JP2009021788A - Information storage system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To keep resistance against data loss and to effectively use the memory capacity of the entire system even when the memory capacities of network apparatuses constituting the system are not uniform. <P>SOLUTION: Respectively different memory capacities are provided in providers P, a memory state report is periodically distributed from the providers P to a network NW, and the providers P and requesters R recognize the memory using conditions of the providers P. When the requester R wants to store a file C in the distributed shared memory 1 of the provider P, a free slot notification request is distributed to the provider P, and when receiving a slot notification response from the provider P, the file C is transmitted to the distributed shared memory 1 of the provider P which has distributed the slot notification response. The number of replicates can be changed corresponding to the importance of the operation setting information of the respective network apparatuses in a distributed storage system. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a network device that controls and monitors a device according to predetermined setting information and an information storage system that multiplexes and stores the setting information between requesters.

  2. Description of the Related Art Conventionally, it has been proposed to solve problems related to data durability in a client-server system by progress of an embedded device network that performs plug-and-play and P2P (Peer-to-peer) communication. The technology for realizing this embedded device network realizes that device data and computer processing programs are distributed over the network. In this embedded device network, in order to maintain data durability, it is necessary to protect the integrity of computer processing programs and data even when a part of the system fails.

  As a technique for maintaining the resistance of computer processing programs and data, generally, computer processing programs and data are stored in a duplicated manner in a network. By duplicating the data, even if a part of the system fails, the data can be restored after the system is restored.

  As described above, in order to maintain the durability of data, conventionally, an information backup server that provides a highly reliable backup of information on a network and provides a highly confidential service, and data distribution among a plurality of terminals are performed. It is considered to use a distributed RAID (Redundant Array of Inexpensive Disks) system or the like. Further, as a technology for backing up data to a server, the following Patent Document 1 and the like can be cited.

Further, conventionally, a technique for storing duplicate copies of information in the entire system by devices constituting a network and a technique for distributing information in the entire system evenly across all devices have been proposed.
JP 2004-102450 A

  However, in the above-described conventional network technology, in a system including an information backup server, connecting the information backup server and the device itself may be a problem of reliability. For example, when connection to the information backup server becomes impossible, the information backup server cannot respond and the backed up computer processing program and data cannot be used, and the possibility of data loss occurs.

  Further, in the system using the above-described distributed RAID, since a plurality of terminals are used, the cost for configuring the system, the maintenance cost, and the repair cost are increased. In addition, the capacity of the system increases and power consumption increases.

  Furthermore, since the technology for storing the entire system information in the device by duplicating or distributing it is necessary for all the devices that make up the network to have the same capacity for data storage, this technology is adopted for devices with small memory capacity. Cannot join the network, and the tolerance to data loss is low. Therefore, the technology for storing the information of the entire system in an overlapped or distributed manner in the device is a technology that can be applied only to a highly functional device having a high memory capacity, and the cost of the entire system is increased.

  Furthermore, the technology for storing the information of the entire system by duplicating the information in the device has a huge amount of duplicate data, and each device needs to be able to store the information of the entire system. Not suitable for devices with low capacity. In a network employing this technology, if the memory capacity of each device is constant, the information of the entire system is simply proportional to the number of devices, and reliable if the number of devices is higher than the memory capacity of each device. Sex disappears.

  Furthermore, an embedded device network in which devices having a low memory capacity and devices having a high memory capacity such as terminals are mixed is different, but most devices have a small memory capacity. In the technology for duplicating information of the entire system, the amount of data that can be stored in each device is restricted to the device having the smallest memory capacity, and only a part of the range of the memory capacity of the device having the high memory capacity can be used. Therefore, there is a problem that the memory capacity that can be used in the entire embedded device network is wasted.

  Therefore, the present invention has been proposed in view of the above-described circumstances, and maintains resistance to data loss in the network, and also reduces the memory capacity of the entire system even when the memory capacity of network devices constituting the system is not uniform. An object of the present invention is to provide an information storage system that can be used effectively.

  The present invention provides an information storage system that includes a plurality of network devices, and each network device stores operation setting information necessary for its operation in another network device. Alternatively, all network devices store operation setting information transmitted from other network devices, and have backup storage means that function as a backup memory for the operation setting information of the other network devices. An information storage means for storing its own operation setting information, a communication means connected to another network device via a communication line and communicating with the other network device, and an own storage device stored in the information storage means. Storage control means for storing operation setting information in backup storage means of other network devices Equipped with a.

  In such an information storage system, in order to solve the above-described problems, the storage control means sets in advance the importance of its own operation setting information and operation setting information of other network devices connected via a communication line. The operation setting information is stored for backup so that the number of copies of the operation setting information of the network device having the higher importance is larger than the number of copies of the operation setting information of the lower importance among the operation setting information of the plurality of network devices. The storage control means acquires the memory state of the backup storage means from the network device having the backup storage means, and own operation setting information for the number of copies based on the importance of the own operation setting information. If the network device has the largest number of copies of operation setting information than the number of copies based on importance, Constant to, instead of the operation setting information of the network devices, and stores the operation setting information for itself.

  According to the information storage system of the present invention, the operation setting information to be stored in the backup storage means is dynamically changed according to a request from another network device. Therefore, even if the importance is set in the operation setting information In addition, it is possible to optimize the holding state of operation setting information of a plurality of network devices, maintain resistance against data loss in the network, and even if the memory capacity of the network devices constituting the system is not uniform, the memory capacity of the entire system Can be used effectively.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  The present invention uses the distributed shared memory 1 in the entire system to dynamically change the operation of storing redundant information necessary for a plurality of network devices, thereby causing data loss of information necessary for the network devices. This is applied to a distributed storage system that avoids the problem.

  Specifically, in the distributed storage system, each network device (device controller described later) has its own operation setting information (configuration description file described later) and operation settings of other network devices connected via a communication line. The importance of information is set in advance. Each network device is configured to increase the number of copies of the operation setting information of the network device having the higher importance than the number of copies of the operation setting information of the lower importance among the operation setting information of the plurality of network devices. The information is stored in the backup storage means, and the storage control means acquires the memory state of the backup storage means from the network device having the backup storage means (distributed shared memory to be described later), and determines the importance of its own operation setting information. If it is determined that the operation setting information corresponding to the number of replicas based on the degree cannot be stored, the network device having the largest number of replicas of the operation setting information than the number of replicas based on the importance is determined, and the network Instead of the device operation setting information, its own operation setting information is stored.

