JP2006261949A - Distributed processing system dynamically changing processing execution node, and information processing device used in distribution processing system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a distributed processing system, etc. which dynamically changes an processing execution node according to present status information of each node. <P>SOLUTION: The system discriminates nodes to be always targets updating present information if a change occurs, and nodes to be targets to notice only that there happens a change, not content of the change if a change occurs in each node, and manages the present information regarding to each node. In a case that external input information is input, the most appropriate processing execution node is decided according to the present information of each node. In the decision, communication cost is used as a reference so that communication amount of the distributed processing system can be mostly reduced and efficiently processed. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a technique for dynamically changing a process execution node in a distributed system.

  A distributed processing system is a system in which processing is shared by a plurality of nodes (server-side or client-side information processing devices) connected by a network and the load is distributed. In such a distributed processing system, processing corresponding to the external input information is performed at any node in the system by an input from outside the system (hereinafter, a node performing processing corresponding to the external input information is designated as “ Called “Process Execution Node”).

  A node that can be a process execution node in the distributed processing system is a node having the external input information and the current state information of the node corresponding to the content of the input information. Here, the current status information of a node is information that the node has in the current status (newest status).

  By the way, if all nodes know the current status information of each node in advance, in the system, the node itself that has received the external input information can be a process execution node. However, if the current status information of each node that changes from moment to moment is updated in all nodes, the amount of communication increases and the efficiency of the entire system may be reduced. This problem is particularly serious in a system where the communication capacity in the network is not very large.

In addition, as a well-known document relevant to this application, there exist the following, for example.
JP 2000-112904 A

  The present invention has been made in view of the above circumstances, and eliminates the necessity of updating the current status information of each node that changes from time to time in all nodes, and processes according to the current status information grasped by each node. An object of the present invention is to provide a distributed processing system that dynamically changes execution nodes, thereby reducing communication load and increasing the efficiency of the entire system, and an information processing apparatus used in the distributed processing system.

  In order to achieve the above object, the present invention takes the following measures.

The first aspect of the present invention is:
A distributed processing system that includes a plurality of nodes that are connected via a network and that each execute predetermined information processing, wherein each of the plurality of nodes includes the plurality of nodes and the network that connects each node. The first information on the structure, the list of the current status information delivery destination nodes that are other nodes of the delivery destination that delivers the current status information of the own node, and the change notifying that a change has occurred in the own node Second information including a list of change flag delivery destination nodes that are other nodes to which the flag is delivered, presence / absence of current status information of other nodes, current status information of other nodes, and latest information of current status information of the other nodes Storage means for storing the third information including the received time, reception means for receiving the external input information, and a first node for processing the external input information If it is determined that the determination process of the first node is performed by the own node or the other node based on the third information, and the determination process of the first node is determined by the other node Using the first information and the second information, a communication cost that is an index that quantitatively indicates a communication load in the distributed processing system is calculated for each node, and the determination is performed based on the calculated value. First determination means for determining a second node to execute processing; request means for requesting the second node to perform the determination processing via the network; and the determination of the first node And a second determination unit that executes the determination process of the first node using the third information when it is determined that the process is to be performed by the own node. It is a processing system.

  According to a second aspect of the present invention, there is provided an information processing apparatus that configures a node used in a distributed processing system, is connected to another node via a network, and executes predetermined information processing. When the first information on the structure of the network connecting between the nodes, a list of current information distribution destination nodes that are other distribution destination nodes that distribute the current status information of the own node, and changes in the own node Second information including a list of change flag delivery destination nodes that are other nodes of the delivery destination that delivers a change flag that informs that there has been a change, presence / absence of current status information of other nodes, current status information of other nodes, Storage means for storing the third information including the time when the latest information on the current status information of the other nodes is received, receiving means for receiving external input information, and processing for the external input information. Determining means for determining whether the determination process of the first node is performed by the own node or another node based on the third information, and the determination process of the first node is performed by another node If it is determined, using the first information and the second information, the communication cost, which is an index that quantitatively indicates the communication load in the distributed processing system, is calculated for each node, and the calculated value is Based on first determination means for determining a second node for executing the determination processing, request means for requesting the second node to perform the determination processing via the network, and the first Second determination means for executing the determination process of the first node using the third information when it is determined that the determination process of the node is performed by the own node. An information processing device characterized by .

