JP2008141361A - Data distribution method, data distribution system and node device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To achieve multicast communication so that even when a fault occurs in a relay node, data can be surely distributed to a distribution destination node without leakage and duplication of data. <P>SOLUTION: A data distribution method is characterized in that: when a distribution source node and a relay mode detect a fault of a relay node one stage lower, a relay node two stages lower is set as a relay node one stage lower to change a configuration of a tree structure; the relay node which has detects the fault acquires a sequence number of the last reception data of a relay node which has become a lower node by the configuration change, and data of the sequence number and following data which exists in data storage queue is set as transmission data to the lower node, and requests distribution of data of the pertinent sequence number to the lower node when data node not present in the data storage queue is present, out of data from the sequence number of the last reception data in the lower node to the sequence number of the last reception data of the node which has detected the fault. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a data distribution method, a data distribution system, and a node device, and in particular, data that can distribute a large amount of data to a large number of servers with high reliability without data leakage or duplication. The present invention relates to a distribution method, a data distribution system, and a node device.

  In general, one-to-many communication that transfers the same data to a plurality of destinations is called multicast communication. Examples of applications that are data transferred by such communication include software distribution and video streaming. In addition to these, a form in which data such as documents and files is distributed to a large number of servers in an in-company network is conceivable. In this case, data is continuously generated unlike software distribution, but data loss is not allowed unlike video data distribution.

  In multicast communication, the network load to which the distribution source node belongs increases as the number of destinations increases to thousands or tens of thousands. Therefore, the load on the distribution source node can be increased by installing a relay node and making the network a tree structure. For example, Patent Document 1 and the like have proposed a method for reducing the above.

  On the other hand, as a data distribution method capable of reliably transferring data without causing data leakage or duplication, when transmitting data from the transmission node to the reception node, the data to be transmitted is temporarily stored in the storage device of the transmission node. A method called a store-and-forward method in which data is transferred after being stored is known. In this method, the transmission node erases the temporarily stored data when the reception response is returned, and retransmits the data when no response is returned. In addition, the receiving node stores the reception state indicating which data has been received, thereby discarding the received data transmitted twice to prevent duplication of data. The store-and-forward method can perform reliable data distribution without duplication even when a failure occurs in a server or a network by performing the above-described operation. In addition, this method can guarantee the order of data distribution by using first-in first-out storage of temporary data in the storage device.

Between the nodes in the tree-structured network described above, specifically, between the distribution source node and the relay node, between the relay node and the next-stage relay node, and between the relay node and the distribution destination node By using a store-and-forward method for communication, reliable data transfer to a large number of distribution destinations can be realized.
JP 2005-25622 A

  In the conventional technology that distributes data by applying the store-and-forward method to the tree-structured network described above, when a failure occurs in a relay node, the data is distributed to the lower nodes of the failed node. It has the problem of disappearing. In order to solve such a problem, there is a method in which data can be distributed to all nodes by re-forming the tree structure of the network when a node failure occurs. If the data stays in the relay node where the error occurs, the data will be lost.

  An object of the present invention is to solve the above-mentioned problems of the prior art and realize multicast communication, and even when a failure occurs in a relay node, data leakage and duplication are eliminated and data is reliably transmitted to a delivery destination node. An object of the present invention is to provide a data distribution method, a data distribution system, and a node device that can be distributed and can reduce the amount of data retransmission traffic.

  According to the present invention, the object is to provide a data distribution method in a data distribution system that distributes data from a distribution source node to a plurality of distribution destination nodes via a relay node, the data distribution system including the distribution source node and a relay A node and a plurality of distribution destination nodes are configured to be connected to each other so as to form a tree-structured network in which the relay node is provided with one or more stages, and the distribution source node and the relay node are: Information on the lower node of the lower stage and the lower node of the lower stage is held in the storage means, and when a failure of the lower relay node is detected, the lower relay node is relayed one lower stage. This is accomplished by changing the configuration of the tree structure as a node and updating the information of the lower node one step lower and the lower node two steps lower in the storage means.

  The object is configured such that a distribution source node, a relay node, and a plurality of distribution destination nodes are connected to each other so as to form a tree-structured network in which the relay node is provided in one or more stages. , A relay node in a data distribution system that distributes data from the distribution source node to a plurality of distribution destination nodes via a relay node, wherein the relay node includes a lower node that is one level lower and a lower node that is two levels lower Holding means for holding information in the storage means, means for detecting a failure of the relay node one level below, and detecting a failure of the relay node one level below, lowers the relay node that is two levels below by one level A means for changing the structure of the tree structure as a relay node, a means for receiving data from an upper node, a data accumulation queue for accumulating and storing received data, and the data accumulation queue at each lower node. Recording means for recording in the storage means which data has been transmitted, means for obtaining the serial number of the last received data of the relay node that has become a lower node due to the configuration change, and the last existing in the data accumulation queue This is achieved by including means for making data after the serial number of the received data transmission data to the lower node.

  According to the present invention, even when a failure occurs in a relay node, data leakage and duplication can be eliminated and data can be reliably delivered to a delivery destination node.

  Hereinafter, embodiments of a data distribution method, a data distribution system, and a node device (hereinafter simply referred to as a node) according to the present invention will be described in detail with reference to the drawings.

  FIG. 1 is a block diagram illustrating a configuration example of a tree-structured distribution network constituting a distribution system according to an embodiment of the present invention.

  The distribution system according to the embodiment of the present invention includes a distribution source node 100, a first-stage relay node 101, a second-stage relay node 111 and 112, a third-stage relay node 121 to 124, and a final destination distribution. It is composed of destination nodes 131-138. The distribution source node 100 distributes data to the first-stage relay node 101 which is a lower node. The relay node 101 distributes the same data to the second-stage relay nodes 111 and 112 which are lower nodes. The relay node 111 distributes the data received from the first-stage relay node 101 that is the upper node to the third-stage relay nodes 121 and 122 that are the lower-level nodes. Similarly, the relay node 112 distributes the data received from the first-stage relay node 101 that is the upper node to the third-stage relay nodes 123 and 124 that are the lower-level nodes. The relay nodes 121 to 124 in the third stage distribute the received data to the distribution destination nodes 131 to 138, which are lower nodes, respectively.

