WO2001037490A1

WO2001037490A1 - Communication path management system, communication path managing apparatus, and computer-readable recorded medium on which communication path managing program is recorded

Info

Publication number: WO2001037490A1
Application number: PCT/JP1999/006442
Authority: WO
Inventors: Minoru Harada; Satoshi Kajiyashiki; Hitoshi Matsumori; Masato Semii; Toshiharu Kawanishi
Original assignee: Fujitsu Limited; Izuta, Naoki
Priority date: 1999-11-18
Filing date: 1999-11-18
Publication date: 2001-05-25

Abstract

A communication path management system comprising a currently-operating communication path (PH1), a stand-by communication path (PH2) additionally provided to the currently-operating communication path (PH1), nodes (301 to 3012) connected to the currently-operating communication path (PH1) in normal operation and adapted to conduct intercommunication with each other and to conduct re-transmission multiple times if a communication error occurs, a termination node (30e), and a management node (10) which polles the termination node (30e) every predetermined time and switches the communication path from the currently-operating communication path (PH1) to the stand-by communication path (PH2) when a fault of the currently-operating communication path (PH1) is detected from the results of the response, whereby the burden on the managing personnel is lightened, the cost is lowered, and the reliability is improved.

Description

Description Communication path management system, communication path management device, and computer-readable recording medium storing communication path management program

The present invention relates to a communication path for connecting a plurality of nodes, wherein the communication path management system manages a duplicated communication path having at least an active system and a standby system, a communication system, a communication management device, and a communication system. The present invention relates to a computer-readable recording medium on which a storage management program is recorded. Background art

In recent years, instead of the conventional stand-alone type of computer, a form in which a plurality of computers are connected via a communication path under the concept of network-combining has become mainstream. In this type of connection, the communication path is at least duplicated between the active system and the standby system, and when an error occurs in the communication path, the reliability is improved by switching from the active system communication path to the standby system communication path. . FIG. 13 is a block diagram showing a configuration of a conventional communication path management system. In this diagram, nodes 1, ~ 1 _N to convenient N nodes between the working communication paths P, or the communication path management system that connects the standby communication path P ₂ is shown. Each of the nodes 1 to 1 _N corresponds to a communication network card or the like to be inserted into an expansion slot of a computer (not shown), and has an active communication path P or a standby communication path P ₂ . Sends and receives data to and from the communication partner node via the network.

Node 1, is the primary system reception port Bok 2, and standby transmission port 3, includes a node 1 ₂ working system transceiver port - provided with the door 2 ₂ and the standby transmission port 3 2. In the same manner, the node receives the active transmission / reception port 2 _L and the standby Node 1 M has an active transmission / reception port 2 _M and a standby transmission / reception port 3 M. The node 1 _N also includes a primary system reception port 2 _N and standby transmission port 3 _N. Thus, node 1, each of ~ 1 _N, and a sending and receiving ports of two systems of current use system transmission port and the standby system transmission port.

The working communication path P, which is used as a communication path during normal operation, interconnects the working transmission / reception ports 2, _2N . On the other hand, the standby communication path P _2, when a communication failure occurs in the working communication path for some reason, working communication path [rho, is a communication path for the backup to be used in place of. The standby communication path P 2 is the standby system transmission port 3 when communication failure occurs, interconnecting to 3 _N.

Further, each of the nodes to ~ 1 _N, the management of the communication path used to communicate, understand Roh once configuration, a communication path management program for performing switching control of the communication path is implementation Bok. Therefore, node 1! 1 to 1 _N perform communication path management and the like based on the communication path management program. The management of the communication path as used herein means managing the communication state of the active communication path P 1 and the standby communication path P ₂ . The grasping of the node configuration, location of the node ~ 1 _N (Adoresu), connection order mechanism, means to grasp the connection state. Information on this node configuration is implemented in advance in each of the nodes 1 i to 1 _N by the administrator in advance. Communication path switching control refers to switching the communication path from the active communication path P 1 to the standby communication path P ₂ when a communication failure occurs.

In the above configuration, for example, if communication between the node and his node 1 _N, working communication path is used. That is, the node selects the active transmission / reception port 2, as the used port, and then transmits data to the node 1 _N from the active transmission / reception port 2,. Thus, the Isseki de, is received in the working system transmission port 2 _N nodes 1 _N and through the working communication paths.

In this case, node 1 of the standby transmission port 3, and transmission node 1 _N standby Receiving port 3 _N is not used for communication, but is in a state where data can be received.

Then, the node 1 _N receives the data from the node Mr. transmits a de one evening for responses from the working transmission port 2 _N to node Mr. As a result, the response data is received by the active transmission / reception port 2 of the node 1 via the active communication path. Thereafter, between the node and his node 1 _N, communication is performed via the working communication paths.

Here, if a failure occurs in the active communication path P, during communication of data from the node to the node 1 _N and communication becomes impossible, the data is transmitted to the active transmission / reception communication port of the node 1 _N. 2, not received. Therefore, in this case, from node 1 _Ν to node 1! Since no response data is sent to the node, the node determines from the reception timeout that the failure has occurred in the active communication path. Accordingly, node 11, after switching the transmission port active system reception port Bok 2, standby transmission port 3 from the, the standby transmitting and receiving i port 3, resend the data from the node 1 _N. That is, in this case, is the communication path is switched from working communication paths to the standby communication path P _2. Node 1 notifies the administrator that a failure has occurred in the active communication path.

The retransmitted data is received by the standby system transmission port 3 _N nodes 1 _N via the standby communication path P _2. This standby node 1 _N, after the communication paths recognizes that it has been switched current for system communication path P, from the standby communication path P _2, a reception port to be used for communication from the working transmission port 2 _N switch to the system transmit and receive port 3 _N. Next, the node 1 _N transmits response data to the node from the standby transmission / reception port 3 _N instead of the active transmission / reception port 2 _N.

Thus, the data standby communication path P ₂ node via said standby transmission port 3, is received. Thereafter, between the node 1 and the node i _N, failure is instead working communication paths P, generated, communicate via standby communication path P ₂ is performed. Note that the above-mentioned nodes and nodes In the same manner as the de 1 _N, switching control of the communications path, switching of communication paths is performed. Incidentally, in the conventional communication path management system as described above, sometimes failed node 1, because doing each is independently switching control of the communications path ~ 1 _N, is a complex protocol with the switching control . Therefore, in the conventional communication path management system, the data capacity of the communication path management program for realizing a complicated protocol naturally increases, so a large program storage area must be secured.

Also, in the conventional communication path management system, the administrator sends information about the node configuration in advance to the node 1! I mentioned the fact that it is necessary to implement ment to ~ 1 _N. For this reason, in the conventional communication path management system, the administrator has to update the information on the node configuration each time the number of nodes increases or decreases and the operating state of the nodes, which imposes a heavy workload. There was a problem of being large.

Further, in the conventional communication path management system, node 1, to 1 for the _N Te to Baie must Inpurimento information relating to the same communication path management Puroguramuyanoichido configuration, the cost of memory to store these And the cost of maintenance was very high.

In addition, it was stated that the conventional communication path management system has a function to notify the administrator that a failure has occurred in the communication path. In addition, in conventional communication path management systems, the administrator must search for the location where a failure has occurred, so it takes unnecessarily long time to recover from the failure and the reliability is low. Atsu 7 Disclosure of the Invention

Accordingly, the present invention provides a communication path management system, a communication path management device, and a computer recording a communication path management program, which can reduce the workload of an administrator, reduce costs, and improve reliability. The purpose is to provide a readable recording medium. The communication path management system according to the present invention includes a first communication path (corresponding to an active communication path according to an embodiment described later) and a second communication path (to be described later) provided in parallel with the first communication path. and corresponds to the standby communication path PH ₂ of the embodiment), usually carried out respectively connected between mutual communication in the first communication path when a luck, a plurality of nodes to perform retransmission at a communication error occurs plurality of times mutually between (described later to an embodiment of the node 3 0! to 3 0 _{1 2} and the terminal node 3 0, corresponds to), and is connected to the first communication path during normal operation of the plurality of nodes A management node for performing inter-communication (corresponding to a management node 10 of an embodiment described later), and the management node is a terminal node of the plurality of nodes (an embodiment described later) predetermined time interval to the corresponding) to the terminal node 3 0 _e Abnormality detecting means (corresponding to a path abnormality detecting section 15 of an embodiment described later) for detecting an abnormality in the first communication path based on the response result, and Switching control means for switching a communication path from the first communication path to the second communication path when the communication path is detected (corresponding to a path switching control unit 17 of an embodiment described later). And According to the present invention, during normal operation, each of the plurality of nodes performs mutual communication via the first communication path. Further, the abnormality detecting means polls the terminal node at predetermined time intervals, and when a response corresponding to the polling is received, the first communication path is assumed to be normal.