  With regard to the importance in this distributed storage system, for example, the operation setting information of a device controller connected to a device related to disaster prevention related to human life is highly important. Similarly, the operation setting information of the device controller connected to the fire (heat detection) sensor and the smoke sensor is also set to be highly important in the distributed storage system. Furthermore, the operation setting information of a device controller that protects assets such as a device controller connected to a device related to crime prevention such as an entrance / exit camera and a security sensor is highly important.

  On the other hand, the operation setting information of an air conditioner that is not directly related to a person's name, property, and crime prevention, or a device controller that is connected only to a lighting device is less important.

[Example of basic configuration of distributed storage system]
For example, as shown in FIG. 1, the distributed storage system includes a plurality of device controllers (network devices) D1 to D4 (hereinafter collectively referred to as “device controller Di”) in a network NW that is a general-purpose bus. Applies to connected distributed storage systems. Examples of the network NW that is a general-purpose bus include the Internet using IP (Internet Protocol). Each of the plurality of device controllers D1 to D4 can perform P2P (Peer-to-peer) communication, and can exchange information with each other.

  Each device controller Di is connected to a subnetwork in which a plurality of end devices are connected to a field bus. The device controllers D1 to D4 store configuration description files CFGi (C1 to C4) that store information for controlling and monitoring a plurality of terminal devices connected to the fieldbus. The device controllers D1 to D4 store the configuration description file CFGi of the other device controller Di in their own distributed shared memories 1-D1, 1-D2, 1-D3, 1-D4, respectively. This distributed shared memory 1 is provided separately from the storage means for storing its own configuration description file CFGi, and functions as a backup storage means for the configuration description file CFGi of another device controller Di.

  In the example shown in FIG. 1, the distributed shared memory 1-D1 of the device controller D1 has a memory capacity (three slots) that can store three configuration description files CFGi of other device controllers Di. The distributed shared memory 1-D2 of D2 has a memory capacity (two slots) that can store two configuration description files CFGi of other device controllers Di, and the distributed shared memory 1-D3 of the device controller D3. Has a memory capacity (3 slots) capable of storing three configuration description files CFGi of other device controllers Di, and the distributed shared memory 1-D4 of the device controller D4 has a configuration of the other device controller Di. Memory capacity that cannot store description file CFGi (distributed shared memory 1 ) Has become a state. The distributed shared memory 1-D1 stores configuration description files CFG2, CFG3, CFG4, the distributed shared memory 1-D2 stores configuration description files CFG1, CFG3, and the distributed shared memory 1-D3. The configuration description files CFG1, CFG2, and CFG4 are stored, and the configuration description file CFGi is not stored in the distributed shared memory 1-D4.

  Therefore, this distributed storage system has a distributed shared memory 1 with different capacities for each device controller Di, and each device controller Di follows the requests from other device controllers Di and other network devices. The configuration description file CFGi stored in its own distributed shared memory 1 is dynamically changed. That is, each device controller Di has a number of copies of the configuration description file CFGi of the network device controller Di having a higher importance than the number of copies of the configuration description file CFGi having a lower importance among the configuration description files CFGi of the plurality of device controllers Di. The configuration description file CFGi is stored in the distributed supply memory 1 so as to increase the number.

[Basic operation of distributed storage system]
As shown in FIG. 2, this distributed storage system updates the distributed shared memory 1 stored therein in response to the update of the configuration description file CFGi of another device controller Di, and As shown, it has a function of providing a configuration description file CFGi stored in its own distributed shared memory 1 to another device controller Di.

  In the distributed storage system, for example, when the configuration description file CFG4 + is updated by the operator, the device controller D4 converts the configuration description file CFG4 stored in another device controller Di to the configuration description file CFG4 +. Broadcast a file update request to be updated to the network NW. When the device controller D1 and the device controller D3 storing the configuration description file CFG4 receive the file update request, the configuration description file CFG4 stored in their own distributed shared memories 1-D1 and 1-D3, Update to the configuration description file CFG4 + attached to the file update request. Further, since the device controller D3 does not store the configuration description file CFG4, it discards the file update request.

  Also, the device controller ID for the device controller D4 to be replaced with a new device controller Di as shown in FIG. 3A by the operator to become the device controller D4 as shown in FIG. 3B. Is added, the device controller D4 broadcasts a file transfer request including the device controller ID to the network NW in order to request its own configuration description file CFGi.

  As shown in FIG. 3C, when the device controller D1 storing the configuration description file CFG4 receives the file transfer request, the configuration description file CFG4 stored in its own distributed shared memory 1-D1. Is returned to the device controller D4. As a result, the device controller D4 can obtain the configuration description file CFG4 for communicating with the end device of the fieldbus. Here, the device controller Di3 cancels the file transfer request from the device controller D4 because the processing load including the storage control unit 12 is currently high. The device controller D3 discards the file transfer request because it does not store the configuration description file CFG4.

  As described above, in the distributed storage system, when the configuration description file CFGi of another device controller Di connected to the network NW is updated, or when the other device controller Di is exchanged and does not hold the configuration description file CFGi. In addition, a configuration description file CFGi owned by itself can be provided to the other device controller Di. Therefore, in this distributed storage system, as described below, even when the memory capacity of each network device is different and the importance of the configuration description file CFGi of each network device is different, the configuration descriptions of other network devices are mutually The storage state of the file CFGi is dynamically changed to maintain the resistance against data loss of the configuration description file CFGi of each network device.

[Components of distributed storage system]
Such a distributed storage system is not limited to the case where the distributed shared memory 1 is used between the device controllers Di connected to the field bus and controlling the end devices connected to the field bus, as described above. The distributed shared memory 1 can be used also between other network devices connected to the NW.

  In the distributed shared memory 1, the number of copies of the configuration description file CFGi of each network device is different. For example, the number of copies of the configuration description file CFGi based on the importance of the first device controller Di in the distributed storage system is “5”, and the number of copies of the configuration description file CFGi based on the importance of the second device controller Di is “3”, and the number of copies of the configuration description file CFGi based on the importance of the third device controller Di is “1”.

  On the other hand, the distributed storage system dynamically changes the usage status of the distributed shared memory 1 between network devices connected to the network NW, and the configuration description file CFGi of each network device has a configuration based on importance. As many as the number of copies of the description file CFGi, the status of being backed up to other network devices is retained.

  As shown in FIG. 4, the types of network devices in this distributed storage system include a requester Ri for storing its own file using another distributed shared memory 1 and another distributed shared memory 1 as another type. There is a provider Pi that provides network devices. The requester Ri and the provider Pi use the distributed shared memory 1 to avoid data loss of the configuration description file CFGi. In the example shown in FIG. 4, requesters Ra, Rb, and Rc that are network devices that do not have the distributed shared memory 1, and providers P1, P2, P3, P4, and P5 that provide the distributed shared memory 1 to other network devices. It shows the case where and exist. The device controller Di described above corresponds to the provider Pi because it has the distributed shared memory 1, and also corresponds to the requester Ri in the sense that the configuration description file CFGi is stored in another device controller Di.

