JPH08172446A - Method for relaying message between networks - Google Patents

Abstract

PURPOSE: To enable an appropriate processing node to surely receive and process a message when the node is present in the same sub-network. CONSTITUTION: Nodes 111 and 112 for a processing and the node 310 for relay are connected to the sub-network 10, the node 320 for the relay and the nodes 211-214 for the processing are connected to the sub-network 20 similarly and they are respectively logically circularly connected. Also, the nodes 310 and 320 for the relay are connected to the sub-network 30 for the relay. A cyclic counter area is provided in the message in addition to a destination address, an origin address and data. A message emission node adds one to the cyclic counter area in the message every time the message is circulated once in the present sub-network. The node for the relay relays and transfers the message only when the value of the cyclic counter area of the inputted message exceeds a prescribed value.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an inter-network message relay method, and more particularly, to a message distribution between sub-networks in a distributed system consisting of sub-network groups connected with logically annularly connected nodes. Relay transfer method.

[0002]

2. Description of the Related Art A distributed system in which a plurality of sub-networks are each connected to a group of nodes that are logically connected in a ring, and these sub-networks are interconnected via a relay node and a relay network. Has become popular.

Conventionally, in such a distributed system,
When relaying a message between sub-networks,
The relay node holds the relay address information containing the destination address group of the message to be relayed, and the destination address that matches the content of the destination address area in the message input from one of the sub-networks is the relay address information. If it is included in the message, the message is relayed to the other sub-network. This conventional inter-network message relay method is shown in FIGS.
This will be described below.

FIG. 6 shows an example of a distributed system to which a conventional inter-network message relay method is applied. Figure 6
In the processing node 111, the processing node 112,
The relay node 310 is connected to the sub-network 10 and is logically linked in this order in a circular shape. Similarly,
The relay node 320, the processing node 211, the processing node 212, the processing node 213, and the processing node 214 are connected to the subnetwork 20 and are logically connected in this order in a circular shape. Further, the relay node 310 and the relay node 320 are connected to the relay subnetwork 30. The relay nodes 310 and 320 hold relay address information 312 and 322, respectively.

FIG. 7 shows an example of a message format in the conventional inter-network message relay method. The message 700 includes a destination address area 711, a source address area 712, and a data area 720.

FIG. 8 is an operation flow of the relay node, FIG.
Is an operation flow of the processing node. In the following, in the distributed system of FIG. 6, the message relay operation in the relay node 310, using the processing node 111 connected to the sub-network 10 as a message issuing node,
A conventional inter-network message relay method will be described by taking the operations of the processing nodes 111 and 112 as examples. Note that the processing relating to the message whose destination address is its own node in the processing node is omitted in FIG. 9 and the following description.

[1] Message Relay Operation from Subnetwork 10 to Subnetwork 20 in Relay Node 310 A message 700 is input from the subnetwork 10 to the relay node 310. Destination address area 71 of input message 700
It is determined whether or not an address that matches the content of 1 exists in the relay address information 312 that is held, and if there is, it goes, and if it does not, it goes. The entered message 700 is sent to the subnetwork 20.
The message is sent to the relay sub-network 30 so as to be transferred to the relay node 320 connected to, and the processing relating to the input message 700 ends. The input message 700 is sent to the sub-network 10, and the processing relating to the input message 700 is completed.

[2] Operation in Processing Nodes 111 and 112 A message 700 is input from the sub-network 10 to the processing node 112. Similarly, the message 700 is input to the message issuing / processing node 111 through the network. Source address area 7 of the input message 700
It is determined whether the contents of 12 match the own node address. If they match, the process proceeds. If they do not match, the process proceeds. Here, the processing node 111 determines a match, and the processing node 112 determines a mismatch. The message issuing processing node 111 determines whether or not the message 700 needs to be retransmitted to the sub-network 10, and if there is a need, it goes to the other side, and if not, ends the processing related to the message 700. It should be noted that the need for retransmission is, for example, when the message 700 is normally captured by the processing node in the destination address area, and the need for retransmission is, for example, a retransmission request returned from the processing node in the destination address area. This is the case when there is no processing node that should fetch the message (try again). The processing node 312 sends the message 700 as it is to the sub-network 10, and the message 70
The process for 0 is terminated.

[0009]

As described above, in the message relay method in the conventional distributed system composed of the sub-network groups to which the logically annularly connected node groups are connected, the relay node has another sub-node. The relay address information containing the destination address group of the message to be relayed to the network is held, and the destination address that matches the content of the destination address in the message input from the sub-network connected to the relay node is the destination address. If the message is included in the relay address information, the message is uniquely transferred to the relay node connected to another sub-network, which causes the following problem.

