JP5008075B2 - Web service system - Google Patents

Description

  The present invention relates to a web service system that provides a web service using a plurality of components, and in particular, from among a plurality of components that are distributed and arranged in an overlay network as alternative components for components used in the web service. The present invention relates to a Web service system to be searched.

  In recent years, the number of Internet users has increased explosively. This is because the environment for using the Internet has been established. For example, communication services such as WWW (World Wide Web) and electronic mail are generally used. High-speed access lines such as ADSL (Asymmetric Digital Subscriber Line), CATV Internet, FTTH (Fiber To The Home), etc. are generally used.

  As the number of Internet users increases, there is an increasing expectation for enhancement of services that can be enjoyed using the Internet. On the other hand, a distributed object technique has been proposed as a technique capable of providing a service using the Internet. Examples include CORBA (Common Object Request Broker Architecture) and DCOM (Distributed Component Object Model). In these distributed object technologies, if the architecture of the distributed object technology is different, interconnection between systems having different architectures is not easy. For example, in order to perform communication between CORBA and DCOM, a mechanism for converting both protocols and bridging them is required.

  Also, even between systems built with the same distributed object technology, messages are exchanged via the Internet, so the firewall settings must be changed. For example, CORBA uses a protocol called IIOP (Internet Inter-ORB Protocol). The firewall settings must be changed so that IIOP can be used. Therefore, these distributed object technologies are easy to use in a closed network such as an intranet or a LAN (Local Area Network). However, it is difficult to support various communication protocols and programs on the Internet. Nevertheless, with the spread of distributed object technology, there is an increasing expectation for a technology that can use objects operating on various platforms without being bound by a specific technology. In such a situation, as a technology that can be used between systems with different platforms, a web service using a versatile communication means has been proposed (for example, see Non-Patent Document 1).

  In Web services, technologies such as SOAP (Simple Object Access Protocol) and WSDL (Web Services Description Language) based on XML (eXtensible Markup Language), which are Internet standard technologies, are used as standard specifications. For this reason, distributed objects can be used across platform walls. A new Web service can also be constructed by using an existing Web service as a component. Furthermore, since the web service is constructed by combining existing components, it is easy to construct the web service, and it is possible to reduce the time and cost required for constructing the service.

  However, in the Web service, as the number of users of the Web service increases, the load on the node where the component used for the Web service is increased, and the processing capacity of the component decreases. Further, when a failure occurs in a component, a service using the component in which the failure has occurred cannot be provided. On the other hand, by using a plurality of components having the same function distributed and arranged on the network, the load on one component can be reduced and the scalability of the Web service can be improved. In addition, when a failure occurs in a component, the service is searched by searching for and using a component having the same function as that of the component in which the failure has occurred among a plurality of components distributed and arranged in a plurality of nodes. You can resume. Here, a Web service that uses a plurality of components having the same function distributed and arranged on a network is particularly referred to as a distributed Web service.

  However, in the distributed Web service, no clear method has been established for the alternative component search method at present. Generally, as a search method for alternative components, a reactive search method, a proactive search method, and the like can be given. Here, as a reactive search method, there is a search method based on flooding. Proactive search techniques include database management search techniques.

  Here, the search method based on flooding is a method of broadcasting a search message and selecting an alternative component from information included in a response message when searching for an alternative component (see, for example, Non-Patent Document 2). .

On the other hand, the search method based on database management is a method in which a node having a database periodically sends a control message to maintain information on the substitute component, and selects the substitute component from information registered in the database.
Masaru Shimamoto, Akira Enomoto, Susumu Nishimoto, etc., “Web Service Complete Construction Guide”, Nikkei Business Publications, 2001 Takahisa Okazaki, Hideo Kamada, Kazuhiko Kinoshita, Hideki Tode, Kozo Murakami, “Highly efficient flooding control considering multiple requirements”, IEICE Technical Report, IN2005-59, July 2005

  However, for the search method by flooding, the reliability of the information obtained by using this search method is high and suitable for obtaining dynamically changing information. There is a problem that the discovery of alternative components is slow.

  On the other hand, for the search method based on database management, the optimal alternative component can be found immediately from many candidates. However, if a failure occurs in the database holding the alternative component information, the alternative component search is performed. There is a problem of becoming impossible.

  Therefore, the present invention has been made in view of the above problems, and when a failure occurs in a component, an optimal substitute component is quickly selected from a plurality of components distributed and arranged on the network. An object of the present invention is to provide a Web service system that uses a search method for searching with the Internet.

  In order to achieve the above object, a Web service system according to the present invention has the following features.