  As shown in FIG. 5, the distributed storage system outputs a memory status report periodically or randomly by the shared memory providing function of each provider Pi, and receives the memory status report by the requester Ri. Then, when the requester Ri has a slot in the distributed shared memory 1 of the provider Pi and wants the provider Pi to back up its own configuration description file CFGi, the requester Ri sends a request for acquiring an empty slot to the provider Pi. A response from Pi can be obtained.

[Configuration example of provider Pi]
Next, a configuration example of the provider Pi in the distributed storage system to which the present invention is applied will be described. The provider Pi is connected to a field bus like the device controller Di described above, and controls and monitors a terminal device connected to the field bus.

  The provider Pi is configured as shown in FIGS. As shown in FIG. 6, the provider Pi uses a device embedded network technology (EMIT (Embedded Micro Internetworking Technology), hereinafter referred to as EMIT technology)) to communicate with the terminal device and the provider Pi. A, the terminal device and the provider Pi perform baseband transmission that modulates the pulse width to perform signal transmission / reception, a field bus B that performs communication in a predetermined multiplex transmission method (Nmast method), and a LON (Local Operating System). A field bus C that performs communication by a distributed control network method called “Network” is connected. Note that the field bus may be connected to the provider Pi not only in the case of using the EMIT technology, the enmast system, and the LON, but in any system.

  A plurality of end devices such as end devices A0 to A2, S3 to S7,... Are connected to the field bus A, and a human sensor is included as the end device S3. A plurality of end devices such as end devices S0, S1, A1, A3, A4... Are connected to the field bus B, and the end device A1 includes illumination. Further, a plurality of end devices such as end devices A0 to A2, S3 to S5,... Are connected to the field bus C, and a temperature sensor is included as the end device S2.

  The provider Pi includes a P2P communication processing unit 11 connected to a network NW that performs P2P communication with other providers Pi and requesters Ri, a storage control unit 12, and the above-described distributed shared memories 1-D1, 1-D2, 1-. Distributed shared memory 13 corresponding to D3, 1-D4, configuration description information driver 14, variable table 15, application processing unit 16, fieldbus monitoring / control unit 17, and a plurality of fields connected to itself Field bus interfaces 18A, 18B, 18C corresponding to buses A, B, C are provided.

  The fieldbus interfaces 18A, 18B, and 18C are set so as to perform communication by a method corresponding to each fieldbus A, B, and C. For example, the fieldbus interface 18A employing the EMIT technology and the end device connected to the fieldbus interface 18A are configured as shown in FIG.

  The terminal devices on the field buses A, B, and C are equipped with a MOS (Micro Object Server) function unit 21 that transmits and receives control data of various terminal devices, status signals of the terminal devices, and the like between the EMIT middlewares. 18A includes an OAS (Object Access Server) function unit 31 that performs routing processing between EMIT middleware, end device recognition processing, and the like. A user terminal (not shown) that manages the end device has a control target. An OAL (Object Access Library) function unit (not shown) that outputs a control / monitoring request to the terminal device and displays the status of the terminal device is mounted. Although FIG. 8 illustrates a case where only the OAS function unit 31 is mounted on the fieldbus interfaces 18A, 18B, and 18C, the OAL function unit may be mounted.

  In the terminal device, an application processing unit 22 and an interface module 23 are connected to the MOS function unit 21.

  For example, when the application processing unit 22 functions as a sensor, an application code for acquiring a detection value (I / F value) is stored, and the detection value is held and transferred to the MOS function unit 21.

  The interface module 23 performs processing such as a data link layer in an OSI (Open Systems Interconnection) reference model. For example, processing according to Ethernet (registered trademark) is performed in the data link layer. Capabilities such as a communication protocol version and a communication speed of the interface module 23 are stored in the capability table 24 and can be acquired by the OAS function unit 31.

  The MOS function unit 21 functions as EMIT middleware between the application layer by the application processing unit 22 and the transport layer by the interface module 23. The MOS function unit 21 is given an object ID 21d by the OAS function unit 31 of the fieldbus interface 18A in order to identify that the end device is an object in EMIT.

  The MOS function unit 21 performs an operation defined by a function 21a, an event 21b, and a variable 21c. The operation of the MOS function unit 21 is defined as a service table 21e and can be acquired as an interface ID by the OAS function unit 31 of the fieldbus interface 18A.

  When the terminal device is a sensor, the function 21a is a detection value output function, the event 21b is an event for outputting the detection value when the detection value becomes a predetermined value, for example, and the variable 21c is the detection value itself. Such a terminal device generates an event at event 21b when the detected value reaches a predetermined value, and outputs the detected value of variable 21c by function 21a.

  A communication module 32 is connected to the OAS function unit 31 of the fieldbus interface 18A.

  The OAS function unit 31 is connected to a fieldbus monitoring / control unit 17 described later, and communicates with a terminal device of the fieldbus A according to the control of the fieldbus monitoring / control unit 17. The fieldbus monitoring / control unit 17 is supplied with a control / monitoring request and returns a response to the control / monitoring request when the user and the administrator want to communicate to control / monitor a certain end device. .

  The communication module 32 communicates with the interface module 23 of the end device, and communicates with the communication module 32 of the apparatus in which the other OAS function unit 31 is mounted. The communication module 32 performs processing such as a data link layer in the OSI reference model. When each terminal device is configured by various interface modules 23, the communication module 32 performs communication protocol version and communication for each device including the terminal device and other OAS function units 31 according to the control of the device access controller 31 c. Communication processing is performed according to the capability of the end device such as speed.

  The OAS function unit 31 functions as EMIT middleware between the process of the fieldbus monitoring / control unit 17 and the process of the communication module 32. The OAS function unit 31 includes a device access server 31a, service information 31b stored in a predetermined memory, and a device access controller 31c.

  The device access server 31a receives a request from the client via the communication module 32, the fieldbus monitoring / control unit 17, and the P2P communication processing unit 11 in order to perform communication with the client (OAL). The device access server 31a controls the operation of transmitting the control / monitoring request to the end device via the device access controller 31c and the communication module 32. Thus, the device access controller 31c transmits a request from the client to the end device, and controls the end device.

  The device access controller 31 c searches the terminal device service table 21 e and the capability table 24 via the communication module 32 in order to obtain the terminal device interface definition and service information. Thereby, the device access controller 31c determines what service (function 21a, event 21b, variable 21c) each end device has, and indicates the correspondence between the object ID 21d of each end device and the service list. Service information 31b is created. Further, the device access controller 31c recognizes what capability (communication protocol version, communication speed) each terminal device has.