  The example shown in FIG. 1 shows the number of lower nodes connected to the lower level of the relay node as two, but the number of lower nodes connected to the lower level of the relay node in each stage is actually several tens or more. Is assumed. In the example shown in FIG. 1, a tree-structured network is configured by connecting three-stage relay nodes between a distribution source node and a distribution destination node. However, the number of relay nodes is one. There may be three or more stages, and the number of stages may be arbitrary.

  The network constituting the distribution system shown in FIG. 1 is a network that can transmit distribution data and the like between nodes at normal time, and the actual network is at least in order to cope with a failure such as a relay node. A communication path that can be connected from a certain node to a node that is two stages lower is provided.

  As shown in FIG. 1, the data distribution system according to the embodiment of the present invention has a tree structure of the network in which the relay nodes are arranged, so that data is distributed directly from the distribution source node to the distribution destination node. The load on the distribution source node can be greatly reduced. For example, when there are 1000 distribution destination servers, ten first-stage relay nodes are arranged below the distribution source server, and ten second-stage relays are provided as subordinate nodes of the first-stage relay node. By providing a tree-structured network in which nodes are arranged and 10 distribution destination nodes are arranged below each of the relay nodes in the second stage, the distribution source node can only distribute data to 10 nodes. Thus, data can be distributed to 1000 distribution destination nodes, and the communication load on the distribution source server can be extremely reduced.

  FIG. 23 is a block diagram showing the configuration of the relay node. The illustrated relay node 20 includes, as hardware, a memory 21 that is a storage unit and a CPU 22. Note that the configuration of the relay node 20 illustrated in FIG. 23 is described as the configuration of the relay node, but the distribution source node may also have the same configuration.

  The memory 21 of the relay node 20 includes a table group 23 including a data accumulation queue 210, a distribution management table 220, a redistribution management table 240, and a tree structure management table 270, and data receiving means 2301 for processing normal data reception. Redistribution data receiving means 2302 for processing reception of redistribution data, data transmission means 2303 for processing normal data transmission, redistribution data transmission means 2304 for processing transmission of redistribution data, and data to be redistributed to lower nodes Redistribution requesting means 2305 for requesting the distribution of the node to the upper node, redistribution request receiving means 2306 for processing when the redistribution is requested from the lower node, tree structure reconfiguration for instructing the tree structure reconfiguration when the relay node fails Configuration instruction unit 2307, tree structure reconstruction response unit 2308 responding to a tree structure reconstruction instruction, Trough a functional unit group by the relay node connection means 2309 to be connected is stored as a relay node.

  Each means constituting the above-described functional unit group is configured as a program, and the program constituting each means is executed by the CPU 21 so that each function required by each means is constructed.

  FIG. 2 is a diagram for explaining a detailed configuration in the table group 23. Next, this will be described. As the table group 23, a data accumulation queue 210, a distribution management table 220, a redistribution management table 240, and a tree structure management table 270 are provided.

  In the data accumulation queue 210, distribution data a211, distribution data b212,..., Distribution data N213 received from the upper node are stored until distribution to all lower nodes is completed.

  The distribution management table 230 includes a reception sequence number management table 221 and a transmission management table 231. In the reception sequence number management table 221, a final reception sequence number 222 that is the sequence number of the distribution data last received from the upper node is set. The transmission management table 231 includes a transmission management entry a232, a transmission management entry b233,..., A transmission management entry N234 for managing the distribution data stored in the data accumulation queue 210 for each lower node.

  The redistribution management table 240 includes a redistribution sequence number management table 241 and a retransmission management table 251. In the redistribution sequence number management table 241, a redistribution request entry pointer 242 is set, and a redistribution request entry a243, a redistribution request entry b244, and a redistribution request entry c245 are provided for each unit for requesting redistribution of data to an upper node. It is chained. The retransmission management table 251 includes a retransmission management entry a252, a retransmission management entry b253,..., And a retransmission management entry N254 for managing the redistributed data stored in the data accumulation queue 210 for each lower node. From each retransmission management entry, a retransmission request entry a261, a retransmission request entry b262, and a retransmission request entry c263 for managing a retransmission request notified from a lower node are chained.

  The tree structure management table 270 includes an upper node table 271 that manages upper nodes, a lower node table 272 that manages lower nodes, and a lower node connection table 273 that manages combinations of lower nodes and their lower nodes. .

  FIG. 3 is a diagram for explaining the configuration of each table constituting the tree structure management table 270. The configuration of each table shown in FIG. 3 is shown as the contents of the tree structure management table 270 included in the relay node 101 in the system shown in FIG. 1, but the contents of each table differ for each relay node.

  As shown in FIG. 3A, the upper node table 271 includes items of an upper node name 301 and an upper node address 302. In the illustrated example, the items of the upper node name 301 include a node 101 and an upper node address 302. IP_101 is set in the item.

  As shown in FIG. 3B, the lower node table 272 is configured by providing records having items of lower node names 310 and lower node addresses 320 as many as the number of connected lower nodes. In the illustrated example, a record of the node 121 and the IP_121 and a record of the node 122 and the IP_122 are set as a set of the lower node name 310 and the lower node address 320.

  The lower node connection table 273 is configured by including a plurality of records each including a lower node name 330, a two-stage lower node name 340 which is a lower node name, and a two-stage lower node address 350 which is an address thereof. Is done. In the illustrated example, as a set of a lower node name 330, a two-stage lower node name 340, and a two-stage lower node address 350, node 121, node 131, IP_131 record, node 121, node 132, IP_132 record, node 122, node Four records are set: 133, IP_133 record, node 122, node 134, and IP_134 record.

  FIG. 4 is a diagram for explaining the configuration of distribution data and the configuration of the distribution management table 230. Next, this will be described.

  The distribution data includes a node pointer unit 410, a serial number 411 that is a consecutive number uniquely assigned to the distribution data by the distribution source, and a data entity 412 that represents the substance of the distribution data.

  In the transmission management entry a232 of the transmission management table 231, a transmission management record aa421 and a transmission management record ab422 for managing the transmission state of each distribution data stored in the data accumulation queue 210 are chained in the order of the serial numbers 411 of each distribution data. Yes. Similarly, in the transmission management entry b233, the transmission management record ba423 and the transmission management record bb424 are chained in the order of the serial numbers 411 of each distribution data, and in the transmission management entry N234, the transmission management record Na425 and the transmission management record Nb426 are distributed in the distribution data. Are chained in the order of serial numbers 411.

