JP2007189560A - One-to-multiple type message transmission system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To suppress an influence when a node defect occurs by transmitting a message to a plurality of receiving nodes without causing response explosion and grasping message receiving situations of the receiving nodes in one-to-multiple type message transmission. <P>SOLUTION: The receiving nodes R are arranged in a matrix, and a transmitting node S transmits a first message to the first receiving node of each column of the matrix, transmits a second message to the first receiving node of each row, receives success response signals from the last receiving nodes of each column and row and discriminates a transmission result and a defective receiving node from receiving situations of the success response signals. Each receiving node transfers the received first message to the next receiving node of the same column, the last receiving node of each column transmits a success response signal to the transmitting node and transfers the received second message to the next receiving node of the same row, and the last receiving node of each row transmits a success response signal to the transmitting node. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a one-to-many message transmission system.

  In a distributed application using a plurality of computers (nodes), it is necessary to transmit and receive messages (application data) between the nodes. Then, it is required to deliver the message reliably (reliable message transmission / reception). HTTPR (HTTP Reliable), ebXML Message Service, WS-Reliable Messaging, WS-Reliability, and the like have been proposed as techniques for realizing highly reliable message transmission / reception between one-to-one nodes on the Internet.

  The basic principle of these techniques is that, as shown in FIG. 12, a positive response in which a receiving node transmits a message response confirmation (referred to as ACK) to the transmitting node, and based on that, the transmitting node confirms message delivery. Confirmation.

  On the other hand, depending on the application, highly reliable message transmission / reception between one-to-many nodes may be desired. In this case, when the above-described technique based on the positive response confirmation is used, as shown in FIG. 13, when the number of receiving nodes increases, the number of ACK transmissions to the transmitting node increases explosively. Will occur. In other words, the increase in ACK causes a problem that the load on the transmitting node increases drastically and becomes unscalable with respect to the number of receiving nodes.

  Technologies that can solve the problem of response explosion include (1) tree structure ACK method, (2) negative response confirmation <NACK) method, and (3) redundant transmission / forward error correction (FEO) method. It has been.

  As shown in FIG. 14, the tree structure ACK scheme reduces the number of response confirmations to the transmission node by arranging the reception nodes in a tree structure, and the upper nodes in each layer aggregate the response confirmations from the lower nodes. It is a method to make it. When each transmitting node cannot receive ACK for a certain period of time, it retransmits the message to the receiving node according to the tree structure (see, for example, Non-Patent Document 1).

  The negative response confirmation <NACK) method is a method for notifying the transmitting node or another receiving node that the receiving node has not received the message (Non-Patent Document 2). This method generally considers that the occurrence frequency of NACK is lower than that of ACK, and can suppress the response explosion stochastically. Each transmitting node retransmits a message to its receiving node only when it receives a NACK.

Redundant transmission / forward error correction (FEO = Forward Error Correction) is a method in which a transmission node receives the received message by transmitting the same message multiple times or by transmitting and receiving the original message and the forward error correction code together. In this method, a message that could not be reconstructed can be reconstructed to eliminate the retransmission of the message (Non-patent Document 3).
TRAM: A Tree-based Reliable Multicast Protocol, Sun Microsystems Laboratory, 1998 http://research.sun.com/technical-reports/1998/abstract-66.html Negative-ACKnowledgment (NACK) -Oriented Reliable Multicast (NORM) Protocol, lETF RFC3940, 2004http = // www.ietf.org/rfc/rfc3940.txt The Use of Forward Error Correction (FEC) in Reliable Multicast, IETF RFC3453,2002http: //www.ietf.org/rfc/rfc3453.txt

In the tree structure ACK scheme, when a failure occurs in a receiving node close to the transmitting node, many receiving nodes are affected. In addition, message transmission / reception is interrupted to a normal node in which no failure has occurred. In addition, the transmitting node can grasp the reception status of the entire node group constituting the leaf of the tree structure from which ACK does not return, but cannot grasp the reception status of all the receiving nodes.

  In addition, in the negative response confirmation (NACK) method, since the positive response confirmation (ACK) is not explicitly transmitted, if the NACK cannot be transmitted / received, it is assumed that the NACK was not transmitted / received normally but was transmitted / received. It will be regarded. Therefore, a mechanism that can reliably transmit and receive NACK is required. In addition, the transmitting node grasps the reception status of the receiving node when NACK does not come for a certain period of time, but NACK cannot reliably confirm that the receiving node has received the message.