  As described above, according to the present invention, it is not necessary to update the current status information of each node that changes every moment, and the process execution node is dynamically changed according to the current status information grasped by each node. Thus, it is possible to realize a distributed processing system that reduces the communication load and increases the efficiency of the entire system, and an information processing apparatus used in the distributed processing system.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the following description, components having substantially the same function and configuration are denoted by the same reference numerals, and redundant description will be given only when necessary.

(First embodiment)
FIG. 1 is a diagram showing a configuration of a distributed processing system 1 according to the present embodiment. As shown in the figure, the distributed processing system 1 is composed of eight nodes from node A to node H connected by a network. Each node is network-connected to at least one other node, and can exchange data directly or indirectly with other nodes.

  In the following description, the own node is referred to as “own node”, and the other nodes are referred to as “other nodes”. Each node is assumed to know the network configuration of the distributed processing system 1 in advance. Furthermore, when the network configuration is dynamically changed, the latest network configuration can be acquired.

  Next, the internal configuration of each node will be described. Although each node has approximately the same configuration, the internal configuration of the node B will be described here for the sake of specific description.

  FIG. 2 is a block diagram showing the configuration of the node B. As shown in the figure, the node B includes the current state information distribution node management unit 10, the node configuration management unit 12, the node current state information management unit 14, the own node current state information distribution unit 16, the own node current state information management unit 18, and the own node. A current state information change flag distribution unit 20, an external input information analysis unit 22, a process execution unit 24, a process request unit 26, another node current state information change flag processing unit 28, another node current state information acquisition unit 30, and a communication interface unit 32. ing.

  The status information distribution node management unit 10 stores a status information distribution destination list. This current status information distribution destination list is a list of current status information distribution destination nodes (hereinafter simply referred to as “current status information distribution destination node list”) that distributes the current status information when each node is periodically or when there is a change. )) And a list of change flag distribution destination nodes that distribute only a change flag notifying the change but not the content of the change (hereinafter referred to simply as “change flag distribution destination node list”). ) Is defined for each node. FIG. 3 shows an example of the current status information distribution destination list stored by the current status information distribution node management unit 10.

  The node configuration management unit 12 includes node configuration information including at least each node corresponding range connected to the distributed processing system and a direct connection destination node (that is, a node directly connected without passing through another node). Remember and manage. FIG. 4 is a diagram illustrating an example of node configuration information stored in the node configuration management unit 12. As shown in the figure, in the node configuration information, each coordinate representing the corresponding range of each node in the coordinate space and a direct connection destination node (see FIG. 1) are defined for each node.

  The node current status information management unit 14 stores other node current status information including the presence / absence of predetermined information (for example, remaining battery amount) regarding other nodes and the time (time stamp) when the latest information of the predetermined information is received for each node. To remember. FIG. 5 is a diagram showing an example of other node status information stored in the node status information management unit 14. In the example shown in the figure, for the nodes A, B, C, D, and E, the remaining battery amount, the time stamp at which each remaining battery amount was received (updated), and the presence of data are F, G Since no data has been received for each of the nodes H and H, the fact that there is no data is stored as other node status information.

  Upon receiving a distribution request via the node status information management unit 14, the own node status information distribution unit 16 obtains a status information distribution destination node list from the status information distribution node management unit 10 and distributes the status information.

  The own node status information management unit 18 updates the current status information of the own node in the node status information management unit 14 periodically or when there is a change in the own node. In addition, when updating the current state information of the own node, the own node current state information management unit 18 requests distribution to the own node current state information change flag distribution unit 20 and the own node current state information distribution unit 16.

  Upon receiving a distribution request from the current node status information management unit 18, the own node status information change flag distribution unit 20 obtains a list of change flag distribution destination nodes from the current status information distribution node management unit 10, and distributes the current status information change flag. .

  The external input information analysis unit 22 determines candidate nodes from among the nodes A to H on the system using external input information received from the outside via the communication interface unit 32.

  The processing execution unit 24 performs processing node determination processing, which will be described later, and control of other components related thereto.

  The processing request unit 26 calculates a communication cost, which will be described later, based on the node configuration information stored in the node configuration management unit 12 and the current status information distribution destination node list stored in the current status information distribution node management unit 10. In addition, in the processing node determination process described later, the processing request unit 26 determines to which node the processing load is requested based on the value of the communication cost, and determines another node according to the determination. Requests processing to the processing execution unit 24 in its own node.