  Each of the transmission management records chained to each transmission management entry in the transmission management table 231 includes a lower-level node name 451 of a transmission destination, a state 452 indicating a transmission state, and a distribution indicating the position in the data accumulation queue 210 of distribution data. It consists of a data pointer 455. The serial number of the transmission data can be confirmed by referring to the serial number 411 of the distribution data indicated by the distribution data pointer 455. The state 452 includes three states of “not yet” 461 indicating unsent, “done” 462 indicating completion of transmission, and “re” 463 indicating waiting for retransmission. One state is set. Each transmission management record is pointed to the corresponding transmission management record aa 421 from the node pointer portion 410 of the distribution data, and transmission management records for managing the same serial number are chained in order, such as transmission management record ba423 and transmission management record Na425. Has been.

  Each of the transmission management entries has a head pointer 431 that points to the head of the transmission management record chain, a final pointer 432 that points to the last, and a state 452 of the transmission management record that is “unknown”, and is the highest (younger) corresponding serial number. It consists of a transmission head pointer 433 indicating a transmission management record and a lower node name 434 indicating a transmission destination node.

  FIG. 5 is a diagram for explaining the configuration of the retransmission management table 251 in the redistribution management table 240.

  In the transmission management entry a252 of the retransmission management table 251, a retransmission management record aa521 and a retransmission management record ab522 for managing the transmission state of each distribution data stored in the data accumulation queue are chained in the order of the serial numbers 411 of the respective distribution data. . Similarly, in the retransmission management entry b253, the retransmission management record ba523 and the retransmission management record bb524 are chained in the order of the serial numbers 411 of the respective distribution data, and the retransmission management record N525 and the retransmission management record Nb26) are chained in the transmission management entry N254. ing.

  Each retransmission management record includes a destination lower node name 551, a state 552 indicating a retransmission state, and a distribution data pointer 553 indicating a position of distribution data in the data accumulation queue 210. By referring to the distribution data sequence number 411 indicated by the distribution data pointer 553, the sequence number of the retransmission data can be confirmed. The state 552 includes two states, “completed” 561 indicating completion of transmission and “re” 562 indicating waiting for retransmission, and one state is set according to the state of retransmission data. In addition, retransmission management records ab 522 corresponding to the distribution data node pointer section 410 are pointed to, and in order, retransmission management records bb 524 and transmission management records Nb 526, retransmission management records for managing the same serial number are chained.

  Each of the retransmission management entries includes a top pointer 531 pointing to the beginning of the retransmission management record chain, a final pointer 532 pointing to the end, and the retransmission (highest) (young) retransmission of the corresponding serial number when the retransmission management record state 552 is in the “re” state. It consists of a retransmission start pointer 533 indicating a management record, a lower node name 534 indicating a retransmission destination node, and a retransmission request unit pointer 535 indicating a retransmission request entry.

  Each of the retransmission request entries includes a retransmission request upper sequence number 541 which is the highest (younger) serial number in the range in which the retransmission request is made, and a retransmission request lower sequence number 542 which is the lowest sequence number in the range in which the retransmission request is issued. The

  FIG. 6 is a diagram for explaining the configuration of the redistribution request entry in the redistribution management table 241 in the redistribution management table 240. Each of the redistribution request entries includes a redistribution request upper sequence number 601 that is the highest (youngest) sequence number of the redistribution request range notified to the upper node, and a redistribution request lower sequence number 602 that is the lowest sequence number. .

  FIG. 7 is a diagram for explaining a message format exchanged between nodes. Next, this will be described.

  As shown in FIG. 7A, the connection message a 710 is a message for notifying that an upper node is newly added as a lower node, and includes a node name 711 of the own node and a node address 712 of the own node.

  As shown in FIG. 7B, the connection message b 720 is a message for notifying the upper node of what kind of node is connected to the lower level of the own node. The node name 721 of the own node, the lower level of the lower node It is composed of a node name 722 and a lower node address 723 of the lower node.

  As shown in FIG. 7C, the distribution telegram 730 is a telegram of data distributed from the distribution node. The distribution telegram 730 includes a serial number 731 assigned in ascending order at the distribution source node and distribution data 733 that is the substance of the data. Composed.

  As shown in FIG. 7D, the upper node change message 740 is a message for instructing the lower node to change the upper node, and includes a node name 741 of the upper node and an upper node address 742 of the upper node. The

  As shown in FIG. 7E, the upper node change response message 750 is a response message to the upper node change message 740, and includes the node name 751 of the own node and the last reception number 222 of the reception number management table 221 at the own node. The final reception serial number 752 is set.

  As shown in FIG. 7 (f), the redistribution request message 760 is notified of the data sequence number for receiving and redistributing the upper node change response message 750 or the redistribution request message 760 from the lower node. Redistribution request that is a message for requesting redistribution to an upper node when there is no distribution data of the corresponding sequence number, and that is the node name 761 of the own node and the highest (youngest) serial number that requests redistribution It consists of an upper sequence number 762 and a redistribution request lower sequence number 763 which is the lowest sequence number.

  FIG. 8 is a flowchart for explaining the processing operation at a related node when a new relay node is connected in a tree-structured network. Next, this will be described. In the example described here, the node 111 is newly connected as a lower node of the relay node 101, the node 121 is newly connected as a lower node of the relay node 111, and the node 131 as a distribution destination is newly connected to the lower node of the node 121. It is an example when connected as a node.

(1) When the node 111 starts connection to the network as a relay node, the node 111 sets information of the node 101, which is the upper node, in the upper node table 271 and sets -1 in the final reception sequence number 222 of the reception sequence number management table 221. Then, a connection message a 710 is created and the node 101 is notified of the connection of the own node. The reason why the final reception serial number 222 is set to -1 is to indicate that the local node distribution data is not held at all (steps 810 to 813).

(2) Upon receiving the connection message a 710 from the node 111, the node 101 adds information on the newly connected node 111 to the lower node table 272 (step 814).

(3) On the other hand, when the node 121 connected as a lower node of the new relay node 111 starts connection to the network as a relay node, it creates a connection message a 710 and notifies the node 111 of the connection of its own node (step) 820, 821).

(4) Upon receiving the connection message a 710 from the node 121, the node 111 adds the information of the node 121 included in the connection message a 710 to the lower node table 272, adds a transmission management entry to the transmission management table 231, and retransmits A retransmission management entry is added to the management table 251 and a connection message b720 is transmitted to the node 101 (steps 822 to 825).