Here, if an error occurs in the first communication path for some reason, the response corresponding to the polling from the terminal node is lost, so that the abnormality in the first communication path is detected by the error detecting means. Thereby, the switching control means switches the communication path from the first communication path in which the abnormality has occurred to the second communication path. Thereafter, each of the plurality of nodes performs mutual communication via the second communication path instead of the first communication path.

Also, if an error has occurred in the first communication path and mutual communication has been performed between a pair of nodes before switching by the switching control means, an error has occurred in the first communication path. Therefore, a communication error occurs. As a result, the source node executes retransmission a plurality of times after a predetermined time has elapsed. If retransmission is performed after switching by the switching control means, the normal second communication path can be used, and normal communication between one set of nodes can be performed without a communication error. Is performed. In other words, in this case, the source node automatically switches the communication path from the first communication path to the second communication path while retransmission is performed several times due to the occurrence of a communication error. It was.

As described above, according to the present invention, only the management node centrally manages the first communication path (abnormality monitoring) and performs switching control when an abnormality occurs in the first communication path. Since there is no need to provide a node with a management function and a switching control function, costs can be reduced and reliability can be improved.

In addition, the communication path management device according to the present invention includes a first communication path (corresponding to an active communication path PH, according to an embodiment described later) and a second communication path provided in parallel with the first communication path. (Corresponding to a standby communication path PH ₂ in one embodiment described later) and connected to the first communication path during normal operation to perform mutual communication, and execute retransmission a plurality of times when a communication error occurs. a plurality of nodes a communication path management device applied to (one embodiment described below node 3 0, to 3 0 _{1 2} and corresponds to the terminal node 3 0 _e) and consisting of a communication path management system, normal operation A communication means (corresponding to a communication unit 11 of an embodiment described later) which is connected to the first communication path and performs mutual communication with the plurality of nodes; To the terminal node at specified time intervals Abnormality detecting means (corresponding to a path abnormality detecting unit 15 of an embodiment described later) for performing polling via communication means and detecting an abnormality in the first communication path based on the response result; A switching control means for switching a communication path from the first communication path to the second communication path when an abnormality is detected by the abnormality detecting means (corresponding to a path switching control unit 17 of an embodiment described later) And characterized in that:

According to the present invention, during normal operation, each of the plurality of nodes performs mutual communication via the first communication path. Further, the abnormality detecting means is provided with a terminal node at predetermined time intervals. Polling is performed on one node, and if there is a response corresponding to this polling, it is assumed that the first communication path is normal.

Here, if an error occurs in the first communication path for some reason, the response corresponding to the polling from the terminal node is lost, so that the abnormality in the first communication path is detected by the error detecting means. Thereby, the switching control means switches the communication path from the first communication path in which the abnormality has occurred to the second communication path. Thereafter, each of the plurality of nodes performs mutual communication via the second communication path instead of the first communication path. Also, if an error has occurred in the first communication path and mutual communication has been performed between a pair of nodes before switching by the switching control means, an error has occurred in the first communication path. Communication error. As a result, the source node executes retransmission a plurality of times after a predetermined time has elapsed.

If retransmission is performed after switching by the switching control means, the normal second communication node can be used, so that a communication error does not occur between a pair of nodes. Normal communication is performed. In other words, in this case, the source node automatically switches the communication path from the first communication path to the second communication path while retransmission is performed several times due to the occurrence of a communication error. It was.

As described above, according to the present invention, since the management of the first communication path (abnormality monitoring) and the switching control at the time of occurrence of an abnormality in the first communication path are performed in a centralized manner, a plurality of other Since it is no longer necessary to provide a management function and a switching control function to the system, costs can be reduced and reliability can be improved.

The communication path management device according to the present invention, in the communication path management device, polls the plurality of nodes at predetermined time intervals via the communication means, and based on the response result, It is characterized by including a node configuration recognizing means for recognizing the configuration (corresponding to a node configuration information recognizing unit 13 of an embodiment described later).

According to the present invention, polling is performed from the node configuration recognizing means to a plurality of nodes at predetermined time intervals via the communication means. And the node configuration recognizing means, The node configuration is recognized by recognizing a node that responded to polling and not recognizing a node that did not respond to polling.

As described above, according to the present invention, the node configuration recognizing means recognizes the node configuration at predetermined time intervals based on the response result of the polling. The work load on the administrator can be reduced as compared with the case where the administrator implements the configuration.

Further, in the communication path management device according to the present invention, in the above communication path management device, when the switching control unit switches the communication path from the first communication path to the second communication path, the communication; The second communication source is fixed to the second communication source.

According to the present invention, the switching control means fixes the communication path to the normal second communication after the occurrence of the abnormality, so that the mutual communication after the switching can be continuously performed.

Further, in the communication path management device according to the present invention, in the communication path management device described above, when an abnormality of the first communication path is detected by the abnormality detection means, the communication path is communicated to a node included in the node configuration. A search means (corresponding to a fault location search unit 16 of an embodiment described later) for performing polling via the means and searching for an abnormal location on the first communication path based on the response result. It is characterized by the following. According to this invention, when the abnormality of the first communication path is detected, the search means polls the nodes included in the node configuration recognized by the node configuration recognition means. Then, the search means searches for an error occurrence location on the first communication path by grasping the node that has responded to the polling and the node that has not responded to the polling.

As described above, according to the present invention, after an abnormality has occurred on the first communication path, the abnormality occurrence location is automatically searched for by the search means, so that the administrator has conventionally searched for the abnormality occurrence location. The time required for the search can be reduced as compared with the case. Further, in the communication path management device according to the present invention, in the communication path management device described above, the search means creates a binary tree from the nodes included in the node configuration, based on a binary tree search method using the binary tree. And searching for an abnormality occurrence location on the first communication path.

According to the present invention, using a very efficient search method called a binary tree search method,

(1) Since an error occurrence location on the communication path is searched, the time required for the search can be drastically reduced.

Further, a computer-readable recording medium recording the communication / path management program according to the present invention includes: a first communication path; a second communication path provided in parallel with the first communication path; A communication consisting of a plurality of nodes connected to the first communication path and performing mutual communication, and executing retransmission multiple times when a communication error occurs. Communication applied to the security management system,. A polling of the last node among the plurality of nodes at predetermined time intervals, and based on a response result, determines whether the first communication path is abnormal. An abnormality detection step (corresponding to step SA1 to step SA3 of an embodiment described later) to be detected, and when an abnormality is detected in the abnormality detection step, the communication path is changed from the first communication path to the second communication path. 2 is a communication path management program for causing a computer to execute a switching control step (corresponding to step SA4 of an embodiment described later) for switching between communication and communication.

According to the present invention, during normal operation, each of the plurality of nodes performs mutual communication via the first communication path. In the abnormality detection step, polling is performed on the terminal node at predetermined time intervals, and if there is a response corresponding to this polling, the first communication path is determined to be normal.

Here, if an error occurs in the first communication path for some reason, the response corresponding to the polling from the terminal node is lost, so that the abnormality in the first communication path is detected in the error detection step. Thereby, in the switching control step, the communication path is switched from the first communication path in which the abnormality has occurred to the second communication path. Hereafter, multiple nodes Communicate with each other via the second communication path instead of the first communication path. In addition, if an error occurs in the first communication path and if mutual communication is performed between a pair of nodes before switching is performed in the switching control process, an error occurs in the first communication path. Communication error. Thus, the transmission source node executes retransmission a plurality of times after a lapse of a predetermined time.