  When receiving the current status information change flag from another node, the other node current status information change flag processing unit 28 checks whether or not the current status information of the node is stored in the node current status information management unit 30. If the current status information already exists, the data of the node in the node current status information management unit 30 is deleted, and only the time stamp is stored. If there is no current status information, new current status information is stored.

  FIG. 6 is a flowchart showing a flow of processing executed by the other node current status information change flag processing unit 28. FIGS. 7A and 7B show examples of changing the node status information obtained by the processing according to the flow of FIG. That is, when the status information change flag of the node D is received in the state of FIG. 7A and the node status information is changed for the node D, the node status information is changed as shown in FIG. become. An example of the operation of the other node status information change flag processing unit 28 will be described with reference to each figure as follows.

  That is, for example, it is assumed that each node distributes the current status information and the current status information change flag with the contents shown in FIG. In such a case, the node B does not receive the current status information of the node D, but receives the current status information change flag. When node B has acquired the current status information of node D as a result of the previous processing, the data of the node current status information management unit 14 of node B is as shown in FIG. That is, the node B has the current status information of the node D.

  Further, during the period from when the node B acquires the current status information from the node D to when the node B receives the current status information flag of the node D, the node B has the latest current status information of the node D. In this normal state, the current status information change flag of a certain node is distributed, but if the current status information is acquired for some reason in a node where the current status information is not distributed, the status information change flag is received until the next status information change flag is received. The node is guaranteed to have the latest status information of the node. Thereby, although it is temporary, it becomes possible to reduce the present condition information acquisition cost.

  The other node current status information acquisition unit 30 receives the current status information distributed from the other node, and when the current status information of the other node is acquired by a request from the processing execution unit 24, the relevant node of the node current status information management unit 14 Update current status information for. The node current status information management unit 14 stores the updated current status information together with the time stamp at the time of the update.

  The communication interface unit 32 exchanges information with other nodes and sends the information to each element.

(Processing node decision process)
Next, a process execution node determination process executed in this distributed processing system will be described. This is to determine the most appropriate process execution node from time to time according to the current status information of each node. In this determination, a standard called communication cost, which will be described later, is used so that the communication amount of the present system can be reduced most effectively. As a result, it is not necessary to update the current state information of each node that changes from time to time in all the nodes, and the load in communication can be reduced and the efficiency of the entire system can be increased.

  FIG. 8 is a flowchart showing the flow of the process execution node determination process. In the present embodiment, for the sake of specific explanation, it is assumed that the node B acquires external input information and determines which node is to process the external input information.

  As shown in FIG. 8, first, when the communication interface unit 32 of the node B acquires external input information, it sends it to the external input information analysis unit 22 (step S1). The external input information analysis unit 22 determines a candidate node from the nodes A to H on the system using the received external input information (step S2). This determination is performed as follows, for example.

  FIGS. 9A and 9B are diagrams for explaining an example of candidate node selection processing. The external input information input by the node B includes, for example, position coordinates (X, Y) as shown in FIG. The external input information analysis unit 22 plots the position coordinates and the node configuration information of each node shown in FIG. 4 on the XY plane as shown in FIG. 9B (in FIG. 9B, the node G And the node having the corresponding range including the input position coordinates (X, Y) is selected as a candidate node. In the present embodiment, since the position coordinates shown in FIG. 9A are input, the node F and the node G are selected as candidate nodes as shown in FIG.

  The candidate node selection process can also be executed by the process execution unit 24.

  Next, the communication cost is calculated from the current status information (for example, resource information of each node) of the candidate node, and the processing node is determined from the candidate node based on this (step S3).

  FIG. 11 is a flowchart showing the flow of processing node determination processing in step S3. As shown in the figure, first, the process execution unit 24 of the node B inquires the node current state information management unit 14 about the current state information of the node F and the node G that are candidate nodes (step S30).

  The process execution unit 24 determines whether the current status information of the node F and the node G can be acquired based on the result of the inquiry to the node current status information management unit 14 (step S31). First, the processing node is determined in the own node (step S32).