(5) When the node 101 receives the connection message b720 from the relay node 111, the node 101 adds a record of a set of the nodes 111 and 121 to the lower node connection table 273 (step 826).

(6) In addition, when the node 131 connected as the lower delivery destination node 131 of the node 121 starts connection to the network as the delivery destination node, the node 131 creates a connection message a 710 and notifies the node 121 of the connection of the own node. (Steps 830 and 831).

(7) Upon receiving the connection message a 710 from the node 131, the node 121 adds the information of the node 131 included in the connection message a 710 to the lower node table 272, creates the connection message b, and transmits it to the node 111 ( Steps 832 and 833).

(8) Upon receiving the connection message b 720 from the relay node 121, the node 111 adds a record of a set of the nodes 121 and 131 to the lower node connection table 273 (step 834).

  FIG. 9 is a flowchart for explaining the operation of the reception process at the relay node when the distribution data is received from the upper node. Next, this will be described.

(1) When reception data is detected, reception processing is started. First, the serial number 731 of the received distribution message 730 is equal to the value of the final reception serial number 222 + 1 of the reception serial number management table 221 or the final reception serial number It is determined whether or not the value of -1 is -1 (steps 900 and 901).

(2) If it is determined in step 901 that the serial number 731 of the received distribution message 730 is equal to the value of the final reception serial number 222 + 1, or the value of the final reception serial number is -1, the received data is stored in the data accumulation queue. The data is stored in 210 (step 902).

(3) Thereafter, update processing of the transmission management table at the time of data reception (details of this processing will be described later with reference to FIG. 10), the serial number 731 of the received data is set in the final reception serial number 222, and the upper node A normal response indicating that the data has been normally received is returned, and the processing here ends (steps 903 to 905, 910).

(4) If the result of determination in step 901 is that the serial number 731 of the received distribution message 730 is not equal to the value of the final reception serial number 222 + 1 and the final reception serial number is not -1, redistribution After performing a data reception process (details of this process will be described later with reference to FIG. 19), it is determined whether or not the serial number 731 of the distribution telegram 730 is equal to or smaller than the value of the final reception serial number 222 (steps 906 and 907).

(5) If it is determined in step 907 that the serial number 731 of the distribution telegram 730 is less than or equal to the value of the final reception serial number 222, a normal response that the data has been normally received is returned to the upper node, and in step 907 If the serial number 731 of the distribution telegram 730 is not less than or equal to the value of the final reception serial number 222, an abnormal response is returned to the upper node, and the processing here ends (steps 905, 908, 910).

  FIG. 10 is a flowchart for explaining the processing operation for updating the transmission management table at the time of data reception in step 903 of the flow shown in FIG. 9, which will be described next.

(1) When this process is started, first, the transmission management entry at the head of the transmission management table 231 is updated, and a new transmission management record is added to the end of the transmission management record chained from the transmission management entry. “Not yet” 461 is set in the state 452 of the added transmission management record (steps 1000 to 1003).

(2) Next, it is determined whether or not the value of the transmission start pointer 433 of the target transmission management entry is NULL (no distribution data exists in the data accumulation queue). The pointer 433 is set to point to the added transmission management record (steps 1004 and 1005).

(3) After the processing in step 1005, or in the determination in step 1004, if the value of the transmission start pointer 433 of the target transmission management entry is not NULL, the transmission management record pointed to is the transmission management table. If the transmission management record being pointed to is the final entry of the transmission management table 231, the processing here ends (steps 1006 and 1010).

(4) If it is determined in step 1006 that the pointed transmission management record is the last entry in the transmission management table 231, the next entry of the transmission management record chained from the transmission management table 231 is pointed, Returning to the processing from 1002, a new transmission management record is added to the end of the transmission management records chained from the transmission management entry, and the processing is continued (step 1007).

  FIG. 11 is a flowchart for explaining the operation of the transmission process in the relay node when the distribution data is transmitted to the lower node. Next, this will be described.

(1) After starting the process, first, a data redistribution process is executed. Details of this data redistribution processing will be described later with reference to the flow shown in FIG. 21 (steps 1100 and 1101).

(2) After executing the data redistribution process, it is determined whether or not the value of the transmission head pointer 433 of the transmission management entry in the transmission management table 231 is NULL, and the value of the transmission head pointer 433 is NULL. In this case, since the data to be transmitted does not exist in the data accumulation queue 210, after sleeping for a predetermined time, the process returns to the data redistribution process in step 1101 and the process is repeated (steps 1102 and 1114).

(3) If it is determined in step 1102 that the value of the transmission head pointer 433 is not NULL, the distribution data corresponding to the transmission management record pointed to by the transmission head pointer 433 is transmitted from the data accumulation queue 210 to the lower node, It is determined whether or not a normal response has been received from the lower node (steps 1103 and 1104).

(4) If a normal response is received from the lower node in the determination in step 1104, it is determined whether or not there is a transmission management record next to the transmission management record. If there is, a transmission head pointer of the transmission management entry is determined. If the next transmission management record is set in 433 and does not exist, NULL is set 1107 in the transmission head pointer 433 (steps 1105, 1106, 1107).

(5) After the processing of steps 1106 and 1107, the state 452 of the transmission management record is set to “done” 462, and the state 452 of all the transmission management records corresponding to the distribution data is “done” 462. If the status 452 of all transmission management records is “done” 462, all transmission management records corresponding to the distribution data are released, and the distribution data is deleted from the data accumulation queue 210 (step 1108-1111).

(6) After the processing at step 1111 or when there is a transmission management record state 452 that is not “done” 462 in the determination at step 1109, the processing returns to the data retransmission processing from step 1101. Repeat the process.

(7) When the normal response is not received from the lower node in the determination in step 1104, it is determined whether or not the abnormal response is received from the lower node, and the abnormal response is not received from the lower node. Since the node cannot be confirmed, it is necessary to retransmit, and the process returns to the data retransmission process from step 1101 and is repeated (step 1112).

(8) If it is determined in step 1112 that an abnormal response has been received from the lower node, it is determined that data loss has occurred and the process ends abnormally. This is because recovery by operation is necessary (step 1113).