  In the redundant transmission / forward error correction (FEC) method, it is not possible to guarantee that a message has been transmitted / received reliably, no matter how redundantly transmitted. Even when it is considered that message transmission / reception is performed “reliably”, it is not efficient because the reception side needs to perform redundant transmission to receive the message. In addition, the transmission node cannot grasp the reception status of the reception node.

  The present invention has been made in view of the above problems. In one-to-many message transmission, [a] a message can be transmitted to a plurality of receiving nodes without causing a response explosion, and [b] message reception by a receiving node. It is an object of the present invention to enable message transmission that can grasp the situation and can suppress the influence even when a node failure occurs.

In order to achieve the above object, a one-to-multiple message transmission system according to a first aspect of the present invention provides:
A one-to-multiple message transmission system comprising a computer device constituting a sending node and a plurality of computer devices constituting a receiving node,
The receiving nodes are logically arranged in a matrix;
The sending node sends a first message to the first receiving node in each column of the matrix, sends a second message to the first receiving node in each row of the matrix, and the last message in each column of the matrix. Receiving a success response signal from the receiving node, receiving a success response signal from the last receiving node of each row of the matrix;
The receiving node transfers the received first message to the next receiving node in the same column, the last receiving node in each column transmits a success response signal to the transmitting node, and the received second message is the same. Forward to the next receiving node in the row, and the last receiving node in each row sends a success response signal to the sending node;
It is characterized by that.

  For example, when a success response signal cannot be received from a receiving node in a certain column and a receiving node in a certain row, the transmitting node determines that the receiving node at the intersection of the row and the column has failed.

  For example, when the sending node cannot receive a success response signal from a receiving node in a certain column, the sending node sends a first message to the first or last receiving node of each row of the matrix, and the last or When a success response signal is received from the first receiving node and a success response signal cannot be received from a receiving node in a row, a second message is transmitted to the first or last receiving node in each column of the matrix, and A success response signal is received from the last or first receiving node in the column.

  For example, when the transmission node cannot receive a success response signal from only one of the reception node of a certain column and the reception node of a certain row and can receive a success response signal from the reception node of another column and row, A means for retransmitting the first or second message, receiving a success response signal, and identifying a receiving node in which a failure has occurred may be provided based on the success response signal.

  For example, the transmission node stores definition information that defines a structure of a logical matrix, and first route information that defines a transmission route of the first message is added to the first message to add the first message. A second route information defining a transmission route of the second message is added to the second message and supplied to the receiving node, and the receiving node receives the first or second received The first or second message is transmitted according to the first or second route information added to the message.

  The transmission node may include means for generating first and second messages by dividing a message to be transmitted or means for generating first and second messages by classifying messages to be transmitted. .

In order to achieve the above object, a computer apparatus according to a second aspect of the present invention provides:
A computer device for transmitting a message to computers arranged in a logical matrix,
Means for transmitting the first message to the first receiving node in each column among the receiving nodes logically arranged in a matrix;
Means for transmitting a second message to the first receiving node of each row of the matrix;
Means for receiving a success response signal from the last receiving node of each column of the matrix;
Means for receiving a success response signal from the last receiving node of each row of the matrix;
Is provided.

In order to achieve the above object, a one-to-multitype message transmission method according to a third aspect of the present invention provides:
Sending the first message to the first receiving node in each column of the receiving nodes logically arranged in a matrix, and sending the second message to the first receiving node in each row of the matrix;
The received first message is sequentially transferred between the receiving nodes constituting each column of the logical matrix, and when the last receiving node in each column successfully receives the message, a success response signal is output;
The received second message is sequentially transferred between the receiving nodes constituting each row of the logical matrix, and a success response signal is output when the last receiving node of each row has successfully received the message.
It is characterized by that.

In order to achieve the above object, a computer program according to the fourth aspect of the present invention provides:
Computer
The first message is transmitted to the first receiving node in each column of the receiving nodes arranged logically in a matrix, and the second message is received at the beginning of each row of the receiving nodes arranged in the matrix. Send to the node and receive a success response signal from the last receiving node in each row and column of the matrix;
It functions as a transmission node.

  With the above configuration, [a] a message can be transmitted to a plurality of receiving nodes without causing a response explosion, [b] the message reception status of the receiving node can be grasped from a success response signal, and [c] when a node failure occurs. It is possible to reduce the influence by resending the message.

