JP2014072631A - Duplex network control system and duplex network control method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a duplex network control system and a duplex network control method capable of easily determining in which system a failure occurred in a duplicated network.SOLUTION: A master node and a slave node comprise: a monitoring unit which confirms that mutual access through a second network is possible; and a failure determining unit in which whether a response to the access is received from a partner node or not is determined, access to the other node through the second network is made and the execution result is mutually transmitted and received through a first network when the response to the access is determined not to be received from the partner node, and which of the master node and the slave node has failure is determined by comparing the execution result of the transmission and reception as a result of monitoring by the monitoring unit. The other node comprises a processing unit which transmits the execution result with respect to the access when accessed from the master node and the slave node.

Description

  The present invention relates to a technique for controlling a duplexed network.

  Conventionally, in a system having a duplexed network configuration, there are various techniques for switching to another network when a failure occurs in one network. For example, in Patent Document 1, a communication state of a device is evaluated and digitized from communication state information acquired from each system device constituting the network, and a network with a smaller value has a lower operational risk. The network system is switched.

Patent No. 4641188

  In Patent Document 1 described above, if there is a response from each device within a certain time, it is determined that the communication state is normal, while if there is no response within a certain time, it is determined that the communication state is not normal. The switching of the system is determined based on the number of devices determined not to be normal. However, in Patent Document 1, since it is necessary to collectively determine the communication status for each system apparatus, there is a problem that it is not easy to determine which system has an abnormality.

  The present invention has been made in view of the above, and provides a duplexed network control system and a duplexed network control method capable of easily determining which system has failed in a duplexed network. The purpose is to do.

  In order to solve the above-described problems and achieve the object, a duplex network control system according to the present invention includes a master node, a slave node connected to the master node via a first network, and the master node. A duplex network control system in which another node communicating with the slave node is connected by a second network, wherein the master node and the slave node can access each other via the second network. As a result of monitoring by the monitoring unit that confirms that it is possible, as a result of monitoring by the monitoring unit, it is determined whether or not a response to the access has been received from the counterpart node; If determined, the access to the other node via the second network is performed. The failure determination unit determines whether the master node or the slave node has a failure by transmitting and receiving the execution results to each other via the first network and comparing the transmitted and received execution results. The other node includes a processing unit that transmits an execution result for the access when the access is received from the master node and the slave node.

  The present invention is also a duplex network control method performed in the duplex network control system.

  According to the present invention, it is possible to provide a duplexed network control system and a duplexed network control method that can easily determine which system has a failure in a duplexed network.

It is a figure which shows the structural example of the duplex network control system which concerns on this invention. It is a block diagram which shows the functional structure of a node. It is a block diagram which shows the functional structure of a master node. It is a figure which shows the example of the structure information of the node which the network which a memory | storage part has has. It is a figure which shows the structural example of the duplication network control system which concerns on this invention (at the time of access processing execution). It is a flowchart which shows the process sequence of a failure determination process. It is a flowchart which shows the process sequence of a failure determination process. It is a figure which shows the example of log output setting information.

  Exemplary embodiments of a duplex network control system and a duplex network control method according to the present invention will be explained below in detail with reference to the accompanying drawings.

  FIG. 1 is a diagram showing a configuration example of a duplex network control system according to the present invention. As shown in FIG. 1, the duplex network control system 1000 includes a plurality of nodes 100, a master node 200, a slave node 300, and networks N1, N2, and N3. The networks N1 and N2 are, for example, general communication networks such as a WAN (Wide Area Network) or a LAN (Local Area Network), and the network N3 connects between the master node 200 and the slave node 300. It is a dedicated LAN. First, the node 100 will be described.

  The node 100 is composed of an information processing apparatus such as a PC (Personal Computer) or a server, for example. As will be described later, the node 100 cannot access the master node 200 and the slave node 300 via the networks N1 and N2, and the cause cannot be determined by these nodes. The access processing (for example, processing for reading the shared memory of the slave node 300) executed by the other node 300 is performed by any one of the nodes 100 (basically, the master node 200 or the slave node). When executed on a node close to the node 300), the execution result of the access process is fed back to the master node 200 and the slave node 300.

