JPH08307437A - Faulty node elimination method and device for multi-branch bus type system - Google Patents

Abstract

PURPOSE: To provide a faulty node elimination method and device for a multi- branch bus type system which quickly specifies a faulty node and automatically restores a system. CONSTITUTION: When a fault occurs in a multi-branch bus type system, a fault decision processing part 10 of a master node 1 isolates the isolation control signal IS to collectively isolate all slave nodes N1 to Nn from a multi-branch bus 2. Then a self-return monitoring part 11 sends the return signal addressed to the part 11 itself to the bus 2 to confirm its own normalcy. When the normalcy of the node 1 is confirmed, the part 10 connects successively every slave node to the bus 2 by an individual isolation signal DISC to decide a fault. Finally, the isolation of the signal IS is canceled and all slave nodes are collectively connected to the bus 2 except for the faulty one. In such a procedure, a system can be quickly and automatically restored against its fault.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a multi-branch bus type system composed of a master node and a plurality of slave nodes, and more particularly to a method and apparatus for identifying and removing a faulty node.

[0002]

2. Description of the Related Art In a bus type system in which a plurality of nodes are connected to a bus line, it is necessary to quickly identify a faulty node and disconnect it from the system in order to prevent a communication failure due to a node fault. is necessary.

Conventionally, in a multi-branch bus type system composed of a master node and a plurality of slave nodes, if the master node or the slave node fails due to some cause and communication becomes impossible, the failing node is manually disconnected from the bus. The system was being restored.

[0004]

However, the method of searching for a faulty node and manually disconnecting it from the bus not only requires maintenance personnel for that purpose, but communication is interrupted for a long time, and system reliability is reduced. There was a problem of significantly lowering it.

It is an object of the present invention to provide a method and apparatus for removing a faulty node in a multi-branch bus type system which quickly identifies a faulty node and automatically restores the system.

[0006]

According to the method of removing a faulty node according to the present invention, when it is recognized that a fault has occurred in a multi-branch bus system, a plurality of slave nodes are collectively disconnected from the multi-branch bus, and subsequently a plurality of slave nodes are disconnected. Slave nodes are sequentially connected to the multi-branch bus one by one, and the failure judgment of one slave node connected to the multi-branch bus is performed. It is characterized in that they are collectively connected to.

Further, after the plurality of slave nodes are collectively disconnected from the multi-branch bus, the master node sends a return signal addressed to itself to the multi-branch bus to determine normality, and the master node is normal. If it is determined that the slave nodes are faulty, the failure determination is sequentially performed.

[0008]

[Function] When a system failure occurs, the master node collectively disconnects all the slave nodes from the multi-branch bus, connects the individual slave nodes in sequence, and performs failure determination. And connect. By this procedure, the multi-branch bus type system at the time of failure can be quickly restored. Further, the master node determines whether or not its communication function is broken by sending back the return data addressed to itself to the multi-branch bus, and only when it is normal, the failure judgment of the slave node is executed. Therefore, the failure of the entire multi-branch bus type system including the master node can be automatically determined and recovered.

[0009]

Embodiments of the present invention will now be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of a fault node removal system according to the present invention. In this embodiment,
One master node 1 and n slave nodes N _{1 to} N _n
And are connected by a bus 2 in a multi-branch form. The master node 1 is provided with a failure determination processing unit 10 in addition to communication means and system management means (not shown), and performs failure node identification and system restoration processing when a failure occurs in the system. Furthermore, the master node 1 is provided with a self-loopback monitoring unit 11 for checking whether or not a failure has occurred in its own communication means.

The same bus disconnection / connection function is provided in each of the slave nodes N1 to Nn. Describing the slave node N1 as a representative, the relay 101 is interposed between the bus 2 and the transmission / reception unit 102, and according to a control signal output from the bus disconnection control unit 103, disconnects the slave node N1 from the bus 2 or Performs disconnection / connection operation such as connecting with 2. Other slave node N2
-Nn have the same structure.