  (A) An application server that provides a Web service using a plurality of components, and a plurality of nodes that are connected to the application server via a physical network and in which the plurality of components are distributed. In the Web service system, each of the plurality of nodes includes: (a1) logic formed between one or more components arranged in the own node and one or more components arranged in another node A link database in which link information indicating a link of a network is registered; and (a2) when requested by the application server to search for a component having the same function as a predetermined component used in the Web service, Registered in the link database And based on the link information and the searching one or more components which form a logical network, and a substitute component search processing unit selects the optimum components as a replacement for the given component.

  As a result, one or more components forming a logical network are managed with one or more components arranged in the own node in the link database provided in each node. At the time of search, an optimal alternative component can be found early among the components forming the logical network based on the link database.

  Further, (b) each of the plurality of nodes includes (b1) execution processing means for executing a component used in the Web service among one or more components arranged in the own node; ) Utility information generation processing for measuring the processing capability of the component being executed by the execution processing means, and using the result obtained by the measurement to generate utility information that is an index of utility obtained by executing the component (B3) Among the plurality of nodes, a node that searches for the optimum component is set as a search proxy node, and a node in which a component having the same function as the predetermined component is arranged is searched In the case of a node, the alternative component search processing means in the search proxy node is the search pair. On the basis of the utility information utility information generated by the generating means in the node, the highest component the utility may be elect as the best components.

  As a result, utility information relating to the processing capability of the components forming the logical network is generated. At the time of searching, an optimal substitute component can be found at an early stage from the components forming the logical network based on the utility information.

  Or (c) each of the plurality of nodes has the same function as the first component among the first component arranged in its own node and one or more components arranged in other nodes. An overlay network construction processing unit that forms a logical network with the second component and registers link information indicating the link of the formed logical network in the link database may be provided.

  As a result, a logical network is formed between one or more components having the same function. At the time of search, an optimal alternative component can be found at an early stage from among the components forming the logical network.

  The present invention is not only realized as a web service system, but also a web service control method for controlling the web service system, a web service control program for causing one or more computer systems to execute the web service control method, and web service control It may be realized as a recording medium on which a program is recorded.

  Further, the present invention is realized as an application server constituting a Web service system, an application server control method for controlling the application server, an application server control program for causing a computer system to execute the application server control method, a recording medium on which the application server control program is recorded, and the like. It may be.

  Further, it may be realized as a node constituting the Web service system, a node control method for controlling the node, a node control program for causing the computer system to execute the node control method, a recording medium on which the node control program is recorded, and the like.

  In the present invention, the search range of alternative components is limited by using an overlay network. Furthermore, by using the utility information, a high-performance component is selected within a limited search range. As a result, a high-performance alternative component can be searched at high speed. In addition, the load can be distributed among the nodes, and a scalable system can be configured.

  Furthermore, it is possible to avoid the biased election of a specific component or a component placed on a specific node as an alternative component during disaster recovery, and improve fairness between components or between nodes. Can do.

  Moreover, since it is based on autonomous distributed processing, there is no single point of failure and it has resistance against a plurality of failures.

(Embodiment 1)
Embodiment 1 of the present invention will be described below with reference to the drawings.

  The Web service system in the present embodiment has the features shown in the following (a) and (b).

  (A) A web service comprising an application server that provides a web service using a plurality of components, and a plurality of nodes that are connected to the application server via a physical network and in which a plurality of components are distributed. Each of the plurality of nodes is a link of a logical network formed between (a1) one or more components arranged in its own node and one or more components arranged in another node When the application server requests for a link database in which link information indicating the link information is registered and (a2) search for a component having the same function as a predetermined component used in the Web service, the link database is registered in the link database. Based on link information Te searches one or more components which form a logical network, and a alternate component search processing unit for selecting the optimum components as an alternative to the given component.

  (B) Each of the plurality of nodes includes: (b1) an execution processing unit that executes a component used in the Web service among one or more components arranged in the own node; and (b2) an execution processing unit. A utility information generation processing unit that measures the processing capacity of the component being executed in step (b), generates utility information that is an index of utility obtained by executing the component, using a result obtained by the measurement, (B3) In the case where a node that searches for an optimum component among a plurality of nodes is set as a search proxy node, and a node in which a component having the same function as a predetermined component is arranged as a search target node, the search proxy The alternative component search processing unit in the node uses the utility information generated by the utility information generation processing unit in the search target node. Based on, selects utility the highest component as the best components.

  Based on the above points, the Web service system in the present embodiment will be described.

<Overview>
First, an outline of the Web service system in the present embodiment will be described.

  FIG. 1 is a diagram showing an outline of a Web service system according to the present embodiment. As shown in FIG. 1, the Web service system 100 searches for alternative components using utility information distributed and managed under the following conditions (1) and (2).

  (1) An overlay network is constructed among a plurality of components having the same function.