  FIG. 2 is a block diagram showing a functional configuration of the node 100. As illustrated in FIG. 2, the node 100 includes a communication unit 101, a failure time processing unit 102, and a control unit 103.

  The communication unit 101 is a communication device such as a NIC (Network Interface Card), for example, and transmits / receives various information for communication between its own node and the other nodes 100, the master node 200, and the slave node 300. .

  The failure processing unit 102 receives the execution of the access processing from the master node 200 and the slave node 300 and transmits the result to the master node 200 and the slave node 300 via the communication unit 101.

  The control unit 103 is, for example, a CPU (Central Processing Unit), and controls the operation of each unit of the node 100 described above. The control unit 103 has a memory (not shown) and stores identification information (for example, address information such as an IP address and a MAC address) for identifying the own node. Next, the master node 200 will be described. Although only the master node 200 is described below, the slave node 300 has the same configuration.

  FIG. 3 is a block diagram showing a functional configuration of the master node 200. As illustrated in FIG. 3, the master node 200 includes a communication unit 201, a monitoring unit 202, a failure determination unit 203, a storage unit 204, and a control unit 205.

  The communication unit 201 is, for example, a communication device such as a NIC, and transmits / receives various information for communication between the own node and the slave node 300 and between the own node and the node 100.

  The monitoring unit 202 periodically acquires alive information (for example, key alive packet) from the slave node 300 for confirming whether or not the connection with the own node is in a valid state via the network N3. By doing so, the connection state of the slave node 300 which is the counterpart node is confirmed. For example, when the monitoring unit 202 confirms that the key alive packet is not received every 200 ms, the monitoring unit 202 determines that the slave node 300 is not in a valid state, and disconnects the connection with the slave node 300. Similarly, the monitoring unit of the slave node 300 also acquires the alive information from the master node 200 via the network N3 and confirms the connection state with the master node 200.

  Although not explicitly shown in FIG. 1, the master node 200 and the slave node 300 are connected by a bus (not shown), and the monitoring unit 202 monitors each other's start / stop state. For example, the monitoring unit 202 monitors whether or not the power of the slave node 300 is ON and whether or not the CPU is activated.

  In addition, the monitoring unit 202 periodically monitors the access process for the slave node 300 via the network N1 or N2, and determines whether an abnormality has occurred in the access process. For example, the monitoring unit 202 accesses the shared memory of the slave node 300 every 200 ms via the network N1 or N2, and determines whether the stored data can be read. Similarly to the monitoring unit of the slave node 300, access to the master node 200 is periodically monitored via the network N1 or N2, and it is determined whether an abnormality has occurred in the access. Yes.

  In addition, the monitoring unit 202 performs cyclic communication via the network N1 or N2, periodically transmits data stored in the shared memory of the own node to the slave node 300, and is stored in each other's shared memory. Monitor data synchronization. Similarly, the monitoring unit included in the slave node 300 also performs cyclic communication via the network N1 or N2, periodically transmits the data in the shared memory included in the own node to the master node 200, and is shared with each other. The synchronization of data stored in the memory is monitored.

  In addition, the monitoring unit 202 periodically monitors the nodes connected to the network N1 or N2, and causes the storage unit 204 to store node configuration information (described later) of the latest nodes connected to these networks.

  The failure determination unit 203 checks the connection state and start / stop state of the slave node 300 monitored by the monitoring unit 202, and determines whether or not the connection between the master node 200 and the slave node 300 is possible. Then, it is determined whether or not the power of the slave node 300 is ON or the CPU is activated, and processing for determining the cause (failure determination processing) and access processing are executed. Specific contents of the processing performed by the failure determination unit 203 will be described later using a flowchart.

  The storage unit 204 is a general memory, for example, and stores various types of information necessary for communication with the slave node 300 and the node 100.