The isolation control signal IS is commonly input from the failure determination processing unit 10 of the master node 1 to the bus disconnection control units of the slave nodes N1 to Nn, and individual disconnection control signals are supplied to the respective slave nodes N1 to Nn. Separation signal DISC1
~ DISCn are individually input. Isolate control signal IS and individual disconnection signals DISC1 to DISCn
Is output to each slave node by a line separate from the bus 2.

The isolation control signal IS is a control signal for simultaneously performing disconnection setting or disconnection release for all slave nodes N1 to Nn. On the contrary,
The individual disconnection signals DISC1 to DISCn are control signals for individually disconnecting or connecting to each slave node. It should be noted that how the bus disconnection control unit of each slave node operates the relay depending on the isolation control signal IS and each individual disconnection signal DISC will be described with reference to FIG.
Shown in.

Failure Judgment Processing Operation FIG. 2 is a flowchart showing the operation of the failure judgment processing unit 10 of the master node 1 in this embodiment. The master node 1 normally transmits a control signal to the slave node (S301) and receives a response thereto from the slave node (Yes in S302). If no such response signal is received over bus 2, or if no response signal is received from multiple slave nodes,
The failure determination processing unit 10 of the master node 1 recognizes that the bus 2 cannot be used due to a failure in the system via the multi-branch bus 2 (No in S302).

When such a failure occurs, the failure determination processing section 10 sends an isolation control signal IS for instructing the "isolation setting" to all the slave nodes (S
303), and all individual disconnection signals DISC1 to DISC
SCn is set to "disconnect" (S304). As a result, all slave nodes N1 to Nn are disconnected from the bus 2 as described later.

When the disconnection of all slave nodes is confirmed (S305), the failure determination processing unit 10 identifies and removes the failed node in the following procedure (S306,
S307).

Faulty Node Judgment Removal FIG. 3 is a flowchart showing a faulty node judgment removal subroutine in this embodiment.

First, the self-loopback monitoring unit 11 sends out data addressed to itself (addressed to the master node) to the bus 2 (S40).
1) It is confirmed that the self-addressed data can be normally received (S402). At this time, if the self-addressed data cannot be received within the predetermined time (No in S402), the failure determination processing unit 10 determines that the master node 1 has a failure and ends the failed node determination processing routine.

When the master node 1 is confirmed to be normal by receiving the self-addressed data within a predetermined time (Yes in S402), the failure determination processing unit 10 initializes the variable M (S403), and individually. Slave nodes are sequentially "connected" to the bus 2 by the disconnection signal DISC, and failure determination is performed for each slave node (S404 to S409). More specifically, when M = 1, the individual disconnection signal DISC1 is “connected”.
To the relay node 1 of the slave node N1 (S404).
01 is turned on to connect the transmitting / receiving unit 102 to the bus 2.

Subsequently, the failure determination processing unit 10 sends the data addressed to the slave node N1 to the bus 2 (S405),
It is checked whether the response from the slave node N1 is within a predetermined time (S406). If there is a response from the slave node N1 within this predetermined time (Yes in S406)
s), the slave node N1 is determined to be operating normally, and the individual disconnection signal DISC1 is set to "disconnect" again (S407). On the contrary, if there is no response from the slave node N1 within this predetermined time (No in S406), the individual disconnection signal DISC1 remains set to "connect".

Next, the variable M is the number n of slave nodes.
Is checked (S408), and if not reached, the variable M is incremented (S40).
9), the same processes S404 to S408 are repeated. Therefore, while sequentially incrementing the variable M, the determination removal processing is executed for all the slave nodes N1 to Nn, and the individual disconnection signal DISC of the normal slave node is set to "disconnect" again (S407), and the failure occurs. The individual disconnection signal DISC of the node is left set to "connect" and exits from the faulty node determination removal subroutine (Yes in S408).
s).