  FIG. 12 is a diagram for explaining an example of reconfiguring the tree structure of a network when a relay node fails. The example shown here is an example when the relay node 121 in the system shown in FIG. 1 becomes a failure 1201. When a failure occurs in the relay node 121, the relay nodes 131 and 132 (described below) In the embodiment of the present invention, data is not distributed to the distribution destination node). For this reason, the relay node 111 higher than the relay node 121 can continue data distribution by newly performing the tree structure reconstruction 1211 and 1212 having the nodes 131 and 132 as lower nodes.

  FIG. 13 is a flowchart for explaining a processing operation at a related node in the case of reconfiguring a tree structure of a tree structure network. Next, this will be described. The example described here is processing when a failure 1301 occurs in the node 121 and the node 111 detects the failure.

(1) When the node 111 detects the occurrence of the failure 1301 in the node 121, the node 111 deletes the information of the failure node 121 from the lower node table 272, which is one of the tree structure management tables 270, and then from the lower node connection table 273. The node corresponding to the two-stage lower node name 340 of the record in which the lower node name 330 is the failure node is selected as a node for instructing the configuration change (steps 1302 to 1304).
(2) In the example described, since the information of the nodes 131 and 132 can be obtained as the two-stage lower node name 340 of the record that is the failure node, the node 111 creates the upper node change message 740, This message is transmitted to these nodes (step 1305).

(3) The nodes 131 and 132 that have received the upper node change message 740 set the information of the node 111 in the upper node table 271 in the own node, create an upper node change response message 750, and transmit it to the node 111. (Steps 1306 and 1307).

(4) Upon receiving the upper node change response message 750 from the nodes 131 and 132, the node 111 deletes the record in which the two-stage lower node name 340 matches the node name 751 of the response message 750 from the lower node connection table 273. After the received node information is added to the lower node table 272, a connection message b720 is created, and this connection message b720 is transmitted to the upper node 101 (steps 1308 to 1310).

(5) On the other hand, since the nodes 131 and 132 transmit the upper node change response message 750 to the node 111 and then create the connection message b 720 and transmit the message to the node, the node 111 The connection telegram b720 from 132 is received, the received node set is added to the lower node connection table 273, and a transmission management entry is added to the transmission management table 231 (steps 1311 to 1313).

(6) Then, the node 111 creates a transmission management record, sets the same value as the failure occurrence node (node 121) in the state 452 of each record, and adds 1315 a retransmission management entry to the retransmission management table 251. Then, a retransmission management record is created, and “done” 561 is set in the state 552 of each record (steps 1314 to 1316).

(7) Further, the node 111 creates a retransmission request entry, sets the final reception sequence number 752 of the upper node change response message 750 in the retransmission request upper sequence number 541, sets 0 in the retransmission request lower sequence number 542, The retransmission request entry is pointed 1318 from the corresponding entry in the retransmission management table 251 (steps 1317 and 1318).

(8) After that, the node 111 determines whether or not the upper node change response message 750 has been received from all of the nodes 131 and 132, and if all the response messages 750 have not been received yet, Waiting for reception of the node change response message 750, the processing from step 1308 is repeated (step 1319).

(9) If it is determined in step 1319 that all of the response message 750 has been received, after executing the accumulated data deletion process (details of this process will be described later with reference to FIG. 14), the redistribution request process (this Details of the processing will be described later with reference to FIG. 15 (steps 1320 and 1321).

  FIG. 14 is a flowchart for explaining the operation of the accumulated data deletion process in step 1320 of the flow shown in FIG. 13. Next, this will be described.

(1) When the accumulated data deletion process is started, the node 111 points to the transmission management entry indicating the failure node (here, the node 121) in the transmission management table 231 and points to the first transmission management record. (Steps 1400 to 1402).

(2) It is determined whether or not there is a transmission management record. If the transmission management record does not exist, the processing here ends. If there is, the transmission management record is deleted (steps 1403, 1410, and 1405).

(3) Next, it is determined whether or not the state 452 of all transmission management records corresponding to the serial number 411 of the corresponding distribution data is “Done” 462. All the transmission management records corresponding to the data are deleted, and the distribution data is deleted from the data accumulation queue 210 (steps 1406 to 1408).

(4) After the processing in step 1408 or in the determination in step 1406, if there is a transmission management record whose status 452 is not “done” 462, point to the next transmission management record, and from step 1403 Returning to the process, the process is repeated (step 1409).

  FIG. 15 is a flowchart for explaining the operation of the redistribution request process for notifying the upper node of the redistribution request in step 1321 of the flow shown in FIG. 13, and this will be described next.

(1) When the redistribution request processing is started, the node 111 points to the highest entry of the retransmission management entry corresponding to the node whose configuration has been changed in the retransmission management table 251, and the retransmission request upper sequence number 541 of the retransmission request entry. However, it is determined whether or not the value is smaller than the serial number 411 of the distribution data corresponding to the transmission management record indicated by the transmission head pointer 433 of the transmission management entry (steps 1500 to 1502).

(2) If it is determined in step 1502 that the retransmission request higher sequence number 541 of the retransmission request entry is smaller than the delivery data sequence number 411 corresponding to the transmission management record, the retransmission request entry is transmitted to the retransmission request lower sequence number 542. A serial number 411 + 1 of the distribution data corresponding to the transmission management record pointed to by the head pointer 433 is set (step 1503).

(3) Next, when there is a retransmission request higher sequence number 541 or less in the transmission management record and the state 452 is “not” 461, the state 452 is set to “re” 463, and a retransmission request is made in the retransmission management record. If there is a record having a higher sequence number 541 or lower and a status 552 of “completed” 561, the status 552 is set to “re” 562 (steps 1504 and 1505).

(4) If it is determined in step 1502 that the retransmission request upper sequence number 541 of the retransmission request entry is larger than the distribution data sequence number 411 corresponding to the transmission management record, the corresponding distribution data is stored in the data accumulation queue of the own node. Since it exists in 210, the retransmission request entry is deleted (step 1530).

(5) After the processing in step 1505 or after the processing in step 1530, it is determined whether there is still a retransmission management entry for the node whose configuration has been changed. If there is still a retransmission management entry, the configuration of the retransmission management table 251 The entry next to the resend management entry of the changed node is pointed, and the process returns to step 1502 to repeat the process (steps 1506 and 1507).

(6) If it is determined in step 1506 that there is no retransmission management entry for the node whose configuration has been changed, a redistribution request entry is created, and the highest number among the retransmission request higher sequence numbers 541 is assigned to the redistribution sequence number management. After setting the redistribution request upper sequence number 601 of the redistribution request entry of the redistribution request entry pointer 242 of the table 241 and setting the lowest number among the retransmission request lower sequence numbers 542 to the redistribution request lower sequence number 602, The re-distribution request message 760 is transmitted to the upper node, and the processing here ends (steps 1508 to 1511 and 1520).