  FIG. 4 is a diagram illustrating an example of node configuration information (for example, node configuration information such as a ring map) included in the network N1 or N2 stored in the storage unit 204. As shown in FIG. 4, the node configuration information stores node identification information for identifying a node, address information of the node, and configuration specifications (CPU, memory, disk capacity) of the node in association with each other. ing. In FIG. 4, there is a node identified by the node identification information “N101” as a node of the network N1, the IP address of the node is “123.456.012,” and the CPU, memory, and disk capacity are respectively It shows that it is 3.4 GHz, 2 GB, and 100 GB. This node configuration information is periodically updated to the latest state by the monitoring unit 202.

  The storage unit 204 has a role as a shared memory, and stores various pieces of information to be stored in synchronization between the master node 200 and the slave node 300 in addition to the node configuration information described above. As will be described later, when the master node 200 and the slave node 300 are inaccessible, the failure determination unit 203 selects the above-described node configuration information at that time point as shown in FIG. Access processing is executed for the node 100 (minimum value node) having the IP address closest to the master node 200 or the slave node 300.

  The control unit 205 is a CPU (Central Processing Unit), for example, and controls the operation of each unit of the master node 200. Next, failure determination processing performed in the duplex network control system 1000 will be described.

  6A and 6B are flowcharts illustrating the processing procedure of the above-described failure determination processing. In the following, the failure determination process in the master node 200 is described, but it is assumed that the slave node 300 performs the same process.

  As shown in FIG. 6A, first, the monitoring unit 202 of the master node 200 refers to the alive information received from the slave node 300, confirms whether the slave node 300 is connectable, and the communication determination unit 203 Then, it is determined whether or not the monitoring unit 202 has confirmed that the slave node 300 can be connected (step S601).

  When the failure determination unit 203 determines that the monitoring unit 202 has confirmed that the slave node 300 can be connected (step S601; Yes), the monitoring unit 202 further determines that the slave node 300 is powered on. It is determined whether or not the CPU has been activated (step S602).

  When the failure determination unit 202 determines that the monitoring unit 202 has not confirmed that the slave node 300 is connectable (step S601; No), or the monitoring unit 202 is in a power-on state of the slave node 300. If it is determined that the CPU activation state has not been confirmed (step S602; No), the slave node 300 is stopped or isolated from the network and is operated in a duplex network configuration. If it is not determined, the process returns to step S601 and waits as it is.

  On the other hand, the failure determination unit 203 determines that the monitoring unit 202 has confirmed that the slave node 300 is connectable (step S601; Yes), and the monitoring unit 202 is in the ON state of the slave node 300 and the CPU If the monitoring unit 202 determines that the active state of the slave node 300 has been confirmed (step S602; Yes), the monitoring unit 202 performs access processing to the shared memory of the slave node 300 via the network N1 or N2 (step S603). Then, it is determined whether or not the execution is successful (step S604). For example, the failure determination unit 203 determines whether or not the monitoring unit 202 has successfully executed the shared memory read process of the slave node 300.

  If the failure determination unit 203 determines that the execution of the access process by the monitoring unit 202 is successful (step S604; Yes), the failure determination unit 203 receives a response indicating that the access process executed by the slave node 300 is successful. It is determined whether or not it has been done (step S606). That is, each of the master node 200 and the slave node 300 receives a response to the effect that the access processing has been successfully performed with respect to the slave node 300 and the master node 200, respectively.

  On the other hand, when the failure determination unit 203 determines that the execution of the access processing by the monitoring unit 202 has not been successful (step S604; No), the failure determination unit 203 transmits a command for performing the access processing (for example, to read the shared memory). (Send command) cannot be executed, or the software of the master node 200 cannot be started, and it is determined that the access process cannot be executed due to an abnormality of the master node 200, and error processing such as outputting a log is executed Then, the master node 200 is stopped (step S605), and the failure determination process is terminated. By terminating the processing without performing the subsequent processing at this stage, it is possible to more quickly isolate the fault.

  If the failure determination unit 203 determines that a response indicating that the access process executed by the slave node 300 has been successfully executed is received (step S606; Yes), the communication via the networks N1 and N2 is normal. Judge, return to step S601 and wait. That is, if the access processing is successfully executed in both the master node 200 and the slave node 300 via the network N1 or N2, it is determined that there is no communication failure in the duplex network control system 1000, and the process stands by.