Then, if retransmission is performed after switching is performed in the switching control process, a normal second communication path can be used, so that normal communication between a pair of nodes can be performed without a communication error. Done. In other words, in this case, the source node automatically switches the communication path from the first communication path to the second communication path while retransmission is performed several times due to the occurrence of a communication error. It was.

As described above, according to the present invention, since the management of the first communication path (abnormality monitoring) and the switching control at the time of occurrence of an abnormality in the first communication path are performed in a centralized manner, a plurality of other Since it is no longer necessary to provide a management function and a switching control function to the system, costs can be reduced and reliability can be improved. BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 is a block diagram showing a configuration of an embodiment according to the present invention, FIG. 2 is a block diagram showing a configuration of a management node 10 shown in FIG. 1, and FIG. FIG. 4 is a diagram showing the data structures of the live check packets AC, to AC ₁₂ and AC _e and the response buckets ACK, to ACK ₁₂ and ACK _e shown in FIG. 1. node configuration information table TB used in the embodiment, a diagram showing the contact and path information table TB _2, Fig. 5, the node _{3 0 i (3 0 2 ~3} 0 1 2 shown in FIG. 1, 3 is a block diagram showing parentheses, FIG. 6 is a block diagram illustrating the operation of the same embodiment, and FIG. 7 is a block diagram of the path abnormality detection unit 15 shown in FIG. a flow chart for explaining the operation, FIG. 8, the node 3 0 which shows in FIG. 6, to 3 0 ₁₂ Oyobi 3 0 ₆ theory the respective operations of To Furochiya - a DOO, 9 figures illustrating the operation of the node configuration information recognizing unit 1 3 shown in FIG. 2 FIG. 10 is a flowchart for explaining the operation of the fault occurrence point searching unit 16 shown in FIG. 2. FIG. 11 is a flowchart showing a binary tree search method used in the same embodiment. FIG. 12 is a block diagram showing a modification of the same embodiment, and FIG. 13 is a block diagram showing a configuration of a conventional communication path management system. BEST MODE FOR CARRYING OUT THE INVENTION

The present invention will be described in more detail with reference to the accompanying drawings. FIG. 1 is a block diagram showing a configuration of an embodiment. In this diagram, management node 10 and node 30! A ~ 3 0 _{1 2} and the terminal node 3 0 _e, earthenware pots conveniently 1 four Roh one de between the working communication path PH or communication path management system interconnecting the standby communication path PH _2, is shown .

Management node 10, node 30, ~ 30! Each of the terminal node ₂ and the terminal node 30 _e corresponds to a communication network card or the like inserted into an expansion slot of a computer (not shown), and is used as an active communication path or a standby communication path. Data is transmitted / received to / from the communication partner node via the path PH2. These management nodes 10 and nodes 30i to 30! ₂ and the terminal node 30 _e are each assigned a unique address.

The path switch 20 _m , the path switch 20, to 20! ₂ and the path switch 20 _e are respectively a management node 10 and a node 30! ~ 30! ₂ and the end node 30 _e , and the communication path is changed from the working communication path PH, to the standby communication path PH 2 (or the standby communication path PH ₂ , etc.). Switch to PH,). The path switching unit 20 _ra , the path switching units 20 to _{20 12,} and the path switching unit 20 _e are provided with switching control signals from the path switching control unit 17 (see FIG. 2) of the management node 10. Collectively controlled by S.

The active communication path is used as a communication path during normal operation. The path switch 20 _m and the path switch 20! ~ 2 0 _{1 2} and the path switch 20 _e Management node i 0, node 30, ~ 30! 2, the mutual connecting each terminal node 3 0 _e. On the other hand, the standby communication path PH ₂ is a backup communication path used in place of the active communication path P when a communication failure occurs in the active communication path for some reason. The standby communication path PH _2, the path switcher 20 _m, the path switcher 2 0 - 2 through 0 ₁₂ and the path switcher 2 0 _e, the management node 1 0, node 30, 1-3 0, ₂ management node 1 0 interconnecting each terminal node 3 0 _e is the automatic recognition of node configuration for the node 3 0 _2-3 0 ₁₂ and terminal node 3 0 _e, the detection of the path abnormality, the point of failure It performs search, path switching control when a route is abnormal, and the like.

Here, the configuration of the management node 10 will be described in detail with reference to FIG. In the management node 10 shown in this figure, the communication unit 11 is connected to a path switch 20 _ra (see FIG. 1), and according to a predetermined communication protocol, the nodes 30,. ₁₂ and end node 30. And communication control between them. More specifically, the communication unit 11 transmits the alive check buckets Ad to AC ₁₂ and the alive check bucket AC, to a node 30! ~ 30! ₂ and end node 30 _e .

Each of these § live check packet AC i to Ac _12, the node 3 with respect to 3 ₍₂ in node status (operation stopped, absent) and communication status (normal, abnormal, disabled, node In addition, the alive check packet Ad to AC ₁₂ is a packet for checking an abnormal point (failure point on the active communication path PH, (or standby communication path PH ₂ )). Also used when searching for.

Above § live check bucket DOO AC, each to Ac _12, as shown in FIG. 3 (a), a node 3 0 of the destination, and each address AD A of ~ 3 0, node status or It is composed of a command C _a for the communication check.

On the other hand, the alive check packet AC e is a packet transmitted to the terminal node 30 _e, and the working communication path P between the management node 10 and the terminal node 30 _e. H, a packet for checking the path state (or standby communication path PH ₂₎ (normal, abnormal). Packet AC _e for this § live check, as shown in FIG. 3 (a), consists of the address AD _A destination terminal nodes 30 _e, a command C _A for the path state check I have.

Further, the communication unit 11 receives the response buckets ACK, ACKACK ₁₂ and ACK e transmitted from the nodes 30 to 30, ₂ and the terminal node 3 (K (see FIG. 1)). response for buckets preparative ACK, ～ACK ₁₂ corresponds to Araibuchietsu click for bucket preparative Ad to Ac ₁₂ described above, and buckets bets for receiving responses. Also, bucket ACK for these responses, ～ACK ₁₂ , as shown in FIG. 3 (b), the address AD _B of the management node 1 0 destination, a command C _B for response, node 30, 30, ₂ of each state (operation FS

B (0: operating, 1: stopped).

On the other hand, the response bucket ACK, is a response bucket corresponding to the live check bucket AC. The response for bucket preparative ACK _e, the management Roh destination - the state shown the address of de 1 0 AD _B, and Command C _B for response, (in operation, stopped) condition of the terminal node 3 0 _beta the flag FS _B (0: operating, 1: stopped) is composed from a.

Returning to Figure 2, the command determination unit 1 2, received response for Pas Kek preparative AC K by the communication unit 1 1, the command C _B (FIG. 3 of ～ACK ₁₂ and response for buckets preparative ACK _e (b )), And notifies the node configuration information recognizing unit 13, route error detecting unit 15, and fault location searching unit 16 of the result of the judgment.

Node configuration information recognition unit 1-3, Araibuchi X click for packet AC _T ~ eight Ji ₁₂ nodes 3 0 via the communication unit 1 1, was sent to ~ 3 0, _2, response for bucket preparative ACK! according existence and content of the reception of ~ACK _12, recognizes the node configuration of the communication path management system shown in Figure 1. The information of the node configuration here includes nodes 30, −30, ₂ and terminal node 3 (connection order of K, address, node state, communication state, and the like. Further, the node configuration information recognizing unit 13 reflects the recognition result of the node configuration in the node configuration information table TB, which is stored in the storage unit 14 (see FIG. 4 (a)). The node configuration information table TBt are those used in the management of the node configuration, the connection order information CN _C, and an address AD _C, node status flag FNc and the communication status flag FC _C. Connection order information CN _C a node 3 0 is information indicating the respective order of connection to 3 0, ₂ and the terminal node 3 0 _e.

