

  In step S17b subsequent to step S17a, a process is performed in which the redundancy that has increased excessively is reduced, that is, the number of nodes holding the data is changed. To change the possession node, for example, the following procedure may be executed.

  The decrease number Sd of possession nodes necessary for setting the current redundancy to the upper limit S2 or less is calculated. The holding node is inquired whether the data can be deleted (step S17a). The node having the quick response is given priority, and the Sd node is determined as the holding node to be deleted. An instruction is issued to the holding node to be deleted, and the data is deleted (step S17b).

  In step S18, the contents described in the monitoring data list of the monitoring target data list holding unit 202 are updated for each data whose holding node has been changed.

  In step S19, the update contents of the monitoring data list are transmitted to the other monitoring nodes, that is, all the nodes holding the monitoring data list, and the monitoring data list held by each node is updated.

  This completes the redundancy control process. That is, the process returns to the first redundancy detection process and waits for the next polling execution.

(Processing example 1)
Hereinafter, processing example 1 of redundancy detection and redundancy control will be described with reference to FIGS.

  7 to 11 are diagrams showing the state of the network 1 in each step (redundancy detection processing and redundancy control processing) shown in FIG. In each processing example, the network 1 is composed of nodes PC1 to PC9, as shown in each figure.

<Initial state>
FIG. 7A shows the connection in the initial state. Among the nodes constituting the network 1, PC1, PC2, PC3, and PC9 are nodes that hold a monitoring data list, that is, nodes that perform monitoring.

  FIG. 7B shows the contents of the monitoring data list. In the monitoring data list, all data held in the nodes PC1 to PC9 is described as monitoring target data as monitoring target data. Further, it is assumed that the redundancy level is set such that the lower limit value S1 = 2 and the upper limit value S2 = 2, that is, the redundancy level is a constant value “2”.

  It is assumed that distributed sharing for providing data with redundancy uses RAID 1 mirroring technology. As shown in FIG. 7A, data C and data D are distributed and shared by PC1 and PC9 in the initial state. Data E and data F are distributed and shared by PC2 and PC3. Any data is the number of owned nodes (= 2) satisfying the redundancy “2” set as a reference. These states are described in the monitoring data list of FIG.

  In this state, the redundancy detection processing polling (step S11 in FIG. 6) is performed. If there is no change in the network shown in (a), the redundancy of each data does not change, so the monitoring shown in (b). There is no change in the contents of the data list.

<Disconnect node>
In FIG. 8A, the PC 9 is suddenly disconnected from the network, and then the redundancy detection process (step S11 to step S13 in FIG. 6) is performed by an arbitrary node (PC1) that has first polled. Indicates broken state.

  Since PC1 does not respond to polling to PC9 (step S11), it detects that the redundancy of data C and data D has changed from “2” to “1” (step S12).

  Due to this change, the redundancy of data C and data D deviates from the reference value “2” (step S13), so that the redundancy control process (after step S14) is started.

  FIG. 8B shows the description content of the monitoring data list of the PC 1 updated at that time. The PC 9 disappears from the data C and data D possession nodes, and the redundancy is “1”.

<Adding owned nodes>
FIG. 9A shows a state in which the redundancy control processing (steps S14 to S19 in FIG. 6) has been performed by the PC 1 continuously.

  The PC 1 determines that the redundancy of the data C and the data D has changed to “1” that is below the reference lower limit value “2” (step S14), and whether or not the data C and the data D can be held in other nodes. Inquiry polling is performed (step S16a).

  The number of additional owned nodes required to increase the redundancy “1” to “2” is one. Here, the PC 8 that responded the earliest is determined as the additional holding node, and it happens that the PC 1 holds the data C and the data D, so that a copy of the data C and the data D is sent to the PC 8 (step S16b). Or PC1 may instruct | indicate transmission to another possession node. Further, data C and data D may be additionally held in different nodes.

  As a result, the redundancy of data C and data D satisfies the lower limit value “2”.