  On the other hand, when the data held by the node current status information management unit 14 is the content of FIG. 5, for example, the data of the node F and the node G are not held. Therefore, the process execution unit 24 determines that acquisition is impossible in step S31, and requests the process request unit 26 to determine processing nodes from another node (step S33).

Next, the processing request unit 26 searches the data of the current status information distribution node management unit 10 and the node configuration management unit 12 to search for “communication cost required to acquire necessary current status information” and “self Two communication costs of “communication cost necessary for requesting processing from node to node” and the sum thereof are calculated (step S34). In this embodiment, the following equation (1) is adopted as the communication cost.

Here, S represents the data size and B _i represents the bandwidth of the communication channel i. The communication cost is the sum of the data size divided by the bandwidth of each communication path.

  For example, assume that the data size of the current status information of each node is 10 and the data size required for requesting the processing is 15. In addition, the bandwidth of each communication path is all 10. At this time, for example, the communication cost when requesting the node A to determine the processing node is the following two totals.

-Acquisition cost of current status information of node F and node G by node A
Node A has the current status information of node F. Therefore, only the current status information of the node G needs to be acquired. In addition to the node G, the node E and the node H have the current status information of the node G (see FIG. 3). When the node A acquires the current status information of the node G from any one of the node E, the node H, or the node G, the respective values are as follows. Communication cost when acquiring from node E = 10 × (1/10) = 1. Communication cost when acquiring from the node H = 10 × (1/10 + 1/10) = 2. Communication cost when acquiring from the node G = 10 × (1/10 + 1/10) = 2. Therefore, the node E has the lowest communication cost. Therefore, the communication cost for the node A to acquire the current status information between the node F and the node G is 1.

・ Communication cost for node B requesting node A
The number of paths from node B to node A is 2 (see FIG. 1). Therefore, the communication cost for the node B to request to the node A is 15 × (1/10 + 1/10).

  The results of the above calculation for each node are shown in FIG. From this result, it can be quantitatively understood that the cost can be minimized when the node E is requested.

  Next, the processing request unit 26 determines the processing node as to which processing node is the processing node based on the total value (index) of the calculated two communication costs, either in its own node or in another node. Is determined (step S35). If it is determined in step S35 that the processing node is determined by the own node, the processing request unit 26 requests the processing execution unit 24 of the own node to perform processing (step S36), and the processing execution unit 24 requests the processing. Is approved, it is determined that the processing node is determined by the own node (step S32). On the other hand, when it is determined in step S35 that the processing node is determined by another node, the processing request unit 26 requests the processing node to determine the processing node having the lowest total value of the two calculated communication costs. If the request is granted in the process execution unit 24 of the node E, it is determined that the process node determination is performed in the node E (another node) (step S37). Note that the external processing information is transferred from the node B to the node E through the route that minimizes the communication cost from the node B to the processing node by the confirmation in step S37.

  Next, a processing node is determined in the own node or another node (step S38). In this determination, a predetermined reference regarding the candidate node, for example, the largest node among the candidate nodes in the remaining battery amount among the candidate nodes is determined as the processing node.

  Note that in the node E, the processing from step S1 to step S3 is executed, and when the result itself determines the processing node, processing node determination based on the above criteria is executed. On the other hand, when it is determined that the node E is a processing node in another node, the same processing is repeated until the processing node is determined.

  Next, when a processing node is determined, processing of external input information is executed in the processing node (step S4 in FIG. 8).

  According to the configuration described above, the following effects can be obtained.

  According to this distributed processing system, in each node, when there is a change, the node that is always subject to update of the current status information, and when there is a change, only the fact that there is a change, not the content of the change, is to be notified The current status information about each node is managed by distinguishing the node. Therefore, it is not necessary for each node to grasp the current status information of all other nodes, and the load on communication can be reduced and the efficiency of the entire system can be increased.

  Further, according to the present distributed processing system, based on a standard called communication cost, a node that can reduce the communication amount most effectively and can determine a processing node efficiently is dynamically determined. As a result, it is not necessary to update the current status information of each node, which changes from moment to moment, in all the nodes, and the load in information communication can be reduced and the efficiency of the entire system can be increased.

(Second Embodiment)
Next, a second embodiment of the present invention will be described. The present embodiment relates to a distributed processing system having a tree structure. Even in such a case, it is possible to adopt the same configuration as in the first embodiment. However, here, a processing node determination method using a virtual node, which is different from the first embodiment, will be described.