Specifically, the connection order information CN _C is Ru node 3-3 0 ₁₂ and the terminal node 3 Oe to the node number assigned respectively (1) - (1 3) der shown in Figure 1 . Here, the node with the youngest node number exists at a position near the management node 10. On the other hand, the node with the oldest node number is located far from the management node 10. Therefore, the node 30 of the node number (1) is located closest to the management node 10, and the terminal node 30 _e of the node number (13) is located farthest from the management node 10. Exists.

Returning to 4 (a), Adoresu AD _C, the node 3 0, -3 0 ₁₂ and end node 3 0, an address of which is assigned to each. The node status flag FN c is set to nodes 30, ~ 30! ₂ and each state of the terminal node 3 0 _e (0: Dynamic Sakuchu, 1: Down, 2: not present) is a flag indicating the. The communication status flag FCc is a flag indicating the status of the communication between the nodes (0: normal, 1: abnormal, 2: invalid (when no node exists)).

Returning to FIG. 2, the path abnormality detection unit 15 transmits the alive check packet AC _e to the terminal node 30β via the communication unit 11 and determines whether or not the response bucket ACK _e has been received. Detects a path error in the active communication path (or standby communication path PH ₂ ). The route abnormality detecting section 1 5, Ru to reflect the detection result of the route the abnormality in the storage unit 1 4 path is stored in the information table TB ₂ (see FIG. 4 (b)). The path information table TB ₂ is working communication paths Roita! And standby communication path PH ₂ Information (path number, the path state, the abnormal point address) and because of a table for managing, routing number LN _D, Path status flag FL _D , error location address Scan ADN _D, and a anomaly address AD F _D.

Routing number LN _D is a working communication paths and their respective on standby communication path PH ₂ granted number. Path state flag FL _D is the active system each柽路state of the communication path PHi Contact and standby communication path PH ₂ (0: normal, 1: abnormal) is a flag indicating.

The error location address ADN _D is a node closest to the relevant location and the management node 1 when an error occurs in any location on the working communication path PH (or the standby communication if path PH ₂ ). This is the address of the node near 0. The abnormal location address AD FD is an address of a node that is the nearest node to the location where the abnormality has occurred and that is far from the management node 10. That is, abnormality occurs point is present between the abnormal point address ADN _D is assigned a node, the abnormality location address AD F _D is assigned the node.

Returning to FIG. 2, the fault occurrence point searching unit 16 performs a well-known binary tree search when the path abnormality detecting unit 15 detects a path abnormality of the working communication path (or the standby communication path PH ₂ ). law, bucket preparative ACt to Ac ₁₂ and response for buckets ACK for Araibuchiwekku, searching for a point of failure (anomaly) using ~ACK _12. The details of the operation of the failure location search unit 16 will be described later.

The path switching control unit 17 is controlled by the path abnormality detection unit 15 and the fault occurrence point searching unit 16 to switch control signals to the path switching units 20, 220 ₁₂ and 20, shown in FIG. Output S. The switching control signal S, the communication path working communication path PH, to the standby system communication path PH ₂ from or is a signal for switching from the standby communication path PH ₂ working communication paths ΡΗ to.

Next, with reference to FIG. 5, the configuration of the node 30 shown in FIG. 1 will be described in detail. In the node 30 shown in this figure, the communication unit 31 includes a path switch 20! Is connected to the (first refer to FIG.), In accordance with a predetermined communication protocol, the management node 1 0, node 3 0 _2-3 0! Controls communication between ₂ and terminating Roh one de 3 0 _e. More specifically, the communication unit 31 transmits an active check bucket from the management node 10. AC, (FIG. 3 (a) refer) and other nodes 3 0 _2-3 0 _12, a function of receiving the buckets preparative response from the terminal node 3 0 _e, § live check bucket preparative AC, the It has a corresponding acknowledgment bucket ACK (see Fig. 3 (b)) and a function to transmit data.

Command determination unit 32 determines the contents of the command C _A contained in § live check packet Ad, and notifies the determination result to the response for the bucket preparative generator 3 3. The response bucket generation unit 33 generates a response bucket ACK, and transmits this to the management node 10 via the communication unit 31. Incidentally, each of the configuration of a node 3 0 _2-3 0, ₂ and the terminal node 3 0 _e is above node 3 0, is the same as the configuration (see FIG. 5).

Next, the operation of the embodiment will be described with reference to FIGS. 6 to 11. Hereinafter describes about the operation when a failure (path error) has occurred between the nodes 3 0 and node 3 0 ₁₂ a working communication paths Pt ^ top shown in FIG. 6 You. FIG. 7 is a flowchart for explaining the operation of the path abnormality detecting unit 15 shown in FIG. 2, and FIG. 8 is a node 30! Shown in FIG. -3 0 ₁₂ and 3 0 is a flowchart for explaining the operation of Noso respectively.

FIG. 9 is a flowchart for explaining the operation of the node configuration information recognizing unit 13 shown in FIG. 2. FIG. 10 is an operation diagram of the failure occurrence point searching unit 16 shown in FIG. It is a flowchart explaining.

In this case, it is assumed that the working communication path PH, shown in FIG. 6, is being used, and that no failure has occurred on the working communication path PH, yet. Oite in this state, the route the abnormality detection section 1 5 of the management node 1 0 that shown in FIG. 2, the flow advances to step SA 1 shown in FIG. 7, with respect to the terminal node 3 0 _beta every predetermined time Then, an alive check is performed to detect a path error (in this case, the path error of the active communication path PHi).

Specifically, the route abnormality detection unit 15 transmits the bucket X _e for alive check to the terminal node 30 _e via the communication unit 11, and then proceeds to step SA 2. S In step SA2, the path abnormality detection unit 15 determines whether or not a response packet ACK, from the terminal node 3 Oe, has been received. Further, the Araibuchiwekku for packets AC _e, the path switcher 2 0 _ra shown in FIG. 6, via the working communication path and the path switcher 2 0 _e, terminal node 3 0 shown in FIG. 5 Received by the communication unit 31 of _e .

Thus, the terminal node 3 0 _e command determination unit 3 2 are stearyl Tsu determination result flop SB 1 shown in FIG. 8, i.e., whether the determination result received the packet AC _e for § live check “Yes” and proceed to step SB2. In step SB2, the command determination unit 32 of the terminal node 3 (the response packet generation unit 33 generates the response packet ACK _e, and then transmits this via the communication unit 31. After transmitting the data to the management node 10, the process proceeds to step SB 3. In step SB 3, the command determination unit 32 determines whether there is data to be transmitted to another node. In this case, the determination result is “No”, the process returns to step SB1, and the above operation is repeated.

The response for packet ACK _e transmitted from the terminal node 3 (the path switching device 20 _e shown in FIG. 6, working communication path PH, and via the path switcher 20 _m, in Figure 2 This is received by the communication unit 11 shown above, whereby the route abnormality detection unit 15 sets the determination result of step SA2 shown in Fig. 7 to "Yes" and returns to step SA1. the abnormality detecting unit 1 5 performs § live check for terminal node 3 0 _e at predetermined time intervals.

Also, in parallel with the operation of the above-described path abnormality detection unit 15, the node configuration information recognition unit 13 operates according to the flowchart shown in FIG. That is, in step SC 1 were shown in the drawing, Roh one de configuration information recognition unit 1-3, to node 3 0-3 0! ₂ shown in FIG. 6 at predetermined time intervals T ₃ ( »fixed time) Perform an alive check.

Specifically, the node configuration information recognition unit 1 3, Araibuchiwekku for packet AC, ~ eight (: _{After 12} nodes 3 0 via the communication unit 1 1, was sent to ~ 3 0 _12, Sutetsu Proceed to SC2. In step SC 2, Roh one de configuration information recognizing unit 1 3, node 3 0, -3 0 ₁₂ response for packet ACK! ~ACK from, determines whether it has received a _2.