  The PC 1 updates the description in the monitoring data list (step S18). In addition, the update contents are notified to the other monitoring nodes (PC2, PC3, and PC9), and based on this, the monitoring data list held by each node is updated (step S19).

  FIG. 9B shows the description content of the monitoring data list updated at that time. The PC 8 is added as a holding node of the data C and the data D, and the redundancy is returned to “2”.

  In this state, the process returns to the polling of redundancy detection processing (step S11 in FIG. 6). If there is no change in the network shown in (a), there is no change in the redundancy of each data. There is no change in the contents of the monitoring data list shown.

<Restoring the node>
In FIG. 10A, redundancy detection processing (steps S11 to S13 in FIG. 6) is performed by an arbitrary node (referred to as PC1) that first performs polling after the PC 9 returns to the network. Indicates the state.

  The PC 1 polls the nodes in the network and obtains a response from the PC 9 (step S11). As a result, the number of nodes holding data C and data D increases, and it is detected that the redundancy has changed from “2” to “3” (step S12).

  Due to this change, the redundancy of data C and data D deviates from the reference value “2” (step S13), so that the redundancy control process (after step S14) is started.

  FIG. 10B shows the description content of the monitoring data list of the PC 1 updated at that time. A PC 9 is added to the node having data C and data D, and the redundancy is “3”.

<Delete possession node>
FIG. 11A shows a state in which the redundancy control processing (steps S14 to S19 in FIG. 6) has been performed by the PC 1 continuously.

  The PC 1 determines that the redundancy of the data C and the data D has changed to “3” exceeding the reference upper limit value “2” (step S15), and the nodes (PC8, PC9, The PC 1) is then polled for an inquiry as to whether the data can be deleted (step S17a).

  The number of possession node deletions required to reduce the redundancy “3” to “2” is one. In this case, the PC 9 that has responded earliest is determined as the holding node to delete, and the PC 9 is instructed to delete data C and data D (step S17b). Alternatively, since the PC 1 happens to hold the data C and the data D, the data held by the PC 1 may be deleted. Further, the data C and the data D may be deleted from different holding nodes.

  As a result, the redundancy of data C and data D satisfies the upper limit “2”.

  The PC 1 updates the description in the monitoring data list (step S18). At the same time, the update contents are notified to the other monitoring nodes (PC2, PC3, and the restored PC9), and based on this, the monitoring data list held by each node is updated (step S19).

  FIG. 11B shows the contents of the monitoring data list updated at that time. The PC 9 is deleted as a node having data C and data D, and the redundancy is returned to “2”.

  In this state, the process returns to the polling of the redundancy detection process (step S11 in FIG. 6).

(Processing example 2)
Hereinafter, processing example 2 of redundancy detection and redundancy control will be described with reference to FIGS. 12 and 13.

  Processing example 2 is an example in which data is divided and distributed and shared. Also, polling for detecting redundancy is not performed periodically, but at the timing of access for data acquisition.

  12 and 13 are diagrams showing the state of the network 1 before and after performing the redundancy detection process and the redundancy control process shown in FIG. Also in this processing example, the network 1 is configured by nodes PC1 to PC9.

<Initial state>
FIG. 12A shows the connection in the initial state. Here, all the nodes (PC1 to PC9) constituting the network 1 are nodes that hold the monitoring data list, that is, perform monitoring.

  FIG. 12B shows the contents of the monitoring data list. In the monitoring data list, all data held in the nodes PC1 to PC9 is described as monitoring target data as monitoring target data. Further, it is assumed that the redundancy level is set such that the lower limit value S1 = 2 and the upper limit value S2 = 2, that is, the redundancy level is a constant value “2”.

  In the distributed sharing in which data is given redundancy, taking data A as an example, data A1 to A5 are divided into five, and each divided data is distributed and shared by two nodes. That is, the number of owned nodes = 2 for any divided data, and the divided data A satisfies the redundancy “2” set as a reference.

  These states are described in the monitoring data list of FIG.