  FIG. 13 is a diagram for explaining the structure (tree structure) of the distributed processing system according to the present embodiment. As shown in the figure, each node is directly or indirectly connected by a tree structure with node A as a vertex. Node F, node G, and node H collectively form one pseudo node (virtual node) X. Here, the virtual node is a pseudo (virtual) node composed of a plurality of different nodes in order to treat a plurality of nodes as one node in network communication. Hereinafter, each node configuring the virtual node is referred to as a configuration node.

  The virtual node is handled as follows in this distributed processing system. That is, all the information related to the configuration nodes F, G, and H is the information of the virtual node X. Accordingly, the nodes F, G, and H in the current information distribution destination list shown in FIG. 3, the node configuration information shown in FIG. 4, the other node current information shown in FIG. 5, etc. are all replaced with the virtual node X.

  Thus, the present distributed processing system in which the information of each of the configuration nodes F, G, and H is replaced with that of the virtual node X can be regarded as being configured by the nodes A to E and X in a pseudo manner. In this pseudo distributed processing system, the process execution node determination process described in the first embodiment can be applied. In the determination process, if it is determined that “the processing node is determined in the virtual node X”, the virtual node is regarded as one distributed processing system, and this is further compared in the first embodiment. The described process execution node determination process may be applied.

  In addition, it is preferable that the current status information of each configuration node in the virtual node is centrally managed by any one of the configuration nodes or a plurality of nodes. As a result, the amount of communication with the outside of the virtual node can be reduced.

  According to the configuration described above, by using a virtual node, in addition to the effects described in the first embodiment, the present invention is not limited to the current network phase structure, and can be fluidized as necessary. The network structure can be changed. This distributed processing system using virtual nodes is particularly beneficial when it has a tree structure. However, the present invention is not limited to the case of having a tree structure.

  Note that the present invention is not limited to the above-described embodiment as it is, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. Specific examples of modifications are as follows.

  (1) Each function according to the present embodiment can also be realized by installing a program for executing the processing in a computer such as a workstation and developing the program on a memory. At this time, a program capable of causing the computer to execute the technique is stored in a recording medium such as a magnetic disk (floppy (registered trademark) disk, hard disk, etc.), an optical disk (CD-ROM, DVD, etc.), or a semiconductor memory. It can also be distributed.

  (2) In each of the above embodiments, when the own node makes a processing node determination processing request to another node, the processing result performed by the other node side (processing node determination requested side) is It has not returned to the processing node determination request side). Therefore, it is preferable that the processing result by the processing node determination request side is fed back to the processing node determination request side as necessary. In such a case, it is preferable to add the cost for returning the processing result to the processing node determination requesting side using the communication cost as in the processing node determination processing.

  (3) In each of the above embodiments, the current status information distribution destination node list (see FIG. 3) is exemplified as having one path from the own node. However, regardless of this, a node having two or more paths may be set as the current status information distribution destination, or different nodes corresponding to a plurality of paths may be selected. It is preferable to arbitrarily select which node is the current status information distribution destination according to the characteristics of the system. The same applies to the selection of the change flag distribution destination node.

  In addition, various inventions can be formed by appropriately combining a plurality of components disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined.

  As described above, according to the present invention, it is not necessary to update the current status information of each node that changes every moment, and the process execution node is dynamically changed according to the current status information of each node, thereby communicating. It is possible to realize a distributed processing system that reduces the load on the system and increases the efficiency of the entire system, and an information processing apparatus used in the distributed processing system.

FIG. 1 is a diagram showing a configuration of a distributed processing system 1 according to the present embodiment. FIG. 2 is a block diagram showing the configuration of each node. FIG. 3 shows an example of the current status information distribution destination list stored by the current status information distribution node management unit 10. FIG. 4 is a diagram illustrating an example of node configuration information stored in the node configuration management unit 12. FIG. 5 is a diagram showing an example of other node status information stored in the node status information management 14. FIG. 6 is a flowchart showing a flow of processing executed by the other node current status information change flag processing unit 28. 7 (a) and 7 (b) are diagrams for explaining the change of the node current status information. FIG. 8 is a flowchart showing the flow of the process execution node determination process. FIGS. 9A and 9B are diagrams for explaining an example of candidate node selection processing. FIG. 10 is a diagram for explaining an example of candidate node selection processing. FIG. 11 is a flowchart showing the flow of processing node determination processing in step S3. FIG. 12 is a diagram showing necessary communication costs in each node. FIG. 13 is a diagram illustrating a schematic structure (tree structure) of the distributed processing system according to the second embodiment.