  FIG. 16 is a flowchart for explaining the operation of receiving a redistribution request by a node notified of the redistribution request from the lower node. Next, this will be described.

(1) When processing is started, a node that has been notified of a redistribution request from a lower node creates a retransmission request entry, and a redistribution request for a redistribution request message 760 in the retransmission request upper sequence number 601 of the redistribution request entry. The higher sequence number 762 is set, and the redistribution request lower sequence number 763 of the redistribution request message is set in the retransmission request lower sequence number 542 (steps 1600 to 1603).

(2) Next, the retransmission request entry is connected to the retransmission management entry corresponding to the node name requested to be redistributed, and the transmission management record has a serial number of the transmission request range and a record having the state 452 of “done” 462 exists. In this case, the state 452 is set to “re” 463, and if there is a record of the transmission request range in the retransmission management record and the state 552 is “done” 561, the state 552 is changed to “re” 562. Set (steps 1604 to 1606).

(3) It is determined whether there is a transmission management entry or a retransmission management entry corresponding to all the serial numbers in the retransmission request range, and if there is a corresponding transmission management entry or retransmission management entry, the data storage queue 210 of its own node Since the distribution data exists, the process here is terminated (steps 1607 and 1620).

(4) If it is determined in step 1607 that there is no corresponding transmission management entry or retransmission management entry, there is no distribution data in the data storage queue 210 of the own node, so a retransmission management record is created and retransmission management is performed. Connected to the entry, the state 552 is set to “re” 562, and then a retransmission management record is created and connected to another retransmission management entry, and the state 552 is set to “done” 561 (steps 1608 and 1609).

(5) Then, a redistribution request message creation process is performed, and the process here ends. Note that the redistribution request message creation processing will be described later with reference to FIG. 17 (steps 1610 and 1620).

  FIG. 17 is a flowchart for explaining the operation of the redistribution request message creation process for requesting data distribution to the upper node in step 1610 of the flow shown in FIG. 16, and this will be described next.

(1) When the process of creating a redistribution request message is started, the node that creates the message creates a redistribution request entry from the retransmission request entry, and all the serial numbers in the request range have already been changed to other redistribution request entries. If all the serial numbers in the request range have already been set in another redistribution request entry, it is not necessary to create a redistribution request message. The process ends (steps 1700 to 1702, 1710).

(2) If it is determined in step 1702 that all serial numbers in the request range are not set in other redistribution request entries, are some serial numbers in the request range already set in other redistribution request entries? If some serial numbers in the request range have already been set in other redistribution request entries, the redistribution request upper sequence number 601 and the redistribution request lower order are set so that the serial numbers do not overlap. The serial number 602 is changed. The redistribution request upper sequence number 601 and the redistribution request lower sequence number 602 will be described later with reference to FIG. 18 (steps 1703 and 1704).

(3) After the processing in step 1704 or in the determination in step 1703, if some serial numbers in the request range are not set in other redistribution request entries, the redistribution request entries are managed in the redistribution serial numbers. After connecting to the table 241 and then creating a redistribution request message 760 and transmitting it to the upper node, the processing here is terminated (steps 1705, 1706, 1710).

  FIG. 18 is a diagram for explaining an example of changing the serial number of the redistribution request entry. Next, this will be described. In the example described here, when the redistribution request entry a1800 and the redistribution request entry b1803 already exist, the redistribution request entry c1810 is newly created by the processing in step 1701 of the flow of FIG. 17 described above. It is an example.

  The redistribution request entry a1800 that already exists has the value of the upper sequence number 1801 for requesting redistribution as 10 and the value of the lower sequence number 1802 as 13, and the existing redistribution request entry b1803 is redistributed. It is assumed that the value of the upper sequence number 1804 for requesting is 19 and the value of the lower sequence number 1805 is 23. The redistribution request entry c1810 newly created by the processing in step 1701 of the flow in FIG. 17 has a value of 12 for the upper sequence number 1811 for requesting redistribution and a value of 15 for the lower sequence number 1812. It is said.

  In this case, since the redistribution request entry c1810 is changed except for the serial numbers 12 and 13, since the redistribution request entry c1810 already knows that the redistribution request has been notified from the existing redistribution request entry a1800. Changed to the redistribution request entry c1820. That is, the changed redistribution request entry c1820 is set to the redistribution request upper sequence number 1821 in which the value of the redistribution request upper sequence number is changed to 14, and the redistribution request lower sequence number remains at the value of 15. 1822.

  Thereafter, it is assumed that a redistribution request entry d1830 having a higher sequence number 1831 requesting redistribution of 18 and a lower sequence number 1822 of 25 is added. In this case, since the redistribution request entry b1803 already indicates that the redistribution request has been notified for the serial numbers 19 to 23, the redistribution request entry d1830 excludes the serial numbers 19 to 23. Changed to That is, the redistribution request entry d1830 is changed to a redistribution request entry da1840 and a redistribution request entry db1850. In the redistribution request entry da1840 created by this change, the value of the redistribution request upper sequence number 1841 is set to 18, the value of the redistribution request lower sequence number 1842 is set to 18, and the redistribution request entry db1850 The value of the serial number 1851 is set to 24, and the value of the redistribution request lower sequence number 1852 is set to 25.

  FIG. 19 is a flowchart for explaining the operation of the redistribution data reception process (step 906 in the flow shown in FIG. 9), which is the result of the redistribution request notified to the upper node.

(1) When the redistribution data receiving process is started, the node that has received the redistribution data has the same sequence number 731 of the distribution telegram 730 as the redistribution request upper sequence number 601 of any redistribution request entry in the own node. It is determined whether it is a value (steps 1900, 1901).

(2) If it is determined in step 1901 that the serial number 731 of the distribution message 730 has the same value as the redistribution request upper sequence number 601 of any redistribution request entry, the serial number of the distribution data existing in the data accumulation queue 210 It is determined whether or not there is a match with 411. If there is no match with the delivery data sequence number 411 in the data storage queue 210, the received re-distribution data is stored in the data storage queue 210 (step 1902). 1903).