  A highly reliable message multicast system according to an embodiment of the present invention will be described below.

  As shown in FIG. 1, the reliable message multicast transmission / reception system according to the embodiment of the present invention includes computer devices 11 connected to each other by a communication network N1.

The computer devices 11 are connected to each other by a communication network N1, and each function as a communication node.
Any one of the computer apparatuses 11 is a transmission node that transmits a message, and any remaining plural computer apparatuses 11 are reception nodes that receive a message from the transmission node.
Each computer apparatus 11 may be a transmission node or a reception node as appropriate according to the message or the content of the work, and the combination and number of transmission nodes and reception nodes are changed and set as appropriate.

Next, a procedure of the entire system for message transmission / reception in the message multicast system configured as described above will be described.
Apart from the physical connection and arrangement of the computer device 11,
One of the plurality of computer apparatuses 11 is a transmission node, and any other plurality of computer apparatuses 11 are reception nodes.
Here, in order to facilitate understanding, the number of receiving nodes is 25.

The transmitting node maintains the logical mesh configuration of the receiving node separately from the physical configuration. Since it is a logical configuration, it does not depend on the physical configuration.
FIG. 2A shows an example of a mesh configuration, in which 25 receiving nodes are arranged in a two-dimensional mesh shape (matrix) of 5 rows × 5 columns.

  This mesh-like grouping on the horizontal axis is called an X group (matrix column), and is individually referred to as groups a to e (a column to e column). The grouping on the vertical axis is referred to as a Y group (matrix row) and is individually referred to as groups A to E (A to E rows).

  As shown in FIG. 2B, the transmission node holds a table defining this mesh configuration.

  The transmitting node sets two messages to be transmitted as one set, and transmits one message to each group of the X group and the other to each group of the Y group. At this time, one message to be transmitted may be divided into two and transmitted.

[Send and receive messages]
When the two messages are MSG-1 and MSG-2, respectively, the transmission node transmits the message MSG-1 to each of the groups a to e (columns) of the X group as illustrated in FIG. As shown in FIG. 3B, the message MSG-2 is transmitted to each of the groups A to E (rows) of the Y group.

  The nodes of each group deliver messages between receiving nodes belonging to the same group in a bucket relay manner. The last receiving node in each group delivers the delivered message (MSG-1 or MSG-2) to the sending node.

  This serves as a positive response confirmation (reception response; ACK) for each group. Instead of using the delivered message as a substitute for ACK, the last receiving node may separately send an ACK message to the transmitting node.

The transmitting node adds information indicating a route for bucket relay based on the node arrangement table TN to the message to be transmitted. For example, in a message based on the SOAP (Simple Object Access Protocol) specification, a node list traced as routing information is added to the SOAP header element.
For example, the transmission node adds the route information shown in FIG. 4A to the message MSG-1 transmitted to the groups a to e.
Further, the transmission node adds the route information shown in FIG. 4B to the message MSG-2 transmitted to the groups A to E.

Each receiving node transmits the received message to the next node according to the route information added to the received message. For example, the receiving node R2 directly receives the message MSG-1 from the transmitting node, and the next receiving node next to itself (R2) from the added route information “R2 → R7 → R12 → R17 → R22 → S”. R7 is specified and the message MSG-1 is transmitted to the node R7. Similarly, the receiving node R2 receives the message MSG-2 from the node R2 node, and receives the next of itself (R2) from the added route information “R1 → R2 → R3 → R4 → R5 → S”. The node R3 is specified, and the message MSG-1 is transmitted to the node R3.
Similarly, for example, when receiving the message MSG-1 and MSG-2, the receiving node R25 specifies the route information added to the header, the node next to itself (R25) being the transmitting node S, Messages MSG-1 and MSG-2 are transmitted to the transmission node S.

[Understanding the reception status of the receiving node by the sending node]
The sending node grasps the reception status of the message at the receiving node.
When the message transmission form as described above is adopted, the reception status of the receiving node cannot be grasped only by one message (MSG-1). For example, as shown in FIG. 5, it is assumed that a failure has occurred in the node R18. In that case, in the transmission of MSG-1, the response confirmation message of group c (message sent from the receiving node R23 to the sending node S) is not sent. Therefore, the transmitting node can grasp that MSG-1 has not (completely) reached each receiving node of group c.