If the destination address is a message having a multicast address, the relay node cannot specify the trigger for relaying the message. Here, in the case where the multicast address message sent from the source node is received by only one processing node and should be processed according to the message content, the processing node that can receive the message is It does not exist between the source node and the relay node,
It is assumed that it exists between the relay node and the source node. In this case, if the relay node relays the message to the relay sub-network, it may be received and processed by the processing node connected to another sub-network, and the same sub-network as the source node. The communication delay time is longer and the response time until the processing result is returned to the source node is longer than in the case where the processing node connected to the node receives and processes. Furthermore, when there is no processing node capable of receiving and processing the message in the other sub-network, the processing related to the message is not performed at all. This problem seems to be solvable by sending the relay message to the original sub-network at the time of relaying the message in the relay node, but in this case, the message is It can be received and processed by multiple processing nodes in duplicate, and the processing results from multiple processing nodes are returned to the source node, increasing the amount of communication traffic in the network, the source node and the processing. Useless processing will occur in the protection node.

An object of the present invention is to provide an appropriate processing node in the same sub-network as described above in a distributed system consisting of sub-network groups connected with logically connected nodes. Nevertheless, the response time of the processing result increases due to the message being transferred to the remote processing node and processed, and the message is not transferred to any processing node and is not processed,
It is an object of the present invention to provide an inter-network message relay method that solves an increase in communication traffic volume, unnecessary processing, and the like due to messages being redundantly transferred and processed by a plurality of processing nodes.

[0012]

In order to achieve the above object, the present invention provides a cyclic counter area in a message,
Updates the value of the tour counter area every time the message goes around the connected subnetwork of the source node,
The relay node connected to a certain sub-network refers to the cyclic counter area in the message sent from the forward adjacent node, and the value of the cyclic counter area is set to a predetermined threshold value (cyclic threshold value). ) Is exceeded, the message is sent to a backward adjacent node of the own sub-network, and if it is exceeded, the message is relayed to another sub-network via the relay sub-network. .

[0013]

The functioning threshold may be held in advance by the relay node, or may be provided in the message together with the functioning counter area, and the message issuing / processing node sets an arbitrary threshold value when transmitting the message. It may be done. In any case, the relay node connected to a certain sub-network refers to the cyclic counter area in the message sent from the preceding node in the sub-network, and the value exceeds the cyclic threshold value. Only at this time, the message is transferred to a relay node connected to another predetermined sub-network via the relay sub-network. As a result, even when the destination address is a message having a multi-cast address, the relay node can accurately determine the timing for relaying the message, and if an appropriate processing node exists in the same sub-network, the processing node Allows you to receive and process messages.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a conceptual diagram showing an example of a distributed system to which the inter-network message relay method of the present invention is applied. This is basically the same as the distributed system of FIG. That is, the processing node 111 and the processing node 11
2. The relay node 310 is connected to the sub-network 10 and is logically linked in this order. Similarly, the relay node 320, the processing node 211, the processing node 222, the processing node 223, and the processing node 22.
4 are connected to the sub-network 20 and are logically connected in this order in a ring shape. Further, the relay node 310 and the relay node 320 are the relay sub-network 30.
It is connected to the. The relay nodes 310 and 320 hold the cyclic threshold value information 311 and 321 respectively,
When the cyclic threshold area is provided in the message, the relay nodes 310 and 320 do not need to have the cyclic threshold information.

FIG. 2 shows an example of a message format in the inter-network message relay method of the present invention. (A) shows the case where the relay node holds the cyclic threshold information in advance, and (b) shows the message. This is a case where a cyclic threshold value region is provided. That is, in (a), the message 200 includes a destination address area 211 and a source address area 2
12, the cyclic counter area 213, and the data area 220. Further, in (b), the message 200 includes a destination address area 211, a source address area 212, a cyclic counter area 213, a cyclic threshold area 214, and a data area 22.
Contains 0. Here, in the case of (a), the message information can be reduced by the cyclic threshold area, and in the case of (b), the cyclic threshold can be set to an arbitrary value according to the message at the source node. it can.

FIG. 3 is an operation flow of the relay node in the inter-network message relay method of the present invention, and FIG. 4 is an operation flow of the processing node. In the description below, the processing node 111 connected to the sub-network 10 of FIG. 1 is used as a message issuing node, and the message format is as shown in FIG. 2A, the message relay operation in the relay node 310 of FIG. Processing nodes 111, 1
The inter-network message relay method of the present invention will be described by taking the operation in 12 as an example. In the processing node,
The processing regarding the message whose destination address is addressed to its own node is omitted in FIG. 4 and the following description.