  When the MOS function unit 21 receives an event packet of an event at the communication module 32, the device access controller 31c analyzes the event packet by the device access controller 31c and notifies the device access server 31a. When a control / monitoring request from a client is transmitted to a terminal device, the MOS function unit 21 that can respond to the control / monitoring request is recognized with reference to the service information 31b. Then, the destination object ID 21d is added to the control / monitoring request, and the packet is transmitted from the communication module 32 to the terminal device in accordance with the capability of the terminal device.

  The OAS function unit 31 is provided with an object ID 31d so that other EMIT middleware-compatible devices can identify the OAS function unit 31. The object ID 31d is stored in a packet to be transmitted when communicating with the MOS function unit 21 of the end device or an EMIT middleware compatible device such as a client (personal computer) in which another OAL is mounted.

  As shown in FIG. 7, the field bus interfaces 18A, 18B, and 18C including the OAS function unit 31 are configured to include a field bus driver 18a and a field bus interface unit 18b.

  When receiving the control data from the fieldbus monitoring / control unit 17, the fieldbus driver 18a refers to the configuration description information from the configuration description information driver 14 to determine which end device is to be operated and how. , Send a command to the end device. When the fieldbus interface unit 18b receives a command from the fieldbus driver 18a, the fieldbus interface unit 18b converts the command into a signal corresponding to the fieldbus A, B, or C, and sends the signal to the fieldbus A, B, or C.

  When the field bus interface unit 18b receives monitor data from the field buses A, B, and C to notify the field bus monitoring / control unit 17 of the state change of the end device and the occurrence of an event, the field bus interface unit 18b stores the monitor data in the field. The data is transferred to the bus driver 18a and sent to the fieldbus monitoring / control unit 17.

  The fieldbus monitoring / control unit 17 refers to the variable table 15 and controls the fieldbuses A, B, and C. As shown in FIG. 9, the variable table 15 is configured by associating terminal device names, terminal device states, variable types, field bus IDs, and field bus addresses for each of the field buses A, B, and C. Yes. Then, the field bus monitoring / control unit 17 monitors the end devices of the field buses A, B, and C, updates the contents of the variable table 15, and refers to the variable table 15 to determine the field buses A, B, and C. Send control data to the end device.

  As shown in FIG. 7, the fieldbus monitoring / control unit 17 is configured by dividing a function into a fieldbus monitoring unit 17A and a fieldbus control unit 17B. When receiving the monitor data from the fieldbus driver 18a, the fieldbus monitoring unit 17A updates the contents of the variable table 15. At this time, the fieldbus monitoring unit 17A recognizes the end device name based on the object ID of the end device included in the monitor data, and rewrites the state of the end device. When the field bus control unit 17B receives control data from the application processing unit 16, the field bus control unit 17B refers to the variable table 15 and the configuration description file CFGi to determine the control data transmission destination field bus, terminal device, and the like. Transfer to the driver 18a.

  The application processing unit 16 controls application processing realized by the end device. The application processing unit 16 refers to the variable table and gives a control / monitoring request to the fieldbus monitoring / controlling unit 17. Accordingly, the end device is operated in accordance with the monitoring request and control request transmitted from the user via the network NW.

  In addition, as shown in FIG. 7, the application processing unit 16 creates an application processing unit 16A that sends control data to the end device, a linkage table 16B in FIG. 11 described later, and a linkage table 16B according to its own configuration description information. You may comprise as the application update part 16C to update.

  The configuration description information driver 14 stores a configuration description file CFGi for controlling the field buses A, B, and C connected thereto. As shown in FIG. 7, the configuration description information driver 14 is connected to the storage control unit 12 and includes a configuration description file storage unit 14a for storing its own configuration description file CFGi. When a plurality of configuration description files CFGi are transmitted from another device controller, the configuration description information driver 14 converts the configuration description file CFGi in the configuration description file storage unit 14a to the configuration description file CFGi selected by the information selection unit 12E. rewrite. Further, in the configuration description information driver 14, for example, when a new end device is connected to the field buses A, B, and C, information about the new end device is added to the variable table 15 by the field bus monitoring unit 17A. Is notified from the provider ID storage unit 19, the change of the field buses A, B, and C is reflected in the configuration description file CFGi. As a result, the configuration description file CFGi can always be kept up-to-date.

  The information (configuration description information) included in the configuration description file CFGi includes a field bus table that defines a field bus connected to itself as shown in FIG. 9, a variable table 15 in FIG. 10, and a configuration shown in FIG. 12 includes a cooperation table for defining a cooperation operation performed between various end devices, and software parameters storing other parameters as shown in FIG.

  The field bus table shown in FIG. 9 includes field bus IDs (fb0, fb1, fb2,...) Of the field buses A, B, and C, and the methods (EMIT, Fieldbus information defining Nmast and Ron). This field bus table is read by the field bus monitoring / control unit 17 and referred to in order to identify the field buses A, B, and C connected thereto.

  The variable table 15 shown in FIG. 10 is configured as described above, and is referred to by the fieldbus monitoring / control unit 17 and the application processing unit 16 to specify the state of the terminal device connected to itself. The variable table 15 is updated by referring to the configuration description information by the variable table updating unit 15A shown in FIG.

  The linkage table shown in FIG. 11 includes linkage event information that defines an event of a terminal device that becomes a trigger for starting a linkage operation, and linkage response device information that defines a terminal device that responds when the linkage occurrence event occurs. , Associated result information defining a result of operating the associated response device is associated. This linkage table is referred to by the application processing unit 16 when an event occurrence is detected by the fieldbus monitoring / control unit 17 via the fieldbus interfaces 18A, 18B, and 18C. If the application processing unit 16 determines that an event corresponding to the cooperation table has occurred, the application processing unit 16 operates so that the cooperation response device satisfies the cooperation result. In addition, this cooperation table may be comprised as a part function (cooperation table 16B) of the application process part 16, as shown in FIG.

  The software parameters shown in FIG. 12 store information types stored as parameters, parameter type information, and parameter values in association with each other. In this example, the size of the distributed shared memory 13, the capacity of its own configuration description file CFGi, and the number of times the terminal device is controlled are stored as parameters. The software parameters include parameters used in the fieldbus interfaces 18A, 18B, and 18C, parameters used in the application processing unit 16, parameters used in the storage control unit 12, parameters used in the P2P communication processing unit 11, and the like. Including.

  For example, the size of the distributed shared memory 13 is referred to by the storage control unit 12 to determine whether or not the configuration description file CFGi of the requester Ri and another provider Pi can be stored. It is determined whether or not the requester Ri and the other provider Pi can store their configuration description file CFGi. Note that the size of the distributed shared memory 1 may be described in slots representing the number of storable configuration description files CFGi instead of in bytes.