In this case, node 3 0 shown in FIG. 6, - 3 node 3 0 of 0 ₁₂ _2, 3 0 ₅ and 3 convenient three nodes that are assumed to be in the stopped state. Therefore, § live check packets AC transmitted from the management node 1 0, of the to Ac _12, § live check bucket preparative AC _2, AC ₅ and AC _8, the node 3 0 _2, 3 05 and 3 0 Not received on ₃ .

-.. The way, other § live check for bucket door _{_{ACi, AC 3, AC 4,}} ACs AC- and AC g ~AC _12, the node _{3 0,, 3 0 3,} 3 0 3 0 s, 3 07 and Received from 303 to ₃₀ ! ₂ .

Thus, node 3 0, 3 0 ₃ 3 0 _4, 3 06, 3 0 ₇ and 3 0 _9-3 0, ₂ of each of the determination result of step SB 1 shown in FIG. 8, "Y es ”and proceed to Step SB2. In step SB2, node 30! , 3 03, 3 0 _4, 3 0 ₆ 3 0 ₇ and 3 03-3 0 ₁₂ each, the response for the packet _{AC K,, ACK 3, ACK} 4, ACK 6, ACKT and ACK ₃ ～ACK ₁₂ Is sent to the management node 10, and then the process proceeds to step SB3.

In step SB3, node 30! , 3 0 ₃ 3 0 _4, 3 0 _s, each 3 07 and 3 0 _3-3 0! _2, it is determined whether there is data to be transmitted to another node, in this case, The determination result is “No”, the process returns to step SB1, and the above operation is repeated.

Then, the above-mentioned response for bucket preparative _{_{ACK!, ACK 3, ACK 4}} , ACK S, the respective AC KT and ACK ₉ ~ACK ₁₂ is received in the communication unit 1 1 of the management node 1 0, node configuration information recognizing The unit 13 sets the determination result of step SC2 shown in FIG. 9 as “Yes” and proceeds to step SC3.

In step SC 3, the Roh one de configuration information recognition unit 1 3, the response for packet ACK,, ACK _3, ACK _4, ACKE, it ACKT and ACK ₉ ~ACK ₁₂ Based on the record of the state flag FS _B (see FIG. 3 (b)), Figure 4 (a) to Bruno one de configuration information table TB shown, the node 3 0! , 3 0 _3, 3 0 _4, 3 0 _s, 3 Omicron tau and 3 0 _3-3 0, ₂ corresponding to each connection order information CN _C, address AD _c, node state flag F Nc and the communication status flag FC After registering _C (see Fig. 4 (a)), return to step SC1 and repeat the above operation.

That is, in this case, as shown in FIG. 6, the nodes 3 0 ^, 3 0 ₃ 3 0 _4, 3 0 3 0 ₇ and 3 0 _3-3 0 _12, with nodes in operation Yes, it was recognized by the node configuration information recognition unit 13 as a configuration node. In this case, the number m of operating nodes (hereinafter referred to as the number m of operating nodes) is “9”. - How, from each node 3 0 _2, 3 0 ₅ and 3 0 ₃ suspended, since the buckets preparative response not sent to the management node 1 0, the node configuration information recognition unit 1 3, in Step SC 2 The determination result is “No”, and the process proceeds to Step SC4. In step SC 4, the Roh one de configuration information recognition unit 1 3, node 3 0 _2, 3 0 ₅ and 3 0 ₈ of the previous node status flags corresponding to each F Nc (FIG. 4 (a) refer) is It is determined whether it is 0 (during operation) or not. If the determination result is “No”, the process returns to step SC1 and repeats the above operation.

- How, if the determination result in step SC 4 is "Y es", Roh one de configuration information sure識部1 3, node 3 0 _2, 3 0 ₅ and 3 0 ₈ communication Fukutai corresponding to each after setting 1 (abnormal) the flag FC _C (see FIG. 4 (a)), the process returns to step SC 1, the above-described operation is repeated.

Here, for some reason, a failure has occurred at the point P in the section between the nodes 30,, and 30] ₂ on the working communication path Ρ ι ι shown in FIG. And. Immediately after this failure occurs, from node 30 to node 30! Assuming that it becomes necessary to transmit data to ₂ , node 3 sets the determination result of step SB3 shown in FIG. 8 to “Yes” and proceeds to step SB4. In Sutetsu flop SB 4, Node 3, after transmitting the data to the node 3 0 _12, the process proceeds to step SB 5. In this case, the data transmitted from the node 3 0, since the failure Ρ point occurs, is not received by the node 3 0 _12. Accordingly, from node 3 0 _12, the response for bucket Bok corresponding to the de Isseki is not transmitted to the node 3 0 Eta.

In step SB 5, the node 3 0 _u is either reply there Luke not from the destination node 3 0, _2, i.e., determines whether it has received a bucket preparative response, in this case, judgments Set the result to "No" and proceed to step SB6. If the determination result of step SB5 is “Yes”, the node 30 returns to step SB1.

In step SB 6, node 3 0 u, after a predetermined time T ₂ wait, the process returns to step SB 1. Thereafter, the node 30 Η retransmits the data at regular time intervals T ₂ until the data transmission to the node 310 ₁₂ ends normally, that is, until the determination result of step SB 5 becomes “Yes”. Keep doing.

Further, immediately after the failure Ρ points shown in Figure 6 occurs, the scan Tetsupu SA 1 shown in FIG. 7, the path abnormality detecting section 1 5 of the management node 1 0, the Araipuchiwekku for packet AC _e After transmitting to the end node 3 via the communication unit 11, the process proceeds to step SA2.

In step SA2, the path abnormality detection unit 15 determines whether or not a response bucket ACK _e has been received from the terminal node 30 _{e in the same} manner as the above-described operation. Their to bucket preparative AC _e for § live check transmitted from the management node 1 0 shown in FIG. 6 is, since the failure occurs in point P, the terminal node 3 (not a received the. Therefore, the terminal node 3 0, from the bucket for response Ozuru pairs § live check packet AC _e ACK, is not transmitted to the management node 1 0.

As a result, the path abnormality detection unit 15 sets the determination result of step SA2 shown in FIG. 7 to “No” and proceeds to step SA3. That is, in this case, the path abnormality of the working communication path P is detected by the path abnormality detection unit 15. In stearyl-up SA 3, the path abnormality detecting section 1 5, working communication paths Roita, corresponding to, FIG. 4 (b) the route information 1 in the path state flag FL _D of the table TB ₂ (abnormal) as shown After setting, go to step SA4. In step SA 4, the route the abnormality detection unit 1 5, a communication path or the normal path (in this case, the standby communication path PH ₂₎ whether is switched to the switching control signal S from the path switching control unit 1 7 Is determined from the output state of In this case, the route abnormality detection unit 15 sets the determination result of step SA4 to "No", and proceeds to step SA5. In stearyl-up SA 5, the path abnormality detecting section 1 5, path abnormality path (in this case, working communication path PH,) normal path (in this case, the standby communication path PH ₂₎ a to switch the path After issuing an instruction to the switching control unit 17, the process returns to step SA1 and repeats the above-described operation.

The path switching control unit 1 7 outputs the switching control signal S corresponding to the instruction path switcher 2 0 "shown in FIG. 6, 2 0! ~ 2 0 ₁₂ and 2 (to. Thus, the communication paths Is automatically switched from the working communication path to the standby communication path PH _2. Therefore, the inter-node communication is performed via the standby communication path PH ₂ thereafter.

- How, when the result of determination in step SA4 is "Ye s", i.e., if the communication path is already switched to the standby communication path PH ₂ from working communication paths PHi, the route the abnormality detection unit 1 5 returns to step SA 1, repeat the operation described above with respect to the standby communication path PH _2. That is, in this case, the communication path is fixed to the standby communication path PH _2.

Further, after the communication path is switched working communication path PH, from the standby communication path PH _2, at step SB 4 shown in FIG. 8, a node 3 0 u node from 3 0 shown in FIG. 6 It is assumed that the data has been retransmitted to ₁₂ . In this case, the data is node 3 0 via path switcher 2 0 standby communication path PH ₂ and the path switcher 2 0 _12! ₂ received.