  Similarly, the transmitting node can recognize that MSG-2 has not reached Group D from ACKs related to other messages MSG-2. From this relationship, it can also be understood that the node R18 belonging to the group c and the group D has failed. That is, the transmitting node can determine whether or not each receiving node has received a message only by receiving a message from the last node of each group.

  When a node failure occurs between the transmission of the message MSG-1 and the transmission of the message MSG-2, the reception status can be grasped by retransmitting the unsent message that could not be confirmed. This point will be described later.

  Further, when a failure occurs in a re-transmission / reception node for an unsent message, it is necessary to transmit a message to a node in which no failure has occurred (a node other than R18 in the group c and group D). In the example of the failure of the node R18 shown in FIG. 5, it is necessary to transmit the message MSG-1 to the node R23 where no failure has occurred and the messages to the nodes R19 and R20. Therefore, it is necessary to retransmit message MSG-2 to nodes R19 and R20 and message MSG-1 to node R23. However, the transmission node S grasps the reception status assuming that the message MSG-1 has not been delivered to all the nodes in the group c and the message MSG-2 has not been delivered to all the nodes in the group D.

  Therefore, the sending node S removes the failed node from routing, and sends the paired message (MSG-1, MSG-2) to the group opposite to the group that sent the message first and in the opposite direction. Resend the message from. That is, MSG-1 is transmitted to each group of Y group, and MSG-2 is transmitted to each group of X group in the opposite direction to the previous time. By doing so, MSG-2 is delivered to the nodes R19 and R20, MSG-1 is delivered to the node R23, and a message is delivered to a node having no failure.

  More specifically, additional information “R5 → R4 → R3 → R2 → R1 → S” for tracing the receiving node of the group A in the reverse order is added to the message MSG-1 to the receiving node R5 of the group A. Send. Similarly, additional information “R5 → R4 → R3 → R2 → R1 → S” for tracing the group B receiving nodes in the reverse order is added to the message MSG-2 and transmitted to the group B receiving node R10. The same applies to the groups C to E. Further, additional information “R21 → R16 → R11 → R6 → R1 → S” for tracing the receiving node of group a in the reverse order is added to the message MSG-2 and transmitted to the receiving node R5 of group a. The same applies to the groups b to e. As shown in FIGS. 4C and 4D, the transmission node prepares route information for the occurrence of this abnormality in advance.

  Then, since the last node of each group transmits a message to the transmitting side, the transmitting node can confirm the transmission (arrival) of the messages MSG-1 and MSG-2 to the non-failed node. During this time, the node that has already received MSG-1 or MSG-2 only relays the message according to the routing information.

  With such a message transmission / reception method, retransmission is possible even when a node failure occurs between the transmission of the first message MSG-1 and the transmission of the next message MSG-2. For example, if message MSG-1 is successfully transmitted to all groups, then message MSG-2 is transmitted, and transmission of group D of this message fails, nodes R16, R17, R18, R19, R20 Not all or some of them have been sent. Even in this case, message transmission to a node other than the failed node can be performed by retransmitting the message in the reverse direction even in X group units, and the failed node can be identified from the result of the retransmission. By the above method, highly reliable one-to-many message transmission / reception can be realized.

  As described above, according to the message transmission system of this embodiment, (1) the same message can be reliably transmitted to a plurality of receiving nodes without causing a response explosion, and (2) the message reception status of the receiving node is the receiving node. (3) It is possible to keep sending and receiving messages while minimizing the effect of a node failure. Furthermore, even in one-to-many highly reliable message transmission / reception, the load can be distributed throughout the system, and high performance is not required for the transmission node.

  Next, the configuration and operation of each transmission node and reception node for enabling the transmission processing of the message as the overall transmission / reception will be described.

  As shown in FIG. 6, the computer device 11 physically includes a communication device 21, a control device 22, a storage device 23, an input device 24, and an output device 25.

  The communication device 21 is an interface for communicating with another computer device 11 via the communication network N1. For example, the network interface card (NIC) is used.

  The control device 22 includes, for example, a CPU (Central Processing Unit) and a RAM (Random Access Memory) serving as a work area, and executes each communication control process described later.

The storage device 23 is composed of a built-in hard disk device or the like, and stores the processing program of the control device 22 and various data.
In the storage device 23, matrix information (FIG. 3 (b)) and route information (FIGS. 4 (a) to (d)) of a receiving node when the computer device 11 becomes a transmitting node are registered.