(3) Thereafter, the retransmission management table update process at the time of receiving the redistribution data is performed, and 1 is added to the value of the redistribution request upper sequence number 601. The retransmission management table update processing will be described later with reference to the flow shown in FIG. 20 (steps 1904 and 1905).

(4) Next, it is determined whether or not the redistribution request upper sequence number 601 is larger than the redistribution request lower sequence number 602. If the redistribution request upper sequence number 601 is larger than the redistribution request lower sequence number 602, a redistribution request entry Is released 1907 and the processing is terminated (steps 1906, 1907, 1910).

(5) If it is determined in step 1901 that the serial number 731 of the distribution message 730 does not have the same value in the redistribution request upper sequence number 601 of any redistribution request entry, the determination in step 1902 If there is a match with the serial number 411 of the distribution data in the accumulation queue 210, or if the redistribution request upper sequence number 601 is not larger than the redistribution request lower sequence number 602 as determined in step 1906, the next re-transmission request The processing here is ended as waiting for the reception of the distribution data (step 1910).

  FIG. 20 is a flowchart for explaining the operation of the retransmission management table update process at the time of receiving redistributed data in step 1904 of the flow shown in FIG. 19, which will be described next.

(1) When the retransmission management table update process is started, the node that has received the redistribution data points to the top entry of the retransmission management table 251 so that the serial number 411 of the distribution data is in ascending order in the retransmission management entry. A new retransmission management record is inserted (steps 2000 to 2002).

(2) It is determined whether or not the redistribution data sequence number is included in the range of the redistribution request upper sequence number 601 and the redistribution request lower sequence number 602 of the already existing retransmission request entry. “Re” 562 is set in the state 552 of the retransmitted management record, and if it is not included, “Done” 561 is set in the state 552 of the added retransmission management record (steps 2003 to 2005).

(3) After the processing in step 2004 or step 2005, it is determined whether or not the value of the retransmission head pointer 533 is NULL. If the value is NULL, the retransmission head pointer 533 indicates the added retransmission management record. In this manner, the value of the retransmission start pointer 533 is set (steps 2006 and 2007).

(4) If the value of the retransmission head pointer 533 is not NULL after the processing in step 2007 or in the determination in step 2006, it is determined whether or not the updated entry is the last entry in the retransmission management table. If it is not the last entry, the next entry in the retransmission management table is pointed, and the process returns to Step 2002 to repeat the process. If it is not the last entry, the process here ends (Step 2008 to 2010).

  FIG. 21 is a flowchart for explaining the operation of the data redistribution process in step 1101 of the flow shown in FIG. 11. Next, this will be described.

(1) When data redistribution processing is started, a node that redistributes data determines whether or not the value of the retransmission start pointer 533 of the retransmission management entry in the relayer management table 251 is NULL. If there is NULL, there is no data to be redistributed, so the processing here ends (steps 2101 and 2130).

(2) If the value of the retransmission start pointer 533 of the retransmission management entry is not NULL in the determination at step 2101, the distribution data corresponding to the retransmission management record pointed to by the value of the retransmission start pointer 533 is retrieved from the data accumulation queue 210. It transmits to the lower node (step 2102).

(3) Thereafter, it is determined whether or not a normal response has been received from the lower node. If a normal response has been received, a retransmission request entry update process (to be described later) is performed according to the flow shown in FIG. It is determined whether or not there is a retransmission management record whose status 552 is “Re” 562 in the retransmission management records located after the next (steps 2103 to 2105).

(4) If there is a retransmission management record in which the status 552 is “Re” 562 in the determination in step 2105, the next retransmission management record is set in the retransmission head pointer 533. NULL is set to the pointer 533 (steps 2106 and 2107).

(5) After the processing of step 2106 or step 2107, the state 552 of the retransmission management record is set to “completed” 561, and the state 552 of all the retransmission management records corresponding to the distribution data becomes “completed” 561. If the status 552 of all the retransmission management records corresponding to the distribution data is not “Done” 561, the processing returns to Step 2101 and is repeated (Steps 2108 and 2109).

(6) If it is determined in step 2109 that the state 552 of all retransmission management records corresponding to the distribution data is “completed” 561, the entire retransmission management record corresponding to the distribution data is released and the distribution data is changed. The data is deleted from the data accumulation queue, and the process returns to the process from step 2101 to repeat the process (steps 2110 and 2111).

(7) If the normal response from the lower node cannot be received in the determination in step 2103, it is determined whether or not the abnormal response has been received from the lower node, and the abnormal response from the lower node cannot be received. Returning to the process from step 2101, the process is repeated, and if an abnormal response is received, it is determined that data loss has occurred and the process ends abnormally. This abnormal termination is because recovery by operation is necessary (step 2112).

  FIG. 22 is a flowchart for explaining the operation of the retransmission request entry update process in step 2104 of the flow shown in FIG. 11. Next, this will be described.

(1) When the retransmission request entry update process is started, 1 is added to the retransmission request upper sequence number 541 to determine whether or not the retransmission request upper sequence number 541 is larger than the retransmission request lower sequence number 542 (steps 2200 to 2202). ).

(2) If the retransmission request upper sequence number 541 is larger than the retransmission request lower sequence number 542 in the determination at step 2202, the retransmission request entry is released, the processing here is terminated, and the retransmission request upper sequence number 541 is retransmitted. If it is not larger than the requested lower sequence number 542, the processing here is terminated without doing anything (steps 2203, 2204).

  Each process in the above-described embodiment of the present invention is configured by a program and can be executed by a CPU included in the present invention. These programs are stored in a recording medium such as an FD, CDROM, or DVD. It can be provided and can be provided by digital information via a network.