Thus, the node 3 0 ₁₂ transmits a response for packets corresponding to the data to node 3 0 ,,. The response for bucket DOO, the path switcher 2 0 _12, is received by the node 3 0 through the standby communication path PH ₂ and the path switching unit 2 0 H. At this time, the node 30 ^ returns the determination result of step SB5 to "Yes", and returns to step SB1. That is, in this case, the data from node 3 0 H to node 3 0 ₁₂ While being performed retransmitted several times, since the communication path is switched working communication path PH, from the standby communications path PH _2, the retransmission of the data after several apparently failure recovery It was done.

On the other hand, in parallel with the operation of the above-described route abnormality detection unit 15, the failure location search unit 16 proceeds to step SD1 shown in FIG. At step SD 1, failure occurrence location searching unit 1 6, the path state flag FL _D shown in FIG. 4 (b) it is determined whether or not 1 (abnormal), this determination result is "No" If so, repeat the same decision. In this case, a failure occurs in point P shown in FIG. 6, and Ri by the path abnormality detecting section 1 5, working communication path PH, corresponding path state flag FL _D is set to 1 (abnormal) Therefore, the failure point searching unit 16 sets the determination result of step SD1 to "Yes" and proceeds to step SD2.

In step SD2 and subsequent steps, the fault location search unit 16 searches for a point P where the fault shown in FIG. 6 has occurred, using the binary tree search method. In summary, the fault location search unit 16 operates based on the node configuration information table TBi and the obtained number m of active nodes (in this case, “9”: see FIG. 6). The binary tree shown in Fig. 11 (a) is created for the nodes inside. The node m / 2 shown in the figure corresponds to the middle node among the number m of operating nodes. The m / 4 node corresponds to the middle node among the number of nodes of mZ 2 or less. On the other hand, the node of 3mZ4 corresponds to the middle node among the number of nodes of mZ 2 or more. Hereinafter, similarly, a binary tree structure is illustrated in FIG.

Here, the fault location search unit 16 transmits the lower node hair live check bucket from the upper node shown in the figure and whether or not the corresponding response bucket has been received. Then, the node to which the live check bucket is to be transmitted is determined, thereby searching for a location where a failure has occurred.

Fig. 11 (b) shows a specific example of the binary tree corresponding to the number of active nodes m = 9 shown in Fig. 6 in the same way as the binary tree shown in Fig. 11 (a). FIG. As shown in the figure, the node 30 ₇ is at the top, the node 3 0 - the one branch Previously node 304 located, the node 301 ₀ to the other branch destination node 30 ₇ is located. Similarly, the two branch destination from node 30 _4, the node 30 ₃ and Contact and node 30 ₆ is positioned, in the previous node 30 _3, node 30, is located. In addition, the two branch destination node 30 _10, is located the node 30 ₉ and node 30 Η, node 30!, In the previous, node 30 ₁₂ is located. Returning to FIG. 10, in step SD2, the failure location searching section 16 sets the destination address of the live check bucket. Specifically, search section 1 6 probe failure occurrence location is the address of the node 3 0 ₇ located at the top of the first 1 view (b), § live check bucket shown in FIG. 3 (a) G After setting to the address AD _{A of} AC ₇ , proceed to Step SD 3.

At step SD 3, the failure occurrence location searching unit 1 6, by controlling the path switching control unit 1 7, from normal path communication paths shown in FIG. 6 (in this case, the standby communication path PH ₂₎ After switching to the line fault occurrence path (in this case, the working communication path PH,), proceed to Step SD4.

At step SD 4, the failure occurrence location searching unit 1 6, after transmitting the node 3 07 Hearaipuchi Eck for packet AC _7, the process proceeds to step SD 5. At step SD 5, failure occurrence location searching unit 1 6 determines whether a response has been received for bucket preparative ACK ₇ from node 3 0 _7. In this case, because a failure does not occur with the management node 1 0 and node 3 0 _7, § live check packet AC ₇ is received by the node 3 0 tau. Accordingly, from node 30 _7, the response for bucket preparative ACK ₇ corresponding to packet ACT for § live check is sent to the management node 1 0.

Then, when the response bucket ACK ₇ is received, the failure location searching unit 16 sets the determination result of step SD5 to “Yes” and proceeds to step SD10. In scan Tetsupu SD 1 0, the failure occurrence location searching unit 1 6, sets a destination address of the path Ketchen Bok for next § live check addresses AD _A (FIG. 3 (a) refer) to. In this case, the failure occurrence location searching unit 1 6, due to receiving a response for buckets preparative ACK7, node 3 0 located at one of the branch destination from the node 3 0 ₇ shown in the first FIG. 1 (b) The address of ₁₀ is used for the live check packet AC _t . After setting to the address AD _A, the process proceeds to step SD 1 1.

In step SD11, the fault location searching unit i6 switches the communication path shown in FIG. 6 from the path error occurrence path (in this case, the working communication path) to the normal path (in this case, the standby communication path PH). After switching to ₂ ), proceed to step SD12. In step SD12, the fault location search unit 16 determines whether or not the search has been completed. In this case, the determination result is "No", and the process returns to step SD3.

In step SD4, the fault location searching unit 16 sets the node 30,. After sending the Hearaibuchi Eck for a packet AC _10, the process proceeds to step SD 5. At step SD 5, failure occurrence location searching unit 1 6 determines whether a response has been received for bucket preparative ACK ₁₀ from the node 30 _10. In this case, since no failure has occurred between the management node ₁₀ and the node 3010, the bucket AC, for the active check. Is transmitted to the management node ₁₀ corresponding to the response packet ACK10.

And the response bucket ACK ,. Is received, the fault location searching unit 16 sets the determination result of step SD5 to “Yes” and proceeds to step SD10. In scan Tetsupu SD 1 0, failure occurrence location searching unit 1 6, since it receives a response for the bucket preparative ACK _1Q, you located at one of the branch destination from the node 30 ₁₀ shown in the first FIG. 1 (b) After setting the address of node 30H to the address ADA of the bucket ACH for live check, proceed to step SD11.

In step SD11, the fault location searching unit 16 switches the communication path shown in FIG. 6 from the path error occurrence path (in this case, the working communication path) to the normal path (in this case, the standby communication path PH). After switching to ₂ ), proceed to step SD12. In step SD12, the fault location searching unit 16 determines whether or not the search has been completed. In this case, the judgment result is “No” and the process returns to Step SD3.

At step SD 3, the failure occurrence location searching unit 1 6, similarly to the operation described above, the normal communication paths shown in FIG. 6 passes (in this case, the standby communication path PH ₂₎ from the line abnormality path ( In this case, switch to the working communication path PH!) And proceed to step SD4.

In step S D4, the fault location search unit 16 sets the node 30! Hair Adip Packet for Ad! Ad! And then go to step SD5. In step SD5, the fault location searching unit 16 determines whether or not a response packet ACKH has been received from the node 30u. In this case, since no failure has occurred between the management node 10 and the node 30u, the response bucket AC corresponding to the alive packet ACu is transmitted to the management node 10.

When the response bucket AC is received, the failure location searching unit 16 sets the determination result of step SD5 to "Yes" and proceeds to step SD10. In step SD10, the fault location searching unit 16 has received the response bucket ACKH, so the last node 30u located ahead of the node 30u shown in FIG. ₁₂ Adoresu of, after was set boss to Adoresu AD _a of Araibuchietsuku for bucket door AC _12, the process proceeds to step SD 1 1.

In step SD 11, the fault location searching unit 16 switches the communication path shown in FIG. 6 from the path error occurrence path (in this case, the working communication path PH,) to the normal path (in this case, the standby communication path). After switching to the path PH ₂ ), proceed to Step SD 12. In step SD12, the fault location search unit 16 determines whether or not the search has been completed. In this case, the determination result is "No", and the process returns to step SD3.

At step SD 3, the failure occurrence location searching unit 1 6, similarly to the operation described above, the normal communication paths shown in FIG. 6 passes (in this case, the standby communication path PH ₂₎ from the line abnormality path ( In this case, after switching to the working communication path), proceed to step SD4.