  On the other hand, when the failure determination unit 203 determines that a response indicating that the access processing executed by the slave node 300 has been successfully executed has not been received (step S606; No), the failure determination unit 203 sends the response via the network N3. A communication failure message indicating that it has not been received is transmitted to the slave node 300 to report that fact (step S607). Similarly, when the slave node 300 has not received the response from the master node 200, the slave node 300 transmits a communication failure message to the master node 200 and reports the fact. For example, if the master node 200 and the slave node 300 do not receive the response within a certain time, the master node 200 and the slave node 300 determine that the timeout has occurred and transmit a communication failure message.

  Then, the failure determination unit 203 refers to the node configuration information stored in the storage unit 204, determines whether or not another node 100 exists in the network N1 or N2 (step S608), and the network N1 or N2 If it is determined that there is no other node 100 (step S608; No), it is determined that it is difficult to determine the failure, error processing such as outputting a log is executed (step S609), and the failure is determined. Interrupt the process. In this case, it may be possible to return to step S601 again after a certain time has elapsed and re-execute the failure determination process, or notify the administrator to that effect and investigate the cause.

  On the other hand, when the failure determination unit 203 determines that another node 100 exists in the network N1 or N2 (step S608; Yes), the failure determination unit 203 refers to the node configuration information illustrated in FIG. The access process is executed for the value node, and the execution result is received (step S610). At this time, the slave node 300 similarly executes an access process for the other nodes 100 and receives the execution result. For example, each of the master node 200 and the slave node 300 uses the IP address of the other node (for example, the slave node 300 that is the other node in the case of the master node 200) that has been the target of the access process as the other node 100. The IP address and the protocol for communicating with the other node 100 are converted, the access process is executed for the other node 100, and the execution result is received. In this way, by performing the access process on the minimum value node, the failure determination process can be executed efficiently and promptly.

  In this embodiment, the failure determination unit 203 executes the access process for the minimum value node. However, if the access process is complicated and takes a long time, further refer to the node configuration information. The node 100 having a high configuration specification may be selected to execute the access process. In this case, the result of the access process can be quickly received even if the processing time is normal, and the entire failure determination process can be quickly executed.

  Then, the failure determination unit 203 transmits the execution result received from the other node 100 to the slave node 300 via the network N3 (step S611), and the slave node 300 similarly receives from the other node 100. The execution result is transmitted to the master node 200 via the network N3 (step S612).

  The failure determination unit 203 confirms the contents of the execution result received by the own node (master node 200) from the other node 100 and the contents of the execution result received from the slave node 300, and whether only the execution result of the own node is abnormal. (For example, whether or not the access process could not be executed and timed out) is determined (step S613), and when it is determined that only the execution result of the own node is abnormal (step S613; Yes), the master node 200 It is determined that the access process cannot be executed due to the abnormality, and an error process such as outputting a log is executed, the master node 200 is stopped (step S614), and the failure determination process is terminated.

  On the other hand, when the failure determination unit 203 determines that only the execution result of the own node is not abnormal (step S613; No), the failure determination unit 203 further determines whether only the execution result of the other node (slave node 300) is abnormal (step). S615) When it is determined that only the execution result of the other node is abnormal (step S615; Yes), it is determined that the access process cannot be executed due to the abnormality of the slave node 300, and error processing such as outputting a log is performed. The slave node 300 is stopped (step S616), and the failure determination process is terminated.

  On the other hand, when the failure determination unit 203 determines that there is no abnormality only in the execution result of the other node (step S615; No), the failure determination unit 203 executes normally on both the master node 200 as the own node and the slave node 300 as the other node. It is determined whether or not the result has been successfully received (step S617).

  If the failure determination unit 203 determines that the execution result has not been successfully received in both nodes (step S617; No), the failure determination unit 203 determines that the failure is in the entire network N1 or N2, and outputs a log. Error processing such as performing is executed (step S618), and the failure determination processing is interrupted. In this case, it may be notified to the administrator and the cause investigated.