It is a block diagram which shows the structural example of the distribution network of the tree structure which comprises the delivery system by one Embodiment of this invention. It is a figure explaining the detailed structure in a table group. It is a figure explaining the structure of each table which comprises a tree structure management table. It is a figure explaining the structure of delivery data, and the structure of a delivery management table. It is a figure explaining the structure of the retransmission management table in a redistribution management table. It is a figure explaining the structure of the redistribution request entry of the redistribution management table in a redistribution management table. It is a figure explaining the message | telegram format transmitted / received between nodes. It is a flowchart explaining the processing operation | movement in the node concerned when a new relay node is connected in the network of a tree structure. It is a flowchart explaining the operation | movement of the reception process in a relay node when distribution data is received from an upper node. 10 is a flowchart for explaining a processing operation for updating a transmission management table at the time of data reception in step 903 of the flow shown in FIG. 9. It is a flowchart explaining the operation | movement of the transmission process in a relay node when transmitting delivery data to a low-order node. It is a figure explaining the example which reconfigure | reconstructs the tree structure of a network at the time of a relay node failure. It is a flowchart explaining the processing operation in the node concerned in the case of reconfiguring the tree structure of a tree structure network. It is a flowchart explaining the operation | movement of the accumulation data deletion process in step 1320 of the flow shown in FIG. It is a flowchart explaining the operation | movement of the redistribution request process which notifies the redistribution request in step 1321 of the flow shown in FIG. 13 to an upper node. It is a flowchart explaining the operation | movement of the reception process of the redistribution request by the node notified from the lower node by the redistribution request. FIG. 17 is a flowchart for explaining the operation of a redistribution request message creation process for requesting data distribution to an upper node in step 1610 of the flow shown in FIG. 16. It is a figure explaining the example of the serial number change of a redistribution request entry. 10 is a flowchart for explaining an operation of a redistribution data reception process in step 906 of the flow shown in FIG. 9. FIG. 20 is a flowchart for explaining the operation of retransmission management table update processing when receiving redistributed data in step 1904 of the flow shown in FIG. 19. FIG. 12 is a flowchart for explaining the operation of data redistribution processing in step 1101 of the flow shown in FIG. 11. 12 is a flowchart for explaining the operation of retransmission request entry update processing in step 2104 of the flow shown in FIG. 11. It is a block diagram which shows the structure of a relay node.

Explanation of symbols

21 CPU
22 Memory 23 Table group 100 Distribution node 20, 101, 111, 112, 121-124 Relay node 131-138 Distribution destination node 2301 Data reception means 2302 Redistribution data reception means 2303 Data transmission means 2304 Redistribution data transmission means 2305 Redistribution Request means 2306 Redistribution request reception means 2307 Tree structure reconfiguration instruction means 2308 Tree structure reconfiguration response means 2309 Relay node connection means 210 Data accumulation queues 211, 212, 213 Distribution data 221 Reception sequence number management table 230 Distribution management table 231 Transmission management Table 240 Redistribution management table 241 Redistribution sequence number management table 251 Retransmission management table 270 Tree structure management table 271 Upper node table 272 Lower node table 273 Lower node table De connection table

Claims (7)

In a data distribution method in a data distribution system for distributing data from a distribution source node to a plurality of distribution destination nodes via a relay node,
The data distribution system is configured such that the distribution source node, relay nodes, and a plurality of distribution destination nodes are connected to each other so as to form a tree-structured network in which the relay nodes are provided in one or more stages. Has been
The distribution source node and the relay node hold the information of the lower node of the lower stage and the lower node of the lower stage in the storage means, and when a failure of the lower relay node is detected, Data that changes the structure of a tree structure in which a certain relay node is a relay node that is one level lower and updates information of a lower node that is one level lower and a lower node that is two levels lower in the storage means Delivery method.   The relay node receives data from the upper node, stores and stores the received data in a data accumulation queue, and records in the storage means which data in the data accumulation queue has been transmitted to each lower node. The serial number of the last received data of the relay node that has become a lower node due to the configuration change is acquired, and the data after the last received data serial number existing in the data accumulation queue is used as transmission data to the lower node, The data distribution method according to claim 1.   The relay node corresponds to an upper node when there is data that does not exist in the data accumulation queue among data from the last received data sequence number of the lower node to the last received data sequence number of the node that detected the failure. 3. The data distribution method according to claim 2, wherein a distribution number is requested to be transmitted.   4. The data distribution method according to claim 3, wherein the relay node notifies the upper node of a distribution request having a serial number that does not overlap with a serial number of data already requested for distribution to the upper node due to another lower node failure. . In a data distribution system that distributes data from a distribution source node to a plurality of distribution destination nodes via a relay node,
The data distribution system is configured such that the distribution source node, relay nodes, and a plurality of distribution destination nodes are connected to each other so as to form a tree-structured network in which the relay nodes are provided in one or more stages. Has been
The distribution source node and the relay node are: a holding unit that holds information of a lower node that is one step lower and a lower node that is two steps lower in a storage unit; a unit that detects a failure of a relay node that is one step lower; Means for changing the structure of a tree structure in which a relay node that is two steps below is detected as a relay node that is one step below when a failure of a relay node that is one step below is detected, and a lower node that is one step below in the holding means And a means for updating the information of the lower-level lower two nodes. A data distribution system comprising: A distribution source node, a relay node, and a plurality of distribution destination nodes are configured to be connected to each other so as to form a tree-structured network in which the relay node is provided in one or more stages. A relay node in a data distribution system that distributes data to a plurality of distribution destination nodes via a relay node,
The relay node has a storage unit that holds information of a lower node that is one step lower and a lower node that is two steps lower, a unit that detects a failure of the relay node that is one step lower, and a relay that is one step lower When a node failure is detected, a means for changing the structure of a tree structure with a relay node that is two steps lower, a means for receiving data from an upper node, and storing and storing received data Data storage queue, recording means for recording in the storage means which data in the data storage queue has been transmitted to each lower node, and the serial number of the last received data of the relay node that has become a lower node due to the configuration change A relay node apparatus comprising: means for obtaining; and means for setting data after the serial number of the last received data existing in the data accumulation queue as transmission data to a lower node. A distribution source node, a relay node, and a plurality of distribution destination nodes are configured to be connected to each other so as to form a tree-structured network in which the relay node is provided in one or more stages. In a data relay processing program at a relay node in a data distribution system that distributes data to a plurality of distribution destination nodes via a relay node,
Storing information in the storage means of the information of the lower node of the lower stage and the lower node of the lower stage, detecting the failure of the relay node of the lower stage, and detecting the failure of the relay node of the lower stage The step of changing the structure of the tree structure in which the relay node that is two steps below is the relay node that is one step below, and information on the lower node that is one step lower and the lower node that is two steps lower in the storage means Updating the data, receiving the data from the upper node, storing and storing the received data in the data storage queue, and recording which data in the data storage queue has been transmitted to each lower node Acquiring the serial number of the last received data of the relay node that has become a lower node due to the configuration change, and the data after the serial number of the last received data existing in the data accumulation queue And a step of transmitting data position to the node, the data relay processing program characterized by executing the steps to the CPU of the relay node.

Images