  The storage control unit 12 uses the distributed shared memory 1 in the entire network NW to perform a process of backing up its own configuration description file CFGi in the distributed shared memory 13 of another provider Pi, the requester Ri, and other providers A process of backing up the Pi configuration description file CFGi to its own distributed shared memory 13 is performed. When the storage control unit 12 backs up its own configuration description file CFGi to another provider Pi, the storage control unit 12 reads out its own configuration description file CFGi stored in the configuration description information driver 14, and receives the other provider from the P2P communication processing unit 11. Send to Pi. Further, the storage control unit 12 receives a backup request from the requester Ri and another provider Pi via the P2P communication processing unit 11 and stores the requester Ri and the configuration description file CFGi of the other provider Pi in the distributed shared memory 13. To do.

  Further, as shown in FIG. 7, the provider Pi is given a device controller ID by the operator's operation on the operation panel, the operation on the dip switch, or the like. This device controller ID is stored in the provider ID storage unit 19.

  As shown in FIG. 7, such a provider Pi includes an information receiving unit 12A, an information updating unit 12B, an information requesting unit 12C, a configuration description information memory 12D, and an information selecting unit 12E as the storage control unit 12.

  When the information requesting unit 12C detects from the software parameters in the configuration description file storage unit 14a that there is no self configuration description file CFGi, the information request unit 12C causes the P2P communication processing unit 11 to transmit a file transfer request including the device controller ID. . When one or more configuration description files CFGi are received by the P2P communication processing unit 11 from another provider Pi in response to the transmission of this file transfer request, the configuration description file CFGi is received by the information reception unit 12A. When the information receiving unit 12A accumulates the plurality of configuration description files CFGi in the configuration description information memory 12D, the information selection unit 12E stores the plurality of configuration description files CFGi stored in a voted form in the configuration description information memory 12D. The configuration description file CFGi that is the latest and not corrupted is selected and made usable by the configuration description information driver 14. As a result, the configuration description information driver 14 updates the configuration description file CFGi stored in the configuration description file storage unit 14a.

  Further, the storage control unit 12 updates the configuration description file CFGi when, for example, the configuration of the field buses A, B, and C is changed or when the provider Pi storing the old configuration description file CFGi is detected. There is a need to. At this time, the storage control unit 12 detects from the software parameter of the configuration description file storage unit 14a that the information update unit 12B needs to update the configuration description file CFGi. Then, the information update unit 12B includes the device ID in the file update request for updating its own configuration description file CFGi and broadcasts it from the P2P communication processing unit 11.

  Further, when the information receiving unit 12A receives the configuration description file CFGi from the requester Ri and another provider Pi, the storage control unit 12 stores the configuration description file CFGi of the requester Ri and other provider Pi in the distributed shared memory 13. Remember. Further, when the storage control unit 12 receives an empty slot acquisition request of the provider Pi from the requester Ri or another provider Pi, the information receiving unit 12A corresponds to the device ID included in the empty slot acquisition request. The configuration description file CFGi is read and transmitted.

[Distributed storage system functions]
As shown in FIG. 13, the distributed storage system including the provider Pi and the requester Ri as described above is installed in the provider Pi and the distributed shared memory monitoring function 41 and the storage control function 42 implemented in the requester Ri. And a shared memory providing function 43.

  The distributed shared memory monitoring function 41 is a function provided in each requester Ri, receives the memory status report S1 broadcast by the memory status report output function unit 43c of the provider Pi, and receives the status of the distributed shared memory 1 And an active monitoring unit 41a for monitoring its own file backup status.

  As shown in FIG. 14, the memory status report S1 includes a device ID of the provider Pi (hereinafter referred to as a provider ID) and the number of slots (configuration) among all slots of the distributed shared memory 13 of the provider Pi. The number of description files CFGi stored) and slot allocation information indicating which requestor Ri is allocated the number of possessed slots is received by the active monitoring unit 41a.

  Upon receipt of the memory status report S1, the active monitoring unit 41a creates information indicating the status of the distributed shared memory 1 and information indicating its own file backup status, and passes them to the storage control function.

  Information indicating the state of the distributed shared memory 1 includes an allocation table indicating which requester Ri's configuration description file CFGi is stored in each provider Pi, a table of the provider Pi to which the self is connected, and a total slot of the provider Pi. , The total number of requesters Ri, the total number of providers Pi, the required number of replicas of the configuration description file CFGi, and the ID of the requester Ri operating in the configuration description file CFGi having the largest number of replicas relative to the required number of replicas. This required number of copies is the number of copies of the configuration description file CFGi based on the importance set in advance in the distributed storage system. The ID of the requester Ri that operates in the configuration description file CFGi having the largest number of copies relative to the required number of copies is based on the current number of copies of each device controller Di and the importance in the preset distributed storage system. It is obtained from the difference from the required number of replicas of each device controller Di.

  The information indicating the self file backup status includes the ID of the backup provider Pi that stores the self configuration description file CFGi, the number of copies of the self configuration description file CFGi in the entire current distributed storage system, and the desired number of requested copies. The number of providers Pi that have available memory in the distributed shared memory 13 and the hit rate obtained by dividing the number of requested replicas by the number of providers Pi that can be stored. This desirable requested number of copies is a value obtained by subtracting the number of copies of the current configuration description file CFGi from the required number of copies of the configuration description file CFGi based on importance.

  The storage control function 42 includes a configuration description information storage unit 42a corresponding to the configuration description file storage unit 14a, a shared memory state file 42b, a backup state file 42c, a memory state detection function unit 42d, and an empty slot question function unit 42e. An empty slot acquisition function unit 42f, a copy number warning function unit 42g, and a slot reassignment function unit 42h.

  The storage control function 42 passes the information indicating the state of the distributed shared memory 1 created by the distributed shared memory monitoring function 41 from the active monitoring unit 41a and the information indicating the backup state to the memory state detection function unit 42d. Then, the shared memory state file 42b and the backup state file 42c are created and can be referred to from the empty slot inquiry function unit 42e and the empty slot acquisition function unit 42f.

  When it is desired to store its own configuration description file CFGi in the provider Pi, the empty slot question function unit 42e broadcasts an empty slot notification request S2 to the network NW and receives an empty slot notification response S3 for the empty slot notification request S2. To do. As shown in FIG. 14, the empty slot notification request S2 includes its own device ID (hereinafter, the device ID of the requester Ri is referred to as requester ID) and the hit rate stored in the backup status file 42c. Including.