  The failure determination unit 203 determines that there is no communication failure in the network N1 or N2 when it is determined that the execution result has been successfully received in both nodes (step S617; Yes), and further, the master node 200 and the slave node The contents of the log information acquired at each node 300 are read to determine which node has a failure, and the cause is determined (step S619).

  For example, the failure determination unit 203 reads log information when the communication unit 201 communicates with the slave node 300 or the node 100, and determines whether an error is recorded. The contents recorded as log information and the contents recorded as errors (error name, code number, etc.) are determined in advance as log output setting information.

  FIG. 7 is a diagram illustrating an example of log output setting information. As shown in FIG. 7, in the log output setting information, a code number for identifying an error code, the name of the error code, and a reference value that is an error based on the error code are associated with each other. In the example illustrated in FIG. 7, the error of the code number “1” is set to be an error when the number of times of normal frame reception from the slave node 300 is 10 times or less per minute. The error “2” indicates that the number of times that the frame is not normally received from the slave node 300 is one or more times per minute, for example, an error setting. The log output setting information is stored in advance in the storage unit 204, for example.

  In this manner, the failure determination unit 203 refers to the error recorded in the log information, determines whether or not a failure has occurred in the master node 200, and determines the cause based on the error name or code number. To do. Similarly to this, the slave node 300 also refers to the error recorded in the log information, determines whether or not a failure has occurred in the slave node 300 that is its own node, and determines its error name or code number according to the error name or code number. Determine the cause. Therefore, even when the failure cannot be determined based on the execution result of the access process, the failure can be isolated. The determination result is transmitted and received between the master node 200 and the slave node 300, and information is shared with each other.

  In the present embodiment, the cause of the failure is determined with reference to the error code as described above. However, the master node 200 and the slave node 300 are synchronized with each other and have the same contents as long as they are in a normal state. Therefore, it is determined whether there is a difference in the output amount of log information in a predetermined time (for example, 1 minute), so that either the master node 200 or the slave node 300 has a failure. It is also possible to determine whether it has occurred. As a result, when there is no need to identify the specific cause of the failure, it is possible to quickly identify the node where the failure has occurred.

  Then, the failure determination unit 203 compares the log information output from the master node 200 that is the own node with the log information received from the slave node 300 that is the other node, and an error is recorded only in the master node 200 and an error is detected. Or not (step S620). If it is determined that an error is recorded only in the master node 200 and is abnormal (step S620; Yes), the error of the master node 200 is caused by the error. Error processing such as outputting a log is executed, the master node 200 is stopped (step S621), and the failure determination process is terminated.

  On the other hand, when the failure determination unit 203 determines that no error is recorded in the master node 200 and that there is no abnormality (step S620; No), an error is recorded only in the slave node 300, resulting in an abnormality. (Step S622), when an error is recorded only in the slave node 300 and it is determined that an error has occurred (step S622; Yes), the error of the slave node 300 occurs due to the error. Error processing such as outputting a log is executed, the slave node 300 is stopped (step S623), and the failure determination process is terminated.

  Then, the failure determination unit 203 indicates that no error has been recorded in the master node 200 and no abnormality has occurred, and no error has been recorded in the slave node 300 and no abnormality has occurred. If it is determined (step S622; No), as in the case of step S618, it is determined that there is a failure in the entire network N1 or N2, error processing such as outputting a log is executed (step S624), and failure determination processing End. In this case, it may be notified to the administrator and the cause investigated. When the process of step S624 is finished, all the fault determination processes shown in FIGS. 6A and 6B are finished.

  When each of the processes in steps S605, S614, and S621 is performed, the master node 200 has an abnormality. Therefore, a switching process such as taking over the same IP address as that of the master node 200 is performed on the slave node 300 side. Is switched to the slave node 300 side.

  On the other hand, when each process of steps S616 and S623 is performed, there is an abnormality in the slave node 300. When each process of steps S609, S618, and S624 is performed, a network other than the master node 200 and the slave node 300 is used. Therefore, the failure determination processing is terminated by interrupting the failure determination processing or executing error processing such as outputting a log, and then the administrator or the like responds to the failure.