Explanation of symbols

A to H, X ... node, 10 ... current state information distribution node management unit, 12 ... node configuration management unit, 14 ... node current state information management, 16 ... own node current state information distribution unit, 18 ... own node current state information management unit, 20 ... own node current state information change flag distribution unit, 22 ... external input information analysis unit, 24 ... processing execution unit, 26 ... processing request unit, 28 ... other node current state information change flag processing unit, 30 ... other node current state information acquisition unit, 32 ... Communication interface

Claims (8)

A distributed processing system comprising a plurality of nodes connected via a network and each executing predetermined information processing,
Each of the plurality of nodes is
First information on the plurality of nodes and the structure of the network connecting the nodes, a list of current status information distribution destination nodes that are other distribution destination nodes that distribute the current status information of the own node, and the own node Second information including a list of change flag delivery destination nodes that are other nodes of the delivery destination that delivers a change flag that informs that a change has occurred when a change occurs, the presence / absence of current status information of other nodes, other nodes Storage means for storing current status information of the second node, third information including the time when the latest information of the current status information of the other node is received,
A receiving means for receiving external input information;
Determining means for determining whether the first node for performing processing on the external input information is to be performed by the own node or another node based on the third information;
When it is determined that the determination process of the first node is performed by another node, the first information and the second information are used to indicate the communication load in the distributed processing system quantitatively. A first determination unit that calculates a certain communication cost for each node and determines a second node that executes the determination process based on the calculated value;
Requesting means for requesting the determination processing to the second node via the network;
A second determination unit that executes the determination process of the first node using the third information when it is determined that the determination process of the first node is performed by the own node;
A distributed processing system comprising:   The distributed processing system according to claim 1, wherein the communication cost is a calculated value based on a sum of a quotient of a data size and a communication bandwidth between nodes.   The first determining means selects at least one candidate node from the plurality of nodes based on the external input information, calculates the communication cost for each of the at least one candidate node, and based on the calculated value The distributed processing system according to claim 1, wherein a second node that executes the determination process is determined.   The said 2nd determination means performs the said determination process of the said 1st node based on the resource information contained in the said 3rd information, The any one of Claim 1 thru | or 3 characterized by the above-mentioned. Distributed processing system. An information processing apparatus that configures a node used in a distributed processing system, is connected to another node via a network, and executes predetermined information processing,
First information on the plurality of nodes and the structure of the network connecting the nodes, a list of current status information distribution destination nodes that are other distribution destination nodes that distribute the current status information of the own node, and the own node Second information including a list of change flag delivery destination nodes that are other nodes of the delivery destination that delivers a change flag that informs that a change has occurred when a change occurs, the presence / absence of current status information of other nodes, other nodes Storage means for storing current status information of the second node, third information including the time when the latest information of the current status information of the other node is received,
A receiving means for receiving external input information;
Determining means for determining whether the first node for performing processing on the external input information is to be performed by the own node or another node based on the third information;
When it is determined that the determination process of the first node is performed by another node, the first information and the second information are used to indicate the communication load in the distributed processing system quantitatively. A first determination unit that calculates a certain communication cost for each node and determines a second node that executes the determination process based on the calculated value;
Requesting means for requesting the determination processing to the second node via the network;
A second determination unit that executes the determination process of the first node using the third information when it is determined that the determination process of the first node is performed by the own node;
An information processing apparatus comprising:   The information processing apparatus according to claim 5, wherein the communication cost is a calculated value based on a sum of quotients of a data size and a communication bandwidth between nodes.   The first determining means selects at least one candidate node from the plurality of nodes based on the external input information, calculates the communication cost for each of the at least one candidate node, and based on the calculated value The information processing apparatus according to claim 5, wherein a second node that executes the determination process is determined.   The said 2nd determination means performs the said determination process of the said 1st node based on the resource information contained in the said 3rd information, The any one of Claim 5 thru | or 7 characterized by the above-mentioned. Information processing device.

Images