  Here, the component is a software part having an interface. An overlay network is a logical network in which virtual links are formed using links in a physical network.

  (2) Utility information regarding components is exchanged between adjacent components in the overlay network. However, utility information is exchanged only between components having links formed on the overlay network.

  Here, the utility information varies between 0 and 1 depending on the processing capability of the node in which the alternative component is arranged, the delay on the network with respect to the node, the number of application servers using the component, and the like. Information including a value (hereinafter referred to as “utility value”). A node is a physical resource of a computer system, such as hardware.

  Hereinafter, the Web service is an application service configured by combining components specified by a URI (Uniform Resource Identifier) using XML (eXtensible Markup Language), HTTP (Hyper Text Transfer Protocol), or the like. .

  The nodes 101 to 105, the application server 106, the user terminal 107, and the like are connected to each other via a physical network, and are TCP / IP (Transmission Control Protocol / Internet Protocol), UDP / IP (User Datagram Protocol / Internet Protocol). Data is transmitted and received in packet units based on a communication protocol such as.

  In searching for an alternative component, it is important to consider the time until service restart and the quality of service after restart. Here, the utility value is used as the latter index value.

  The time until the service is restarted is a value obtained by adding the time until the replacement component is discovered and the RTT (Round Trip Time) of the message notifying that the replacement component is used.

  Further, the utility value and the time until service restart are in a trade-off relationship. This can be seen from the following points. For example, in order to use a high utility component, that is, a component having a utility value close to 1, as a substitute component, it is necessary to obtain and compare utility values of a plurality of components. However, in order to obtain and compare the utility values of a plurality of components, it takes time to search, and the time until service restarts increases.

<Alternative component search method>
Next, an alternative component search method used in the Web service system according to this embodiment will be described.

  It is assumed that the application server provides a web service using one or more components. In addition, the address of the search proxy node for the component is transmitted in advance from the component used for the Web service. In the database provided in the application server, the address of the search proxy node transmitted from the component is registered.

  Here, the search proxy node is a node in which a component having the largest number of links among one or more components adjacent to the component used in the Web service in the overlay network is arranged.

  The application server stops providing the service if a failure occurs in any of the components used to provide the service. Based on a plurality of addresses registered in the database, a search proxy node for a component in which a failure has occurred is specified. A search request for causing the identified search proxy node to search for a component having the same function as the component in which the failure has occurred is transmitted.

  Based on the search request transmitted from the application server, the search proxy node specifies a component having the same function as the component in which the failure has occurred from one or more components arranged in the search proxy node. A component adjacent to the identified component is selected from a plurality of components forming the overlay network. Queries the node (hereinafter referred to as “search target node”) where the selected component (hereinafter referred to as “search target component”) is arranged for information on the utility value and address of the search target component. A utility information inquiry request is sent. At this time, when a plurality of components are selected, utility information inquiry requests are individually transmitted to the plurality of search target nodes.

  The search target node specifies information related to the utility value and address of the search target component from the information managed by the search target node based on the utility information inquiry request transmitted from the search proxy node. The identified information is transmitted to the search proxy node as a response to the utility information inquiry request.

  Based on the information transmitted from the search target node, the search proxy node selects a component having a utility value close to 1 as an optimum component as a substitute for the component in which the failure has occurred from one or more search target components. Component information for identifying the selected component is transmitted to the application server as a response to the search request.

  Based on the component information transmitted from the search proxy node, the application server switches from the component in which the failure has occurred to the component specified by the component information and resumes providing the web service.

<When using>
For example, as shown in FIG. 2, the application server 106 provides a Web service that uses the components b1, c4, and e3. Note that from each component, the address of the node where the component having the largest number of links among one or more components adjacent to each component in the overlay network is transmitted as the address of the search proxy node. In the database provided in the application server 106, the address of the search proxy node transmitted from each component is registered.

  Here, as an example, the address of the node 103 where the component b3 is arranged is transmitted from the component b1. In the database, the address of the node 103 is registered as the address of the search proxy node for the component b1. From the component c4, the address of the node 105 where the component c5 is arranged is transmitted. In the database, the address of the node 105 is registered as the address of the search proxy node for the component c4. From the component e3, the address of the node 105 where the component e5 is arranged is transmitted. In the database, the address of the node 105 is registered as the address of the search proxy node for the component e3.

<In case of failure>
Next, as illustrated in FIG. 3, when a failure occurs in the component c4, the application server 106 stops providing a web service that uses the component c4.

<When searching>
At this time, as shown in FIG. 4, the application server 106 identifies the node 105 as a search proxy node for the component c4 based on the address registered in the database. A request is sent to the identified node 105 to search for an optimum component as an alternative component of the component c4.