  As in Processing Example 1, each time data acquisition is accessed in this state, polling for redundancy detection processing (step S11 in FIG. 6) is performed. If there is no change in the network shown in FIG. There is no change in data redundancy, and redundancy detection and control processing does not proceed.

<Disconnecting nodes and adding owned nodes>
FIG. 13A shows a state in which the redundancy detection process starting from polling is performed after the PC 9 is suddenly disconnected from the network, and the redundancy control process is continuously performed according to the detected change in redundancy. Show. These processes (steps S11 to S19 in FIG. 6) are performed by an arbitrary node (PC1 here) that first accesses to acquire data A after the PC 9 is disconnected.

  In acquiring data A, PC1 polls each node for divided data A1 to A5 for restoring data A, but there is no response from PC9 that should have data A1 (step S11). Then, it is detected that the redundancy of the data A1 has changed from “2” to “1” (step S12).

  Due to this change, the redundancy of data A deviates from the reference value “2” (step S13), and the redundancy control process (after step S14) is started.

  The PC 1 determines that the redundancy of the data A has changed to “1” that is lower than the reference lower limit value “2” (step S14), and polls for an inquiry as to whether or not the data A1 can be held in another node. (Step S16a).

  The data required to increase the redundancy “1” to “2” is data A1, and the required number of additional owned nodes is one. Here, the PC 5 that has responded the earliest is determined as the additional holding node of the data A1, and because the PC 1 happens to hold the data A1, a copy of the data A1 is sent to the PC 5 by itself (step S16b). Or PC1 may instruct | indicate transmission to another possession node.

  As a result, the redundancy of data A satisfies the lower limit “2”.

  The PC 1 updates the description in the monitoring data list (step S18). In addition, the update contents are notified to the other monitoring nodes (all nodes), and the monitoring data list held by each node is updated based on the notification (step S19).

  FIG. 13B shows the description content of the monitoring data list updated at that time. The PC 5 is added as a holding node of the data A1, and the redundancy is “2” as the standard.

  In this state, access for acquiring data A will be performed. However, even if the redundancy detection processing polling (step S11 in FIG. 6) is performed, if the network shown in FIG. Since there is no change in the redundancy of each data, the process does not proceed to the redundancy control process.

  As described above, according to the information management method and the information processing apparatus as a node according to the present embodiment, in a network system in which information is distributed and shared among a plurality of nodes, the retained data is monitored to change the redundancy. To detect. Also, the node that holds the data is changed so that the redundancy is greater than or less than a predetermined value.

  As a result, it is possible to maintain an appropriate redundancy for data retention. Accordingly, it is possible to efficiently use the distributed and shared data without causing unnecessary troubles such as not having unnecessary data and wasting resources, and conversely being unable to acquire data due to insufficient redundancy.

  The scope of the present invention is not limited to the above embodiment. Unless it deviates from the meaning of this invention, those changed forms are also included in the range.

1 is a diagram illustrating an example of the overall configuration of a network 1. FIG. 2 is a diagram illustrating a hardware configuration example of a node (information processing apparatus) 2 configuring a network 1. FIG. It is a figure which shows the example of the connection form of each node 2 which comprises the network 1, ie, the logical topology of a node. It is a figure which shows the connection table TL example of the node 2 linked | related like FIG. 3 is a block diagram illustrating a functional configuration example of a node (information processing apparatus) 2. FIG. It is a flowchart which shows the flow of the typical process from a redundancy detection process to a redundancy control process at the time of data distribution sharing. It is a figure which shows the initial state of the network 1 in the process example 1. FIG. FIG. 8 is a diagram illustrating a state in which a node of the network 1 is disconnected and a redundancy detection process is performed in FIG. 7. It is a figure which shows the state in which the redundancy control process of the network 1 was performed in FIG. FIG. 10 is a diagram illustrating a state in which a node of network 1 is restored in FIG. 9 and a redundancy detection process is performed. It is a figure which shows the state in which the redundancy control process of the network 1 was performed in FIG. It is a figure which shows the initial state of the network 1 in the process example 2. FIG. FIG. 13 is a diagram illustrating a state in which a node of the network 1 is disconnected in FIG. 12 and a redundancy detection process and a redundancy control process are performed.