  This empty slot notification request S2 is received by the shared memory providing function 43 installed in the provider Pi. In response to the received empty slot notification request S2, the shared memory providing function 43 searches for an empty slot by referring to the distributed shared memory 43b corresponding to its own distributed shared memory 13 by the request management function 43a. When there is a slot, it broadcasts an empty slot notification response S3 including its own provider ID to the network NW.

  When this slot notification response S3 is received by the empty slot question function unit 42e of the storage control function 42, the empty slot question function unit 42e notifies the empty slot acquisition function unit 42f of the provider ID included in the slot notification response S3. As shown in FIG. 14, the empty slot acquisition function unit 42f broadcasts a command, a provider ID including an empty slot, a provider list including the provider ID, and an empty slot acquisition request S4 including its own configuration description file CFGi to the network NW. To do.

  In addition, the empty slot inquiry function unit 42e notifies the slot reassignment function unit 42h of the slot notification response S3, and the empty slot acquisition function unit 42f broadcasts the slot reassignment request S5 to the network NW. As shown in FIG. 14, the slot reassignment request S5 stores the command, the provider list, the requester ID of the configuration description file CFGi stored in the use slot by the provider Pi included in the provider list, and the use slot. It includes a new (self) requester ID and a new (self) configuration description file CFGi.

  When the shared memory providing function 43 of the provider Pi receives the empty slot acquisition request S4 by the request management function 43a, the configuration description file CFGi included in the empty slot acquisition request S4 is stored in the distributed shared memory 43b. Further, when the memory state of the distributed shared memory 43b changes due to reception of the empty slot acquisition request S4 or the slot reassignment request S5, the request management function 43a stores the changed slot assignment state stored in the distributed shared memory 43b. Is generated and broadcast from the memory status report output function unit 43c. Thereby, when the active monitoring unit 41a receives the memory status report S1 indicating that the slot allocation status has changed, the shared memory status file 42b and the backup status file 42c are updated.

  In addition, the storage control function 42 is notified of the slot notification response S3 from the empty slot question function unit 42e, and the copy number warning function unit indicates that none of its own configuration description file CFGi is replicated in the distributed storage system. Detect at 42g. If such a configuration description file CFGi is not replicated, the active monitoring unit 41a is notified to that effect, and the current number of replicas of the backup status file 42c is updated to “0”.

[Processing procedure of distributed storage system]
In the distributed storage system having such a function, operation procedures of the distributed shared memory monitoring function 41 and the storage control function 42 will be described with reference to FIGS. 15 and 16.

  The distributed shared memory monitoring function 41 of the requester Ri first performs the processing of step ST1 to step ST8 shown in FIG. 15 to sort out how much its own configuration description file CFGi is replicated to other providers Pi. Therefore, the shared memory state file 42b and the backup state file 42c are updated, and the processes in steps ST11 to ST24 shown in FIG. 16 are performed by the storage control process in the next step ST9.

  In step ST1 of FIG. 15, the active monitoring unit 41a resets the shared memory state file 42b and the backup state file 42c, detects the memory state report S1 distributed from the network NW in step ST2, and in step ST3. It is determined whether or not the memory status report S1 has been detected. If the memory status report S1 can be detected, the process proceeds to step ST4 to update the allocation table of the shared memory status file 42b. This step ST2 is performed in units of time twice as long as a predetermined period for transmitting the memory status report S1 in order to reliably receive the memory status report S1 transmitted every time.

  This allocation table is a table indicating each provider ID and which device (provider Pi, requester Ri) the configuration description file CFGi stored in the provider Pi of the provider ID. In step ST8, the active monitoring unit 41a updates the shared memory state file 42b, and proceeds to the storage control process in the storage control function 42 in step ST9.

  On the other hand, when the memory status report S1 has not been received, the active monitoring unit 41a enters a waiting state for a predetermined time in step ST5, and when the predetermined time has elapsed without receiving the memory status report S1, step ST6. Proceed with the process. The predetermined time is provided at least twice as long as a predetermined interval for transmitting the memory status report S1 in the distributed storage system.

  In step ST6, the active monitoring unit 41a refers to the allocation table included in the shared memory state file 42b, and includes a device table including provider IDs and requester IDs constituting the distributed storage system, the total number of slots in the distributed storage system, Requester information with the maximum number of replicas that stores the largest number of configuration description files CFGi for the total number of requesters, the total number of providers, and the required number of replicas (requester ID, number of extra replicas for the required number of replicas), distributed storage The necessary number of copies of the configuration description file CFGi based on the importance of the configuration description file CFGi in the system, the number of backup providers, the current number of replicas, and the number of providers that can use the memory are calculated.

  Next, the active monitoring unit 41a updates the backup state file 42c in step ST7 and updates the shared memory state file 42b in step ST8 according to the parameters calculated in step ST6, and performs the storage control process in step ST9. Transition.

  As shown in FIG. 17, in the storage control process, first, in step ST11, the memory state detection function unit 42d determines that the number of copies of the current configuration description file CFGi included in the backup state file 42c is the shared memory state file. It is determined whether or not the number of required replications is based on the importance included in 42b. If the number of copies of the own configuration description file CFGi is equal to or greater than the required number of copies, the number of copies of the own configuration description file CFGi is large and there is no risk of data loss, so the process returns to step ST1 in FIG. .

  On the other hand, if it is determined that the number of replicas of the configuration description file CFGi is not equal to or greater than the required number of replicas, in step ST12, the memory status detection function unit 42d determines whether the number of providers that can use the memory in the backup status file 42c is zero. Determine whether or not. If it is determined that the number of providers that can use the memory is 0, the process proceeds to step ST21. If it is determined that the number is not 0, the process proceeds to step ST13.

  In step ST13, the memory state detection function unit 42d needs to add the own configuration description file CFGi in the distributed storage system by subtracting the number of copies of the current configuration description file CFGi from the required number of replicas based on the importance. The requested number of replicas that is a certain number of replicas is calculated, and the process proceeds to step ST14.

  In step ST14, the empty slot inquiry function unit 42e distributes the empty slot notification request S2 to the network NW. At this time, the empty slot inquiry function unit 42e includes in its empty slot notification request S2 a hit rate obtained by dividing its own requester Ri and the number of requested replicas by the number of providers that can use the memory. The empty slot notification request S2 is received by all providers Pi of the distributed storage system, and the request management function 43a of the shared memory providing function 43 refers to its own distributed shared memory 43b and there is an empty slot. Broadcast an empty slot notification response S3 including its own provider ID to the network NW.