The input device 24 includes a keyboard, a mouse, and the like, and supplies various information to the control device 22.
The output device 25 includes a display device and a printer, displays various information on a screen, or generates a hard copy.

  A communication network N1 shown in FIG. 1 is a network that performs data transmission based on a predetermined communication program such as TCP / IP (Transmission Control Protocol / Internet Protocol), and is configured by, for example, a LAN (Local Area Network) or the like. The

Next, the operation of the computer apparatus 11 functioning as a transmission node will be described with reference to FIG.
First, the control device 22 divides (classifies) a message to be transmitted into a first message MSG-1 and a second message MSG-2 (step S11). For example, when there is only one transmission message, one message is divided to generate a first message MSG-1 and a second message MSG-2. Further, when there are a plurality of transmission messages, they are grouped into a first message MSG-1 and a second message MSG-2.

  Subsequently, the control device 22 reads the normal-time X group route information from the storage device 23, adds it to the header information of the first message MSG-1, and transmits it (step S12). Specifically, the route information for group a is added to the header information of the first message MSG-1 and transmitted to the receiving node R1, and the route information for group b is added to the header information of the first message MSG-1. Add the information and send it to the receiving node R2,. . . The route information for the group e is added to the header information of the first message MSG-1 and transmitted to the receiving node R5. After that, timing for timeout is started.

  Subsequently, the control device 22 reads the normal-time Y group route information from the storage device 23, adds the route information to the header information of the second message MSG-2, and transmits it (step S13). Specifically, route information for group A is added to the header information of the second message MSG-2 and transmitted to the receiving node R1, and the route information for group B is added to the header information of the second message MSG-2. Add the information and send it to the receiving node R6. . . The route information for group E is added to the header information of the second message MSG-2 and transmitted to the receiving node R21. After that, timing for timeout is started.

  Subsequently, the control device 22 waits for reception of an ACK (transmitted first message MSG-1 or second message MSG-2) from each group (steps S14 and S15).

During this time, the transmitted message is sequentially transferred through the receiving node, and is returned to the transmitting node from the last receiving node of each group.
If all ACKs are received within the specified time (step S14; Yes), it is determined that transmission of the message to all receiving nodes has been completed normally (step S16).

  On the other hand, when at least one ACK cannot be received within the specified time (step S14; No, step S15; Yes), first, an abnormal (failed) node is identified (step S17).

  More specifically, as shown in FIG. 8, the control device 22 determines that the ACK that cannot be received corresponds to the first message MSG-1 (from the X group) and the second message MSG-2. It is determined whether or not both of those corresponding to (from the Y group) are (step S21).

  If it is determined that the group that cannot receive ACK is both the X group and the Y group (step S21; Yes), a node located at the intersection of the X group that cannot receive ACK and the Y group that cannot receive ACK fails ( (Abnormal)) (step S22). For example, when ACK cannot be received from the group b of the X group and the group b of the Y group, it is determined that the node R7 located at the intersection is abnormal (failure).

  When it is determined that the group that cannot receive the ACK is one of the X group and the Y group (step S21; No), it is determined that any node in the group is abnormal (failure) (step S23).

  As described above, if an ACK cannot be received from only one of the groups a to e in the X group and the groups A to E in the Y group, a failure occurs after a certain node transfers one message. It is assumed that this is the case. In this case, for example, the process of FIG. 9 may be performed to identify the failed node.

  That is, it is determined whether the group that cannot receive ACK is the group a to e in the X group or the group A to E in the Y group (step S31).

  If the group that cannot receive ACK is any of the groups a to e in the X group, a message is transmitted again to the groups A to E in the Y group (step S32), and the group that cannot receive ACK is Y If it is one of the groups A to E in the group, the message is transmitted again to the groups a to e in the X group (step S33).

  Through this process, the control device 22 receives the transmitted ACK, identifies the group that cannot receive the ACK in the X group and the group that cannot receive the ACK in the Y group, and determines the failure location.

On the other hand, after identifying the abnormal node in step S17 of FIG. 7, the control device 22 performs a process of retransmitting the message for which ACK could not be received (step S18).
Details of this processing will be described with reference to FIG.

  First, the control device 22 determines whether or not an ACK has been received from the groups a to e in the X group (step S41).

  If an ACK has not been received from any of the groups a to e in the X group (step S41; Yes), the process proceeds to step S42. On the other hand, when ACK has been received from all the groups a to e in the X group (step 41; No), step S42 is skipped.