In this way, each individual separation signal DI
If SC1 to DISCn are set to "disconnect" if the nodes are normal nodes, and "isolated" is set to "disconnect" in the state where the failed nodes are set to "connect" (S in FIG. 2).
307), the normal node is reconnected to bus 2 and the faulty node is disconnected from bus 2 and excluded from the multi-branch bus system. In this way, the failed node is automatically eliminated, and the multi-branch bus type system is restored.

Bus disconnection control of slave node In the bus disconnection control section of each slave node, ON / OFF control of the relay is executed according to the setting states of the isolation control signal IS and the individual disconnection signal DISC.

FIG. 4 is a flow chart showing the bus disconnection control of the slave node in this embodiment. First, the bus disconnection control unit 103 of the slave node checks the setting state of the isolation control signal IS received from the failure determination processing unit 10 of the master node 1 (S501), and if it is the disconnection setting, The setting state of the individual disconnection signal DISC is checked (S502). Individual disconnection signal DIS
If C is also "disconnected", the relay is turned off to disconnect the slave node from the bus 2 (S503). This disconnection operation is performed in the processing step S30 of the failure determination processing unit 10.
3 and S304 (see FIG. 2) execute the all slave node disconnection command.

When the individual disconnection signal DISC is set to "connect" in step S502, the bus disconnection control unit 103 turns on the relay to connect the slave node to the bus 2. This connection operation is executed by the processing step S404 (see FIG. 3) of the failure determination processing unit 10.

In step S501, when the isolation control signal IS is set to "disconnection release", if the individual disconnection signal DISC is "connection", the relay is turned off to set the slave node to the bus 2 Is separated from (S503). This disconnection operation is executed when it is determined that the slave node has a failure (S30).
7). On the other hand, if the individual disconnection signal DISC is "disconnect", the relay is turned on and connected to the bus 2 (S504). This connection operation is executed when it is determined that the slave node is normal (S407, S307).

FIG. 5 is a time chart showing a specific operation example of this embodiment. However, for simplicity of explanation, it is assumed that one master node and two slave nodes N1 and N2 are connected to the bus 2, and the slave node N1 is normal and the slave node N2 is faulty. To do.

When the failure judgment processing unit 10 of the master node 1 detects that the bus 2 cannot be used (No in S302), the isolation control signal IS is set to "isolation setting" and all the individual isolation signals DISC1 and DISC2 are set. Each is set to "separation" (S303, S304). Subsequently, when the self-loopback monitoring unit 11 of the master node 1 confirms normality by the self-loopback data (S402), the individual disconnection signal DIS
C1 is set to “connect” and only the slave node N1 is connected to the bus 2 to transmit / receive data (S405, S4).
06). If there is a response from the slave node N1 within the predetermined time, the slave node DN1 is determined to be normal, and the individual disconnection signal DISC1 is set to "disconnect" again (S407).

Subsequently, similarly, the individual disconnection signal DISC2 is set to "connection", only the slave node N2 is connected to the bus 2, and data is transmitted / received. However, if a failure occurs in the slave node N2, the master node 1
The failure determination processing unit 10 of cannot receive a response from the slave node N2 within a predetermined time (N of S406).
o). Therefore, the individual disconnection signal DISC2 to the slave node N2 remains set to "connect", and the faulty node determination removal subroutine is ended.

When the isolation control signal IS is set to "disconnect" in the state of setting the individual disconnection signal (S307), the normal slave node N1 receives the individual disconnection signal DISC1 as "disconnect". Since it is in the state, it is connected to the bus 2 (S504), and since the individual disconnection signal DISC2 is in the "connection" state, the failed slave node N2 remains disconnected from the bus 2 (S503). In this way, only the failed slave node is automatically removed, and operation is resumed as a normal system.

In the present embodiment, the failure determination processing is executed from the slave node N1, but it goes without saying that the failure determination processing may be started from the slave node Nn farthest from the master node 1.