[1] Message Relay Operation from Subnetwork Work 10 to Subnetwork 20 in Relay Node 310 From Subnetwork 10 to Relay Node 310
Then, the message 200 is input. Circuit counter area 21 in the input message 200
It is determined whether the value of 3 is larger than the value of the cyclic threshold value 311 that is held, and if it is larger, the process proceeds. If not, the process proceeds. In order to transfer the input message 200 to the relay node 320, it is sent to the relay sub-network 30,
The process related to the message 200 is ended. Input the message 300 to your own subnetwork 1
0, and the processing relating to the message 200 is completed.

[2] Operation in processing nodes 111 and 112 A message 200 is input from the sub-network 10 to the processing node 112. Further, the message 200 is input to the message issuing processing node 111 by circulating the network in the case of the steps of FIG. Source address area 21 of the input message 200
It is determined whether the contents of 2 match the own node address. If they match, the process proceeds. If they do not match, the process proceeds. here,
The message issuing processing node 111 goes to the processing node 112, and the processing node 112 goes to the processing node 112. The message issuing processing node 111 determines whether or not the input message 200 needs to be retransmitted to the sub-network 10, and if there is, it goes to the other side, and if not, ends the processing related to the message 200.
For example, in the case of step of FIG. 3, the processing node between the relay node 310 and the message issuing processing node 111 is the node that should process the message 200, and the node fetches the message 200. In this case, the processing ends here. The message issuing processing node 111 increments the value of the cyclic counter area 213 in the input message 200 by 1, and goes to step G. The processing node 112 receives the input message 200
Is sent to the sub-network 10 as it is, and the process relating to the message 200 ends. Further, if the message issuing processing node 111 also needs to re-send, the message 200 is sent to the sub-network 10 after the processing of the step.
Send to.

In the above description, the message format is shown in FIG. 2 (a), but when the message format shown in FIG. 2 (b) is used, the relay node performs the steps in FIG.
The cyclic counter area 213 and cyclic threshold area 214 in the message 200 may be compared. The cyclic threshold uses the message format shown in FIG. 2B, and
It is also possible to use a method in which a value held in each relay node is added to the value set in the cyclic threshold area 214 in the message.

In the above description, the message relay sub-network circulation counter area 213 adds 1 every time the message relay sub-network sub-network is cycled, but the message issuing processing node The cyclic counter area is initialized to a predetermined value, and is decremented by 1 each time the sub-network circulates. The relay node makes a zero judgment on the value of the cyclic counter area and relays the message to the relay sub-network. It may be transferred.

Further, each relay node between the sub-networks and the relay sub-network (for example, the relay node 310, the relay node 320 and the relay network 30 in FIG. 1) is degenerated and connected to both sub-networks. Even when the message is directly relayed by one relay node, it can be easily realized as a variation of the present invention.

FIG. 5 shows another example of the message format in the inter-network message relay method of the present invention. The message 500 in FIGS. 5A and 5B
2 corresponds to (a) and (b) of FIG. 2, respectively, and has a reception counter area 515 indicating the cumulative number of nodes that have received the message for the processing and a reception threshold area 516 thereof. Is. In addition, reception threshold 5
16 is set by the source node, and the message 500
Represents the number of nodes desired to be processed. The message format of FIG. 5 is such that the node receiving the input message for its processing increments the reception counter area 515 by 1 and sends it again to the same sub-network as the input, thereby It is particularly effective in a distributed system in which the message can be relayed to another sub-network only when the value of the reception counter area 515 is smaller than the value of the reception threshold area 516.

Next, in the explanation of the operation of FIG. 1, the processing of the relay node 320 of the relay destination subnetwork 20 to which the message is relayed from the relay node 310 of the relay source subnetwork 10 will be touched upon. This processing includes the following examples. This is the same when the message is relayed from the relay node 320 side to the relay node 310 side.

[Example 1] The relay node 320 manages status information such as whether each processing node 211, 212, 213, 214 under the node is in operation, which service can be executed, and the like. , Selects an appropriate processing node based on the content of the relayed message and transfers the message. At that time, in the relay node 320, the contents of the source address area in the relayed message are saved in the save area in the relay node 320 or in a predetermined area in the message, and then the address of the own node is saved. To the source address area in the message. If there is no suitable processing node, information to that effect is added to the message and returned to the relay node 310. Also, when receiving the processing result message from the transfer destination processing node,
Relay node 320 relays it to relay node 310. At that time, in order that the return message reaches the processing request source node, the saved original source address is set in the destination address area in the return message.