Explanation of symbols

1 Network (P2P)
2 Information processing device (node)
DESCRIPTION OF SYMBOLS 3 Switching hub 4 Router 5 Authentication server 201 Data holding part 202 Monitoring object data list holding part 204 Redundancy detection part 205 Redundancy operation part 206 Data operation part 207 Data reception part 208 Data analysis part 209 Data creation part 210 Data transmission part 211 Other information holding unit 212 Other operation unit TL connection table

Claims (13)

A method for managing information in a network system in which information is distributed and shared among a plurality of nodes,
A redundancy detection step of detecting a change in redundancy of each information based on a monitoring target data list for monitoring the redundancy of each information distributed and shared;
A redundancy control step of changing the information holding node so that the redundancy of each information is equal to or higher than a predetermined lower limit value according to a change in the redundancy of each information detected by the redundancy detection step; And a method of managing information. In the redundancy control step,
According to a change in the redundancy of each piece of information detected by the redundancy detection step, the information holding node is changed so that the redundancy of each piece of information becomes a predetermined upper limit value or less. The information management method according to claim 1. In the redundancy detection step,
A node holding the monitoring target data list detects a change in redundancy from a change in the number of nodes holding the information by inquiring each node holding each piece of monitoring target information based on the monitoring target data list. The information management method according to claim 1, wherein the information management method is an information management method. In the redundancy control step,
The node that has detected a change in the redundancy of each information in the redundancy detection step changes the holding node according to the change in the redundancy of each information, and the monitoring target according to the change in the holding node The information management method according to claim 1, wherein the data list is updated. The monitored data list is
Information held by the node holding the monitoring target data list, information held by nodes within the broadcast range, information held by a predetermined node, owned by a predetermined user 5. The information management method according to claim 1, wherein at least one of the received information and the information requested to be monitored is described as monitoring target data. The monitoring target data list is held in a plurality of nodes,
In the redundancy control step,
The update contents of the monitoring target data list are notified to each other between the plurality of nodes, and the monitoring target data list held by each node is updated respectively. Information management method described in the section. The information management method according to claim 1, wherein redundancy is provided by using a RAID technology for the distributed sharing of the information. An information processing apparatus as a node in a network system in which information is distributed and shared among a plurality of nodes,
Storage means for holding a monitoring target data list for monitoring the redundancy of each piece of information distributed and shared;
Redundancy detection means for detecting a change in redundancy of each information based on the monitored data list;
Redundancy control means for changing the information holding node so that the redundancy of each information is equal to or higher than a predetermined lower limit value according to a change in redundancy of each information detected by the redundancy detection means; And an information processing apparatus. The redundancy control means includes
According to a change in the redundancy of each piece of information detected by the redundancy degree detecting means, the information holding node is changed so that the redundancy of each piece of information becomes a predetermined upper limit value or less. The information processing apparatus according to claim 8. The redundancy detection means includes
10. The change in redundancy is detected from a change in the number of nodes holding the information by inquiring each node holding each piece of information to be monitored based on the monitoring target data list. The information processing apparatus described. The redundancy control means includes
The holding node is changed according to a change in redundancy of each piece of information detected by the redundancy detecting means, and the monitoring target data list is updated according to the change of the holding node. The information processing apparatus according to any one of claims 8 to 10. The monitored data list is
Information held by the node holding the monitoring target data list, information held by nodes within the broadcast range, information held by a predetermined node, owned by a predetermined user 12. The information processing apparatus according to claim 8, wherein at least one of the stored information and the information requested to be monitored is described as monitoring target data. The redundancy control means includes
13. The update contents of the monitoring target data list are mutually notified with other nodes holding the monitoring target data list, and the monitoring target data list held by each node is shared. The information processing apparatus according to claim 1.

Images