In step SD4, the fault location searching unit 16 sets the node 30! ₂ hair lip After transmitting the packet for AC _ECI2 , proceed to Step SD5. At step SD 5, failure occurrence location searching unit 1 6 determines whether a response has been received for bucket preparative ACK ₁₂ from node 3 0 _2. In this case, since a failure in the point P between the management node 1 0 and node 3 0 ₁₂ is generated, Araibuchiwekku for packet AC ₁₂ is not received by the node 3 0, _2.

Accordingly, from node 3 0 _12, the response for bucket preparative ACK ₁₂ corresponding to § live check packet AC ₁₂ is not transmitted to the management node 1 0. Accordingly, the failure point searching unit 16 sets the determination result of step SD5 to "No" and proceeds to step SD6. In this case, as shown in FIG. 11 (b), since there is no next node to be checked, the fault location searching unit 16 proceeds to step SD7.

In step SD7, the fault location searching unit 16 changes the communication path shown in FIG. 6 from the path error occurrence path (in this case, the working communication path PH,) to the normal path (in this case, the standby communication path PH). After switching to ₂ ), proceed to Step SD8. In step SD8, the fault location searching unit 16 sets the determination result as "Yes" and proceeds to step SD9.

In step SD9, the fault location searching unit 16 identifies a fault (abnormal) location. Specifically, failure occurrence location searching unit 1 6, finally § live Chi nick row ivy node 3 0 _12, disabled in the interval between the nodes 3 0 H in front of the node 3 0 i ₂ After recognizing that the harm occurrence location (in this case, point P) exists, proceed to step SD14.

At step SD 1 4, failure occurrence location searching unit 1 6 registers the information for identifying a fourth view (b) failure occurrence location to the routing information table TB ₂ shown in. In this case, the failure occurrence location searching unit 1 6, the address of the route information and registers the anomaly address A DN _D table TB ₂ Node 3 0 Ryo Ino dress, the abnormal point Adoresu ADF D to node 3 0 ₁₂ After registering, a series of search processing ends. Note that in the binary tree search method, theoretically, step SD 3 to step SD shown in FIG. The loop processing up to 1 2 (step SD 8) is repeated 1 og (m) times. Further, for example, when a failure in a section between the node 30, and node 30 ₃ shown in FIG. 6 occurs, failure occurrence location searching unit 1 6 intends line operates as follows. That is, the failure point searching unit 16 sets the determination result of step SD1 to “Yes” and proceeds to step SD2. At step SD 2, the failure occurrence location searching unit 1 6, similarly to the operation described above, the Adoresu node 30 ₇ located at the top of the first 1 view (b), shown in FIG. 3 (a) After setting the address AD _{A of the} AC ₇ bucket, proceed to step SD 3.

At step SD 3, the failure occurrence location searching unit 1 6, in the same manner as described above, by controlling the path switching control unit 1 7, the communication path shown in FIG. 6 from the standby communication path PH ₂ after switching to the working communication path PH _t, it proceeds non to step SD 4.

At step SD 4, the failure occurrence location searching unit 1 6, after transmitting the node 30 ₇ Hearaibuchi Eck for packet AC _7, the process proceeds to step SD 5. In this case, a received § live check bucket preparative AC ₇ to node 30 ₇ by the above-mentioned disorders fried, from node 30 _7, the response for the bucket preparative ACK ₇ corresponding to § live check packet ACT management Not sent to node 10. Therefore, in step SD5, the fault location searching unit 16 sets the determination result to "No" and proceeds to step SD6.

At step SD 6, the failure occurrence location searching unit 1 6, sets a destination Adoresu bucket Bok for next Araibuchiwekku Adoresu AD _A (FIG. 3 (a) refer) to. In this case, the failure occurrence location searching unit 1 6, for Nakatsu One receives a response for the bucket preparative ACK7, address of the node 30 ₄ located on the other of the branch destination from the node 307 shown in the first FIG. 1 (b) and, after you have set the address of § live check for packet AC ₄ AD, in, proceed to step SD 7.

At step SD 7, the failure occurrence location searching unit 1 6, after switching to the standby communication path PH ₂ communication paths shown in FIG. 6 from the working communication path, step SD Proceed to 8. In step SD8, the fault location search unit 16 determines whether or not the search has been completed. In this case, the determination result is “No”, and the process returns to step SD3. At step SD 3, the failure occurrence location searching unit 1 6, after switching the communication path shown in FIG. 6 from the standby communication path PH ₂ working communication paths Roita, in, the process proceeds to step SD 4.

At step SD 4, the failure occurrence location searching unit 1 6, after transmitting the node 3 0 ₄ Hearaipuchi Eck for buckets preparative AC _4, the process proceeds to step SD 5. In this case, damage by the above-mentioned failure § live check packets AC ₄ such is received by the node 3 0 _4, from node 3 0 _4, packet ACK for the response corresponding to § live check packets AC ₄ ₄ is not sent to management node 10. Therefore, in step SD5, the fault location searching unit 16 sets the determination result to "Noj" and proceeds to step SD6.

At step SD 6, the failure occurrence location searching section 1 6 is set to address the destination address of Araibuchiwekku for bucket Bok following AD _A (FIG. 3 (a) refer). In this case, the failure occurrence location searching unit 1 6, since One failed to receive a response for buckets preparative ACK _4, node 3 located on the other of the branch destination from the node 3 0 ₄ shown in the first FIG. 1 (b) 0 ₃ of the address, after setting the address AD a of § live check for packet AC _3, the process proceeds to step SD 7.

At step SD 7, the failure occurrence location searching unit 1 6, after switching to the standby communication path PH ₂ communication paths shown in FIG. 6 from the working communication path, the process proceeds to step SD 8. In step SD8, the fault location search unit 16 determines whether or not the search has been completed. In this case, the determination result is “No”, and the process returns to step SD3. At step SD 3, the failure occurrence location searching unit 1 6, after switching the communication path shown in FIG. 6 from the standby communication path PH ₂ working communication paths Roita, in, the process proceeds to step SD 4.

At step SD 4, the failure occurrence location searching unit 1 6, after transmitting the node 3 0 ₃ Hearaibuchi Eck for packet AC _3, the process proceeds to step SD 5. In this case, The fried is § live check packet AC ₃ a is received by the node 30 ₃ due to a failure, the nodes from 30 _3, response for bucket preparative ACK ₃ is managed nodes 1 0 corresponding to § live check packet AC ₃ Not sent to Accordingly, in step SD5, the failure point searching unit 16 sets the determination result to "No" and proceeds to step SD6.

At step SD 6, the failure occurrence location searching section 1 6 is set to address the destination address of the next § live check packet AD _A (FIG. 3 (a) refer). In this case, failure occurrence location searching unit 1 6, because the One failed to receive a response for bucket preparative ACK _3, the address of the node 30 located in node 30 ₃ previously shown in the first FIG. 1 (b), § after setting to live check for packet AC, the address AD _a, the process proceeds to step SD 7.

At step SD 7, the failure occurrence location searching unit 1 6, after switching to the standby communication path PH ₂ communication paths shown in FIG. 6 working communication path PH, from, the process proceeds to step SD 8. In step SD8, the fault location searching unit 16 determines whether the search has been completed. In this case, the determination result is “NOJ” and the process returns to step SD3. In step SD3, the fault location searching unit 1 6, after switching the communication path shown in FIG. 6 from the standby communication path PH ₂ working communication paths PH, the proceeds to scan 'Tetsupu SD 4.

In step SD4, the fault location search unit 16 transmits the node 30, the packet AC for hair live check, and then proceeds to step SD5. In this case, since the packet AC for the error check is received by the node 30, the response packet AC K! Corresponding to the packet AC for the error check is transmitted from the node 30 to the management node 10. Is done. Accordingly, in step SD5, the failure point searching unit 16 sets the determination result as “Yes” and proceeds to step SD10.