[Example 2] The relay node 320 sends out the message relayed in the sub-network 20 in the same manner as the source node in the relay-source sub-network 10. At that time, after the contents of the source address area in the relayed message are saved in the save area in the relay node 320 or a predetermined area in the message, the address of the own node is set as the sender in the message. Set in the address area. When the processing result message is received from the processing node that has processed and executed the message, the relay node 320 relays it to the relay node 310. At that time, in order to make it reach the processing request source node, the saved source address is set in the destination address area in the return message. When a message circulates in the relay node 320 because there is no suitable processing node, the relay node 320 uses a circulation counter or the like like the source node in the relay source sub-network 10. And re-send a predetermined number of times,
Similar to the processing in the relay node 310 described above, when the cyclic counter value exceeds the threshold value, information indicating that fact is added to the message and returned to the relay node 310. In this example, it can be considered that the relay node 320 of the relay destination sub-network 20 has both functions of the source node and the relay node 310 in the relay source sub-network 10.

[0027]

As described above, according to the present invention, the value of the cyclic counter area provided in a message for circulating a certain sub-network to which a group of logically connected nodes is connected. When the destination address is a message with a multicast address, the relay node of the relevant sub-network relays the message to another sub-network only when the value exceeds the specified value (cyclic threshold). However, the relay node can accurately determine the timing for relaying the message, and the message is transferred to the remote processing node and processed even though there is an appropriate processing node in the same subnetwork. The response time of the processing result is increased due to the fact that the message is not transferred to any processing node and is not processed. , It is possible to prevent the message increase duplicate communication traffic due to be processed are transferred and wasteful process to a plurality of processing nodes.

The cyclic threshold value may be stored in the relay node in advance or added as a cyclic threshold value area together with the cyclic counter area in the message. In the former case, the message It is possible to keep the amount of increase of the information of only in the cyclic counter area,
In the latter case, there is an advantage that the message issuing processing node can arbitrarily set the cyclic threshold value according to the type and importance of the message. In addition, a cyclic threshold value is set in the cyclic counter area in the message, the value is decremented by 1 every time the message circulates in the sub-network, and the relay node makes a judgment of zero to trigger the message relay. Then, the cyclic threshold area in the message becomes unnecessary.

[Brief description of drawings]

FIG. 1 is a conceptual diagram showing an example of a distributed system to which an inter-network message relay method of the present invention is applied.

FIG. 2 is a conceptual diagram showing an example of a message format in the inter-network message relay method of the present invention.

FIG. 3 is a diagram showing an example of an operation flow of a relay node in the inter-network message relay method of the present invention.

FIG. 4 is a diagram showing an example of an operation flow of a processing node in the inter-network message relay method of the present invention.

FIG. 5 is a conceptual diagram showing another example of a message format in the inter-network message relay method of the present invention.

FIG. 6 is a conceptual diagram showing an example of a distributed system to which a conventional inter-network message relay method is applied.

FIG. 7 is a conceptual diagram showing an example of a message format in a conventional inter-network message relay method.

FIG. 8 is a conceptual diagram showing an example of an operation flow of a relay node in a conventional inter-network message relay method.

FIG. 9 is a conceptual diagram showing an example of an operation flow of a processing node in a conventional inter-network message relay method.

[Explanation of symbols]

 10, 20 Sub-network 30 Relay sub-network 111, 112 Processing node 211-214 Processing node 310, 320 Relay node 311, 312 Cyclic threshold information 200 Message 211 Destination address area 212 Source address area 213 Cyclic counter Region 214 Travel threshold

Claims (3)

[Claims] 1. A configuration in which a plurality of sub-networks each having a plurality of nodes are connected, a relay node is connected to each sub-network, and the relay nodes are connected by a relay sub-network. Then, the nodes connected to each sub-network are logically connected in a ring, and the message from the front adjacent node is sent to the rear adjacent node after performing predetermined processing as necessary, and the message is also transmitted. In an inter-network message relay method in a distributed system in which a relay node relays to another sub-network via a relay sub-network, a cyclic counter area is provided in the message, and the message is connected to the sub-network of the source node. The value of the patrol counter area is updated each time the The relay node connected to the node refers to the cyclic counter area in the message sent from the forward adjacent node, and the value of the cyclic counter area is set to a predetermined threshold value (hereinafter referred to as the cyclic threshold value). ) Is not exceeded, the message is sent to a backward adjacent node of the own sub-network, and if it is exceeded, the message is relayed to another sub-network via the relay sub-network. Inter-message relay method. 2. The inter-network message relay method according to claim 1, wherein the relay node holds a cyclic threshold value in advance. 3. The method of relaying a message between networks according to claim 1, wherein a cyclic threshold area is provided together with a cyclic counter area in the network, and a message origin node arbitrarily sets the value of the cyclic threshold area. A method for relaying messages between networks that features.