[0032]

As described above, in the fault node removing method and apparatus according to the present invention, when a fault occurs in the system, the master node disconnects all slave nodes from the multi-branch bus and sequentially connects the individual slave nodes. After making a failure determination by connecting the slave nodes together, the slave nodes are collectively connected. By this procedure, the multi-branch bus system at the time of failure can be quickly and automatically restored.

Further, the master node determines whether or not its communication function has failed by sending back the return data addressed to itself to the multi-branch bus, and executes the failure determination of the slave node only when it is normal. To do. Therefore, the failure of the entire multi-branch bus type system including the master node can be automatically determined and recovered.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of a faulty node determination removal system according to the present invention.

FIG. 2 is a flowchart showing a schematic operation of a failure determination processing unit 10 in this embodiment.

FIG. 3 is a flowchart showing a fault node determination removal operation of a fault determination processing unit 10 in this embodiment.

FIG. 4 is a flowchart showing bus disconnection control of a slave node in this embodiment.

FIG. 5 is a time chart showing a specific operation example of the present embodiment.

[Explanation of symbols]

 1 master node 2 bus 10 failure determination processing unit 11 self-loopback monitoring unit 101, 201 relay 102, 202 transmission / reception unit 103, 203 bus disconnection control unit DISC1 to DISCn individual disconnection signal IS isolated control signal N1 to Nn slave node

Claims (6)

[Claims] 1. A method of removing a fault node in a system in which a master node and a plurality of slave nodes are connected by a multi-branch bus, wherein when a fault in the system is recognized, the plurality of slave nodes are connected to the multi-branch bus. , The slave nodes are sequentially connected to the multi-branch bus one by one, the failure determination is performed for one slave node connected to the multi-branch bus, and the slave node in which the failure occurs A method for removing a faulty node is characterized in that the plurality of slave nodes are collectively connected to the multi-branch bus except the above. 2. The master node transmits a predetermined control signal to the plurality of slave nodes through the multi-branch bus and receives the response signal to the control signal within a predetermined time, depending on whether the master node receives the response signal or not. 2. The method for removing a faulty node according to claim 1, wherein the occurrence of a fault in the system is recognized. 3. A fault node removal method in a system in which a master node and a plurality of slave nodes are connected by a multi-branch bus, wherein when a fault occurrence in the system is recognized, the slave nodes are set to the multi-branch bus. From the device, and determines whether the master node is normal. When the master node is normal, the slave nodes are sequentially connected to the multi-branch bus one by one, A failure determination is performed for one slave node connected to the multi-branch bus, and the plurality of slave nodes are collectively connected to the multi-branch bus except for the slave node in which the failure has occurred. How to remove failed node. 4. The master node transmits a predetermined control signal to the plurality of slave nodes through the multi-branch bus, and determines whether the master node receives a response signal to the control signal within a predetermined time. 4. The method for removing a faulty node according to claim 3, wherein the occurrence of a fault in the system is recognized. 5. The master node is normal according to whether the master node sends a signal addressed to itself to the multi-branch bus and receives the signal addressed to itself within a predetermined time. The method of removing a faulty node according to claim 3 or 4, characterized in that it is determined whether or not 6. A failure node removing device in a system in which a master node and a plurality of slave nodes are connected by a multi-branch bus, wherein the master node and the plurality of slave nodes are connected separately from the multi-branch bus. A control signal line is provided, each of the plurality of slave nodes has a switch means for connecting and disconnecting the slave node and the multi-branch bus, and a control signal received from the master node through the control signal line. Disconnection control means for controlling connection and disconnection operations of the switch means according to a signal, and disconnecting the plurality of slave nodes collectively when the occurrence of a failure in the system is recognized, and disconnecting the plurality of slave nodes. Connect each slave node to the multi-branch bus one by one A control signal is sent to the disconnection control means through the control signal line, a failure determination is performed for each slave node connected to the multi-branch bus, and the plurality of slaves are excluded except for the slave node in which the failure has occurred. A fault node removal device comprising: fault determination processing means for transmitting a control signal for collectively connecting nodes to the multi-branch bus to the disconnection control means through the control signal line.