  In step S42, the abnormal-time Y group route information is read from the storage device 23, added to the header of the message MSG-1, and transmitted (step S42).

  Specifically, route information for group A (R5 → R4 → R3 → R2 → R1 → S) is added to the header information of the first message MSG-1 and transmitted to the receiving node R5. The group B route information (R10 → R9 → R8 → R7 → R6 → S) is added to the header information of the message MSG-1 and transmitted to the receiving node R10. The same applies to the groups C to E.

  Next, the control device 22 determines whether or not an ACK has been received from any one of the groups A to E in the Y group (step 43).

If an ACK has not been received from any of the groups A to E in the Y group (step S43; Yes), the process proceeds to step S44. On the other hand, when ACK is received from all the groups A to E in the Y group (step 43; No), step S44 is skipped.
In step S44, the abnormal X group route information is read from the storage device 23, added to the header of the message MSG-2, and transmitted (step S42).

  Specifically, the route information for the group a (R21 → R16 → R11 → R6 → R1 → S) is added to the header information of the second message MSG-2 and transmitted to the receiving node R21. Further, route information for the group b (R22 → R17 → R12 → R7 → R2 → S) is added to the header information of the second message MSG-2 and transmitted to the receiving node R22. The same applies to the groups c to e.

  In this way, a message can be reliably transmitted to nodes other than the failed node.

On the other hand, when receiving the message, the receiving node starts the transfer process shown in FIG. First, the received message is stored in the storage device 23 (step S51).
Next, the next node to which this message is to be transferred is determined (step S52). This is obtained by analyzing header information of the received message.
Next, this message is transferred to the determined transfer destination node (step S53).

  As described above, the message transmission system according to the present embodiment is a highly reliable message multicast system, and (A) logically arranges receiving nodes in a mesh shape, and a part thereof (a sufficiently smaller number than the total number). Response explosion can be prevented by having only the receiver side confirm the positive response. Furthermore, reliable message transmission / reception can be performed by resending the message to the receiving node that has not received the ACK for a certain period of time. In addition, (B) Sending node grasps the receiving node that could not receive the message by making two sets of two messages and sending the two messages to the mesh-like receiving node using different paths. it can. Furthermore, (C) when a transmission node detects a node failure, it retransmits the message on the transmission path opposite to the path that was transmitted first, so that even when a node failure occurs, Can deliver messages as much as possible.

  In addition, this invention is not limited to the said embodiment, A various deformation | transformation and application are possible.

  For example, in the above embodiment, the number of receiving nodes is 25 and the matrix is 5 × 5, but the shape of the matrix is arbitrary. For example, route information may be set assuming a 6 × 4 matrix and a 1 × 1 matrix. If the route information can be defined so that the route information passes through all the nodes in the X direction and the Y direction, the number of nodes and the shape of the matrix are arbitrary.

  In addition, it is possible to use the matrix of the above embodiment in combination with other conventional methods. For example, a message may be transmitted to the transmitting node in the matrix of FIG. 2 using another conventional technique, and the transmitting node may deliver the message to the receiving node.

  In the above embodiment, the transmission node generates route information that defines the delivery order of the messages and transmits the route information by adding it to the message. The node may determine the transfer destination of the message based on appropriate identification information in the message and transfer the message.

  Further, in the above embodiment, when an abnormality is desired, a message is transmitted in the reverse direction and in the reverse direction as shown in FIG. 6, but when an abnormality occurs, the message is shown in FIG. As described above, although the route is different from the normal route (the row and the column are reversed), the message may be transmitted in the forward direction in the row and the column.

1 is a block diagram of a message transmission system according to an embodiment of the present invention. (A) is a block diagram of a matrix logically constituted by the transmission node S and the reception node R, and (b) is an example of data defining a logical matrix constituted by the reception node. It is a figure for demonstrating the transmission path | route of the message in the matrix-like node of FIG. It is a figure which shows the example of the route information stored in a transmission node, (a) and (b) are for normal time, (c) and (d) are for abnormal (failure), (a) and (c) Is for the X group, and (b) and (d) are for the Y group. It is a figure for demonstrating operation | movement when a failure generate | occur | produces in receiving node R18 which comprises a matrix. It is a block diagram which shows the structure of the computer apparatus shown in FIG. It is a flowchart for demonstrating operation | movement of a transmission node. It is a flowchart which shows the detail of the failure node discrimination | determination process in the flowchart of FIG. FIG. 9 is a flowchart showing details of processing that is appropriately executed in addition to the processing of FIG. 8 in order to determine a failed node. It is a flowchart which shows the detail of the resending process of the message in the flowchart of FIG. It is a flowchart for demonstrating the process of a receiving node. It is a figure for demonstrating the positive response confirmation between 1 to 1 nodes. It is a figure for demonstrating the positive response confirmation between one-to-many nodes. It is a figure for demonstrating a tree structure ACK system.