  In this way, the master node 200, the slave node 300 connected to the master node 200 via the network N3, and the other nodes 100 communicating between the master node 200 and the slave node 300 are the network N1 or N2. In the connected dual network control system, the master node 200 and the slave node 300 confirm that they can access each other via the network N1 or N2, and the partner node as a result of monitoring by the monitoring unit 202 If it is determined whether or not a response to access has been received from the other node, and it is determined that a response to access has not been received from the other node, access is made to another node 100 via the network N1 or N2 and executed Network N results to each other And a failure determination unit that determines which of the master node 200 and the slave node 300 has a failure by comparing the transmitted and received execution results, and the other nodes 100 are connected to the master node 200. And the failure time processing unit 102 that transmits the execution result for the access when receiving an access from the slave node 300, it is easy to determine in which system the failure has occurred in the duplexed network. It becomes possible. Further, unlike the prior art, it is not necessary to grasp the communication state of each device connected to each system of the duplexed network, so that it is possible to isolate the failure determination with a simple configuration.

  In addition, this invention is not limited to an above-described Example, Various modifications are included. For example, the above-described embodiments have been described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described. Further, a part of the configuration of one embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of one embodiment. Further, it is possible to add, delete, and replace other configurations for a part of the configuration of each embodiment.

1000 Duplex Network Control System 100 Node 101 Communication Unit 102 Failure Processing Unit 103 Control Unit 200 Master Node 201 Communication Unit 202 Monitoring Unit 203 Failure Determination Unit 204 Storage Unit 205 Control Unit 300 Slave Nodes N1 to N3 Network

Claims (7)

A duplex network in which a master node, a slave node connected to the master node via a first network, and another node communicating between the master node and the slave node are connected by a second network A control system,
The master node and the slave node are
A monitoring unit for confirming that each other can be accessed via the second network;
As a result of monitoring by the monitoring unit, it is determined whether or not a response to the access has been received from the counterpart node, and if it is determined that a response to the access has not been received from the counterpart node, Thus, the access to the other node is performed, and the execution results are transmitted / received to / from each other via the first network, and a failure occurs in either the master node or the slave node by comparing the execution results transmitted / received. A failure determination unit for determining whether there is,
The other nodes are
A processing unit that transmits an execution result for the access when the access is received from the master node and the slave node;
A duplex network control system comprising: The master node and the slave node are
It further includes a storage unit that stores log information of processing performed in its own node,
The failure determination unit determines whether the master node or the slave node has a failure by further comparing the log information when the execution result is normally received from the other node. To
The duplex network control system according to claim 1, wherein: The failure determination unit determines that the master node or the slave node has an abnormality when the execution result received from the other node cannot be received by the master node or the slave node due to timeout. To
The duplex network control system according to claim 1 or 2, wherein The failure determination unit, as a result of monitoring by the monitoring unit, if it is determined that the access is not possible from the counterpart node, it determines that its own node is abnormal without performing the access to the other node,
The duplex network control system according to any one of claims 1 to 3, wherein: The failure determination unit receives the execution result by performing the access to a minimum value node among the other nodes connected to the second network.
The duplex network control system according to any one of claims 1 to 4, wherein: The failure determination unit receives the execution result by performing the access to a node having the highest specification among the other nodes connected to the second network.
The duplex network control system according to any one of claims 1 to 5, wherein: A duplex network in which a master node, a slave node connected to the master node via a first network, and another node communicating between the master node and the slave node are connected by a second network A duplex network control method performed in a control system,
A monitoring step for confirming that each other can be accessed via the second network;
As a result of monitoring in the monitoring step, a determination step of determining whether or not a response to the access has been received from the counterpart node;
A first access step of performing the access to the other node via the second network when it is determined in the determination step that a response to the access is not received from the counterpart node;
A processing step of transmitting an execution result for the access when the access is received from the master node and the slave node in the first access step;
A second access step of transmitting and receiving the execution results to and from each other via the first network when the execution results are received in the processing step;
A failure determination step of determining whether the master node or the slave node has a failure by comparing the execution results transmitted and received in the second access step;
A duplex network control method comprising:

Images