  In the next step ST15, it is determined whether the empty slot question response unit 42e of the requester Ri has received the empty slot notification response S3. If the empty slot question function unit 42e has not received the slot notification response S3, In step ST16, it is determined that there is no empty slot in the distributed storage system. Also, the empty slot inquiry function unit 42e receives the slot notification response S3, and if the number of empty slots in the distributed storage system is greater than 0 and less than the requested replication number, the process proceeds to step ST17, and the requested replication number If the number of empty slots is larger than that, the process proceeds to step ST18.

  When the empty slot inquiry function unit 42e determines that the number of empty slots is greater than 0 and less than the requested number of copies, the fact is notified to the empty slot acquisition function unit 42f, and in step ST19, the empty slot acquisition function unit The empty slot acquisition request S4 is distributed to the network NW by 42f, and the process returns to step ST11. This empty slot acquisition request S4 includes a list of provider IDs broadcast when there is an empty slot number by the slot notification response S3, its own provider ID, and the configuration description file CFGi. This list of provider IDs includes a list of empty slots.

  Thus, when the empty slot acquisition request S4 is received by the request management function 43a of the shared memory providing function 43 of the provider Pi, the configuration description file CFGi included in the empty slot acquisition request S4 is stored in the distributed shared memory 43b. Further, the shared memory providing function 43 of the provider Pi is a memory for transmitting a change in the memory state caused by storing the new configuration description file CFGi in the distributed shared memory 43b, and then transmitting the change in the memory state report output function unit 43c. It is reflected in the status report S1.

  In step ST20 after determining that the number of empty slots is larger than the requested number of replicas in step ST18, the empty slot acquisition function unit 42f has its own requested number of replicas, a list of empty slots, its own provider ID, and a configuration description. An empty slot acquisition request S4 including the file CFGi is broadcast to the network NW. As a result, a number of configuration description files CFGi satisfying the required number of replicas are replicated and stored in the distributed storage system.

  Further, in step ST21 after determining that the number of empty slots is 0 in step ST16, the empty slot question function unit 42e operates the requester Ri that operates on the configuration description file CFGi having the largest number of replicas with respect to the required number of replicas. It is determined whether or not the ID can be identified. When the ID of the requester Ri can be identified, in step ST22, the ID of the requester Ri is notified to the slot reallocation function unit 42h, and the slot reallocation request S5 is sent to the network NW by the empty slot acquisition function unit 42f. Broadcast. The slot reassignment request S5 includes a list for designating the slot of the provider Pi in which the configuration description file CFGi of the requester Ri operating with the configuration description file CFGi having the largest number of replicas with respect to the required number of replicas, It includes an ID of a copy number possessing requester, a new requester ID that is a self device ID that newly uses a slot instead of the maximum replicating number possessing requester, and its own configuration description file CFGi. When there are a plurality of requesters Ri having the largest number of replicas with respect to the required number of replicas, the ID of any requester Ri is identified, and the process proceeds from step ST21 to step ST22.

  On the other hand, if the ID of the requester Ri operating with the configuration description file CFGi having the largest number of replicas relative to the required number of replicas cannot be identified, whether or not the current number of replicas of its own configuration description file CFGi is 0 in step ST23. If it is determined that this is not the case, the process returns to step ST11. If the number of copies of its own configuration description file CFGi is 0, the process proceeds to step ST24.

  In step ST24, the empty slot question function unit 42e notifies the copy number warning function unit 42g that the number of copies of its own configuration description file CFGi is 0, and the copy number warning function unit 42g provides a warning including this. The message is notified to the active monitoring unit 41a.

  In such a distributed storage system, the required number of copies of the configuration description file CFGi based on the importance of the first device controller Di is “5”, and the configuration description file CFGi based on the importance of the second device controller Di is Consider the case where the required number of replicas is “3” and the required number of replicas of the configuration description file CFGi based on the importance of the third device controller Di is “1”.

  In this case, the current number of replicas of the first device controller Di is “5”, the current number of replicas of the second device controller Di is “3”, and the current number of replicas of the third device controller Di is “1”. If YES, the determination in step ST11 in the first to third device controllers Di is “YES”.

  Also, the current number of replicas of the first device controller Di is “3”, the current number of replicas of the second device controller Di is “3”, and the current number of replicas of the third device controller Di is “3”. In some cases, only in the first device controller Di, the determination in step ST11 is “NO”. In step ST12, when there is no empty slot in the other device controller Di, the first device controller Di is set as a requester having the largest number of replicas with respect to the required number of replicas in step ST21. The third device controller Di can be selected. As a result, the required number of replicas of the first device controller Di is “5” and the required number of replicas of the third device controller Di is “1”, so that the configuration description file CFGi of the third device controller Di is stored in the configuration description file CFGi. The configuration description file CFGi of the first device controller Di can be replicated in the two used slots to satisfy the required number of replicas of the first device controller Di.

[Operation when Provider Pi and Requester Ri Leave in Distributed Storage System]
Next, an operation for optimizing the distributed shared memory 43b when the provider Pi and the requester Ri are disconnected from the network NW in the distributed storage system described above will be described.

  As shown in FIG. 17A, when the requester R connected to the network NW is detached from the distributed storage system after being removed by the operator, as shown in FIG. 17A and FIG. A departure trigger occurs in the shared memory providing function 43 of the provider P4. This detachment trigger includes a device ID (detachment ID) to be detached.

  When the provider P4 detects the withdrawal trigger, the provider P4 broadcasts a withdrawal request including the device ID of the requester R to leave and a withdrawal command to the network NW. Here, the provider P1 and the provider P4 store the configuration description file CFGi of the requester R. Therefore, the provider P1 and the provider P4 delete the configuration description file CFGi of the requester R as shown in FIG.

  Further, the provider Pi and the requester Ri that leave the network broadcast a withdrawal request including its own device ID to the network NW at the time of the removal work by the operator, and notify the provider Pi that it is leaving, and the provider Pi The configuration description file CFGi may be deleted.

  In such a distributed storage system, as shown in FIG. 18, the provider Pi first performs a leaving trigger check operation by the shared memory providing function 43 in step ST31. The check operation of the departure trigger is performed every preset period.

  In step ST32, it is determined whether a separation trigger has been detected. When a leaving trigger is detected, the shared memory providing function 43 broadcasts a leaving request including a device ID to leave to the network NW. On the other hand, when the departure trigger is not detected, in step ST34, an operation of checking a signal received from the network NW is performed, and in step ST35, it is determined whether or not a departure request is received. If a withdrawal request has not been received, the process returns to step ST31. If a withdrawal request has been received, the process proceeds to step ST36.

  In step ST36, the provider Pi determines whether or not the device ID configuration description file CFGi included in the withdrawal request is stored in its own distributed shared memory 43b. If the ID of the provider Pi and requester Ri to leave is not stored, the process returns to step ST31. If the ID of the provider Pi and requester Ri to leave is stored, the device ID in step ST37. The configuration description file CFGi is deleted, and the process returns to step ST31.