In this case, as shown in FIG. 11 (b), there is no node to be subjected to the next live check, so the fault location search unit 16 proceeds to step SD11. In step SD 11, the fault location searching unit 16 uses the communication path shown in FIG. After switching to the standby communication path PH ₂ the working communication path PH, from, the process proceeds to step SD 1 2. In step SD12, the fault location searching unit 16 sets the determination result as "Yes" and proceeds to step SD13.

In step SD 13, the fault location searching unit 16 identifies a fault (abnormal) location. Specifically, failure occurrence location searching unit 1 6, finally § live Ji nick nodes 3 0 Been, and a section in the failure between nodes 3 0 ₃ following the node 3 0 〖 After recognizing that the point exists, proceed to step SD14. At step SD 1 4, the failure occurrence location searching unit 1 6, Fig. 4 (b) routing information table TB ₂ of anomaly § dress ADN _D to Roh one de 3 0 shown in registers of § dresses together, after registering the address of node 3 0 ₃ to the abnormal point address ADF _D, and ends the series of search processing.

As described above, according to the embodiment, only the management node 10 unitarily manages the active communication path PH, (abnormality monitoring), and controls switching when the active communication path ΡΗ, an abnormality occurs. since to carry out the other of the plurality of nodes 3 0, -3 0 ₁₂ per cent and termination Roh one de 3 0 _e administrative functions, lowering this the force is not necessary to have a switching control function, et al., the cost And reliability can be improved.

According to one embodiment, the node configuration information recognizing unit 13 recognizes the node configuration at predetermined time intervals based on the response result of the polling. As compared to the case where the node configuration is implemented by the administrator, the work burden of the administrator can be reduced.

According to an embodiment, the path switching control unit 1 7, abnormality after securing the communication path to a normal standby communication path PH _2: since the (FIG. 7 step SA 4 reference) to As, Intercommunication after switching can be performed continuously. Further, according to one embodiment, after an error has occurred in the active communication path PH, the error occurrence point search unit 16 automatically searches for the error occurrence point. Time required for searching compared to when searching for Can be shortened.

Further, according to one embodiment, an extremely efficient search method called a binary tree search method is used to search for an abnormal location on the active communication path PH i, so that the time required for the search is reduced. It can be dramatically reduced.

One embodiment according to the present invention has been described above in detail with reference to the drawings. However, the specific configuration example is not limited to the embodiment, and a specific range within a range not departing from the gist of the present invention. Even a design change or the like is included in the present invention. For example, in the above-described embodiment, a communication path management program for realizing the function of the management node 10 is recorded on the computer-readable recording medium 200 shown in FIG. The communication path management program recorded in the recording medium 200 may be read and executed by the computer 100 shown in FIG.

The computer 100 shown in FIG. 12 includes a CPU 101 executing the communication path management program, an input device 102 such as a keyboard and a mouse, and a ROM (ROM) for storing various data. (Read Only Memory) 103, RAM (Random Access Memory) 104 for storing the operation parameters, etc., a reading device 105 for reading the communication path management program from the recording medium 200, and a display It is composed of output devices 106 such as printers and the like, and a bus BU for connecting various parts of the devices.

The CPU 101 reads the communication path management program recorded on the recording medium 200 via the reading device 105, and executes the communication path management program to execute the communication path management program described above. I do. Note that the recording medium 200 includes a portable recording medium such as an optical disk, a floppy disk, and a hard disk, as well as a transmission medium such as a network that temporarily records and holds data. Is also included.

Further, in the embodiment, an example in which the present invention is applied to the _two active communication paths PH and the standby communication path PH2 has been described. It is needless to say that the present invention can also be applied to. Further, in one embodiment, after performing the process of searching for a failure location described with reference to FIG. 10, the process of switching the path in which the path abnormality has occurred to the normal path (see FIG. 7: Step SA5) May be performed.

Further, in one embodiment, the function of the management node 10 is changed to the node 30! Without specially providing the management node 10. Any one of the nodes 30 to 30 and ₂ and the end node 30 _e may have the function of the management node 10 to reduce the cost.

As described above, according to the communication path management system of the present invention, only the management node centrally manages the first communication path (abnormality monitoring) and controls switching when an abnormality occurs in the first communication path. Since it is performed, it is not necessary to provide a management function and a switching control function to a plurality of other nodes, so that there is an effect that the cost can be reduced and the reliability can be improved.

In addition, since the first communication path is centrally managed (error monitoring) and switching control is performed when an error occurs in the first communication path, multiple other nodes have management functions and switching control functions. Since there is no need to perform this, costs can be reduced and reliability can be improved.

In addition, since the node configuration recognizing means recognizes the node configuration at predetermined time intervals based on the response result of the polling, when the administrator implements the node configuration as in the past, This has the effect of reducing the work load on the manager as compared to the case of

In addition, since the switching control means fixes the communication path to the normal second communication path after the occurrence of an abnormality, there is an effect that the mutual communication after the switching can be continuously performed.

In addition, after an error has occurred in the first communication path, the error occurrence location is automatically searched for by the search means. There is an effect that the time required for the search can be reduced. In addition, using an extremely efficient search method called a binary tree search method, Since the location where the abnormality has occurred is searched, the time required for the search can be significantly reduced. Industrial applicability

As described above, the communication path management system, the communication path management device, and the computer-readable recording medium that records the communication path management program according to the present invention are provided by a LAN (Local Area Network) or the like that interconnects the computers. This is at least useful for redundant communication paths.

Claims

The scope of the claims

1. The first communication path,

A second communication path provided for the first communication path,

A plurality of nodes that are connected to the first communication path to perform mutual communication during normal operation, and execute retransmission a plurality of times when a communication error occurs,

A management node connected to the first communication path during normal operation and performing mutual communication with the plurality of nodes;

With

The management node comprises:

Polling the terminal node of the plurality of nodes at predetermined time intervals, and based on a response result, detecting an abnormality in the first communication path; and detecting an abnormality by the abnormality detecting means. Switching control means for switching a communication path from the first communication path to the second communication,

A communication path management system.

2. The first communication path, the second communication path connected to the first communication path, and the first communication path are connected to the first communication path during normal operation to perform mutual communication, and are retransmitted when a communication error occurs. , A communication management device which is applied to a communication path management system including a plurality of nodes each executing

Communication means connected to the first communication path during normal operation and performing mutual communication with the plurality of nodes;

Abnormality detecting means for polling a terminal node of the plurality of nodes at predetermined time intervals via the communication means, and detecting an abnormality in the first communication path based on a response result; ,

Switching control means for switching a communication path from the first communication path to the second communication path when an abnormality is detected by the abnormality detecting means; A communication path management device comprising:

3. A node configuration recognizing means is provided for polling the plurality of nodes at predetermined time intervals via the communication means, and recognizing a node configuration based on a response result. 3. The communication management device according to claim 2.

4. The switching control means, when switching the communication path from the first communication path to the second communication path, fixes the communication path to the second communication path. 4. The communication path management device according to item 2 or 3.

5. When the abnormality detecting unit detects an abnormality in the first communication path, polls the nodes included in the node configuration via the communication unit, and based on the response result, 4. The communication path management device according to claim 3, further comprising a search unit that searches for a location where an abnormality has occurred on the first communication path.

6. The search means creates a binary tree from the nodes included in the node configuration, and searches for an abnormality occurrence location on the first communication path based on a binary tree search method using the binary tree. The communication path management device according to claim 5, wherein:

7. The first communication path, the second communication path connected to the first communication path, and the first communication path are connected to the first communication path during normal operation to perform mutual communication, and are retransmitted when a communication error occurs. A computer-readable recording medium that stores a communication path management program applied to a communication path management system including a plurality of nodes each of which executes a plurality of times.

An abnormality detection step of causing a terminal node of the plurality of nodes to perform polling at predetermined time intervals, and detecting an abnormality of the first communication path based on a response result; A switching control step of switching a communication path from the first communication path to the second communication node when an abnormality is detected in the abnormality detecting step;

And a computer-readable recording medium storing a communication path management program for causing a computer to execute the program.