Explanation of symbols

11 Computer Device 21 Communication Device 22 Control Device 23 Storage Device 24 Input Device 25 Output Device N1 Communication Network

Claims (9)

A one-to-multiple message transmission system comprising a computer device constituting a sending node and a plurality of computer devices constituting a receiving node,
The receiving nodes are logically arranged in a matrix;
The sending node sends a first message to the first receiving node in each column of the matrix, sends a second message to the first receiving node in each row of the matrix, and the last message in each column of the matrix. Receiving a success response signal from the receiving node, receiving a success response signal from the last receiving node of each row of the matrix;
The receiving node transfers the received first message to the next receiving node in the same column, the last receiving node in each column transmits a success response signal to the transmitting node, and the received second message is the same. Forward to the next receiving node in the row, and the last receiving node in each row sends a success response signal to the sending node;
A one-to-many message transmission system characterized by the above. The sending node is
When a success response signal cannot be received from a receiving node in a certain column and a receiving node in a certain row, it is determined that the receiving node at the intersection of the row and the column is faulty.
The one-to-many type message transmission system according to claim 1, wherein: The sending node is
When a success response signal cannot be received from a receiving node in a column, a first message is transmitted to the first or last receiving node of each row of the matrix, and a successful response is received from the last or first receiving node of each row of the matrix. Receive the signal,
When a successful response signal cannot be received from a receiving node in a row, a second message is transmitted to the first or last receiving node in each column of the matrix, and a successful response signal is received from the last or first receiving node in each column. Receive,
The one-to-many type message transmission system according to claim 1 or 2, characterized by the above. The sending node is
When a successful response signal cannot be received from only one of the receiving node of a certain column and a receiving node of a certain row, and a successful response signal can be received from the receiving node of another column and row,
Means for retransmitting the first or second message, receiving a success response signal, and identifying a receiving node in which a failure has occurred by the success response signal;
The one-to-many type message transmission system according to any one of claims 1 to 3, wherein The transmission node stores definition information that defines a structure of a logical matrix, and the reception node adds first route information that defines a transmission route of the first message to the first message. A second route information defining a transmission route of the second message is added to the second message and supplied to the receiving node;
The receiving node transmits the first or second message according to the first or second route information added to the received first or second message.
The one-to-many type message transmission system according to any one of claims 1 to 4, wherein The transmission node includes means for dividing a message to be transmitted to generate first and second messages, or means for classifying the message to be transmitted and generating first and second messages.
The one-to-multiple message transmission system according to any one of claims 1 to 5, wherein A computer device for transmitting a message to computers arranged in a logical matrix,
Means for transmitting the first message to the first receiving node in each column among the receiving nodes logically arranged in a matrix;
Means for transmitting a second message to the first receiving node in each row of the receiving nodes arranged in the matrix;
Means for receiving a success response signal from the last receiving node of each column of the matrix;
Means for receiving a success response signal from the last receiving node of each row of the matrix;
A computer device comprising: The first message is transmitted to the first receiving node in each column of the receiving nodes logically arranged in a matrix, and the second message is transmitted to each row of the receiving nodes arranged in the matrix. Send to the first receiving node,
The received first message is sequentially transferred between the receiving nodes constituting each column of the logical matrix, and when the last receiving node in each column successfully receives the message, a success response signal is output;
The received second message is sequentially transferred between the receiving nodes constituting each row of the logical matrix, and a success response signal is output when the last receiving node of each row has successfully received the message.
A one-to-multitype message transmission method characterized by the above. Computer
The first message is transmitted to the first receiving node in each column of the receiving nodes arranged logically in a matrix, and the second message is received at the beginning of each row of the receiving nodes arranged in the matrix. Send to the node and receive a success response signal from the last receiving node in each row and column of the matrix;
A computer program that functions as a transmission node.

Images