  In this way, when the provider Pi and requester Ri leave the network NW, the withdrawal request including the device ID is broadcast to the network NW and the configuration description file CFGi of the device ID is deleted from the provider Pi. Another configuration description file CFGi can be stored in the slot in which the configuration description file CFGi is stored. Therefore, the number of copies of the configuration description file CFGi can be increased to further improve data loss tolerance. In addition, the distributed shared memory 43b in the distributed storage system can be used more effectively.

  The above-described embodiment is an example of the present invention. For this reason, the present invention is not limited to the above-described embodiment, and various modifications can be made depending on the design and the like as long as the technical idea according to the present invention is not deviated from this embodiment. Of course, it is possible to change.

  That is, the above-described distributed storage system has been described with respect to a system including the provider Pi and the requester Ri, and a system including only the device controller Di. However, files and data stored in devices such as sensors are stored in the distributed shared memory 1. You may make it carry out distributed storage.

1 is a system diagram showing a configuration of a distributed storage system to which the present invention is applied. FIG. 11 is a system diagram for explaining an operation when a configuration description file is exchanged in a distributed storage system to which the present invention is applied. (A) is an explanatory diagram in which a device controller is exchanged, (b) is an explanatory diagram in which an ID is added to the exchanged device controller and broadcasts a request, and (c) is a configuration in which the exchanged device controller is configured. It is explanatory drawing which receives the response containing a description file. 1 is a system diagram for explaining types of network devices in a distributed storage system to which the present invention is applied. FIG. It is a block diagram for demonstrating the characteristic operation | movement in the distributed storage system to which this invention is applied. It is a block diagram which shows the structure of a device controller in the distributed storage system to which this invention is applied. It is a block diagram which shows the detailed structure of a device controller in the distributed storage system to which this invention is applied. FIG. 2 is a block diagram showing a configuration of a device and fieldbus interface adopting EMIT technology in a distributed storage system to which the present invention is applied. It is a block diagram of a fieldbus table. FIG. 3 is a configuration diagram of a variable table included in a configuration description file stored in a device controller in a distributed storage system to which the present invention is applied. In the distributed storage system to which the present invention is applied, it is a configuration diagram of a linkage table included in a configuration description file stored in a device controller. FIG. 4 is a diagram showing software parameters included in a configuration description file stored in a device controller in a distributed storage system to which the present invention is applied. It is a block diagram which shows the functional structure in the distributed storage system to which this invention is applied. It is a block diagram explaining the content of the signal in the distributed storage system to which this invention is applied. It is a flowchart which shows the process of the distributed shared memory monitoring function in the distributed storage system to which this invention is applied. It is a flowchart which shows the process of the shared memory provision function in the distributed storage system to which this invention is applied. It is a figure for demonstrating operation | movement when a network device detaches | leaves in the distributed storage system to which this invention is applied. 5 is a flowchart for explaining an operation when a network device is detached in a distributed storage system to which the present invention is applied.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Distributed shared memory 11 P2P communication processing part 12 Storage control part 12A Information receiving part 12B Information update part 12C Information request part 12D Configuration description information memory 12E Information selection part 13 Distributed shared memory 14 Configuration description information driver 14a Configuration description file storage Unit 15 Variable table 15A Variable table update unit 16 Application processing unit 16A Application processing unit 16B Linkage table 16C Application update unit 17 Fieldbus monitoring / control unit 17A Fieldbus monitoring unit 17B Fieldbus control unit 18A, 18B, 18C Fieldbus interface 18a Fieldbus driver 18b Fieldbus interface unit 19 Provider ID storage unit 21 MOS function unit 21a Function 21b Event 21c Variable 2 d object ID
21e service table 22 application processing unit 23 interface module 24 capability table 31 OAS function unit 31a device access server 31b service information 31c device access controller 31d object ID
32 Communication Module 41 Distributed Shared Memory Monitoring Function 41a Active Monitoring Unit 42 Storage Control Function 42a Configuration Description Information Storage Unit 42b Shared Memory Status File 42c Backup Status File 42d Memory Status Detection Function Unit 42e Slot Question Function Unit 42f Slot Acquisition Function Unit 42g Replication number warning function part 42h Slot reallocation function part 43 Shared memory providing function 43a Request management function 43b Distributed shared memory 43c Memory status report output function part

Claims (5)

In an information storage system that has a plurality of network devices, and each network device stores operation setting information necessary for its own operation in other network devices,
Some or all of the plurality of network devices store operation setting information transmitted from other network devices and function as backup memory for operation setting information of the other network devices. Having means,
Each network device
Information storage means for storing its own operation setting information;
A communication means connected to the other network device via a communication line to communicate with the other network device;
Storage control means for storing the own operation setting information stored in the information storage means in a backup storage means of another network device,
The storage control means
The importance of the operation setting information of the own network and the operation setting information of the other network devices connected via the communication line is set in advance, and the operation setting information of less importance among the operation setting information of the plurality of network devices Store the operation setting information in the backup storage means so as to increase the number of copies of the operation setting information of the network device having a higher importance than the number of copies,
The storage control means acquires the memory state of the backup storage means from the network device having the backup storage means, and stores its own operation setting information for the number of copies based on the importance of the own operation setting information. If it is determined that it is impossible, the network device having the largest number of copies of the operation setting information than the number of copies based on the importance is determined, and the operation setting information of the network device is used instead of the operation setting information of the network device. An information storage system characterized by storing information. The network device having the backup storage means has a memory status report output means for notifying the status of its own backup storage means to other network devices,
The storage control means refers to the status of the backup storage means notified from the memory status report output means, and stores its own operation setting information in a network device having the backup storage means. The information storage system according to claim 1. The storage control means
Inquires of the network device having the backup storage means whether or not its own operation setting information can be stored, and other network devices that have received a response of usage status information indicating that its own operation setting information can be stored Sends its own operation setting information, stores the own operation setting information in the other network device,
The network device having the backup storage means returns the memory state of its own backup storage means when inquired from another network device whether or not the operation setting information of the other network device can be stored. The information storage system according to claim 1, wherein:   The network device having the backup storage means is self-backup when a network device newly participating in the communication line is connected and inquired whether or not operation setting information can be stored from the new network device. 2. The information storage system according to claim 1, wherein the memory information of the storage means is returned.   When the network device having the backup storage unit receives a request to discard the operation setting information in order to leave the communication line, the network device deletes the operation setting information of the network device to be removed from the backup storage unit. The information storage system according to claim 1, wherein:

Images