  Accordingly, the node 105 selects components c2 and c3 adjacent to the component c5 having the same function as that of the component c4 among the plurality of components forming the overlay network. The utility information of the selected components c2 and c3 is individually acquired from the nodes 102 and 103 where the selected components c2 and c3 are arranged. Based on the two pieces of utility information acquired individually, the component c2 is selected as the component having the highest utility, that is, the component having a utility value close to 1.

<When resuming>
Then, as illustrated in FIG. 5, the node 105 transmits component information for specifying the selected component c2 to the application server 106 as a response to the request.

  Based on the component information transmitted from the node 105, the application server 106 resumes the provision of the Web service using the component c2 as an alternative to the component c4 in which the failure has occurred.

<Overlay network construction method>
Next, an overlay network construction method used in the Web service system according to this embodiment will be described.

  The performance of the alternative component search method described above depends greatly on the topology of the overlay network.

  For example, in order for a search proxy node to search for a component with high utility, it is desirable to form links with more components. However, utility information that must be managed increases as the number of links increases. For this reason, in a large-scale network, it is desirable to set an upper limit on the total number of links of one or more components arranged in each node from the viewpoint of management load. In general, one or more components are arranged in one node.

  Therefore, it is desired to create a topology that can be efficiently searched while achieving load distribution among the components. For this reason, an overlay network is constructed here so that the following specifications are satisfied.

  Specifically, in the overlay network, when the number of links of the component ai arranged in the node i is dia, the restriction expressed by the following formula (1) is provided for the total number of links di of the node i. . Here, const is a constant determined by the processing capability of the node. However, here, in order to simplify the story, a constant value is assumed for all nodes.

  (1) di = dia + dib + dic +... ≦ const

  For each component, it is assumed that at least two links are formed in the overlay network. This is because, when only one link is formed, if a failure occurs in a link destination component, the link source component is isolated in the overlay network. This is because information regarding the utility value, address, etc. of the search target component is not queried at the time of search although it is the search target component.

  Further, it is considered that the time required for the search and the traffic amount flowing on the physical network are related to the number of hops between the nodes of the physical network. Therefore, when constructing an overlay network, an attempt is made to sequentially form a link with one or more components arranged in a node having a small number of hops in the physical network.

  When the upper limit of the total number of links of the node is reached at the time of link formation, any link of the components arranged in the node is deleted. Thereby, fairness is maintained between one or more components arranged in the node. Here, the fairness means that, when searching for an alternative component, the alternative component can be found to the same extent in any component.

<Construction procedure>
Next, the procedure of an overlay network construction method that satisfies the above specifications will be described.

  6 and 7 are diagrams showing the procedure of the overlay network construction method in the present embodiment. As illustrated in FIG. 6, here, as an example, it is assumed that a component ai is newly arranged in the node i. In this case, first, LLN (Least Link Number) is set as the lower limit of the number of links of the component. MLN (Max Link Number) is set as the upper limit of the total link number di of the node i. By arranging the component a at the node i, the expected value E (dia) of the average number of links of the components arranged at the node i is calculated based on the following equation (2).

  (2) E (dia) = di / (number of components arranged in node i)

  Next, as long as the total number of links di of the node i is not greater than or equal to the MLN (S101: No), an attempt is made to form a new link with any one of a plurality of components constituting the overlay network. At this time, the link destination component is selected by the following procedures (1) to (3) (S102).

  (1) A component having the same function as that of the component ai is selected from among a plurality of components constituting the overlay network.

  (2) Among the components selected in step 1, select components that are not linked to component ai.

  (3) Among the components selected in the procedure 2, the component arranged in the node nearest to the node i in the physical network forming the overlay network is selected.

  Here, as an example, it is assumed that the component aj arranged in the node j adjacent to the node i is selected as the link destination component.

  Next, when the total link number dj of the node j is greater than or equal to the MLN set in the node j (S103: Yes), one of the links of one or more components arranged in the node j is deleted. (S104). At this time, as shown in FIG. 7, a link to be deleted is selected in the following procedures (1) to (3) (S111, S112), and the selected link is deleted (S113).

  (1) Among the one or more components arranged at the node j, the component having the largest number of links is selected (S111).

  (2) Among the one or more components that form a link with the selected link source component, the component with the largest number of links is selected (S112).

  (3) Deletes the link formed by the component selected in step 1 and the component selected in step 2 among the one or more links formed by one or more components arranged in node j. (S113).

  For example, it is assumed that in the procedure 1, the component bj is selected as the component having the largest number of links among one or more components arranged in the node j. Assume that component bk is selected as the component with the largest number of links among the one or more components that form links with component bj in procedure 2. In this case, the link olb (j, k) formed by the component bj and the component bk is selected in the procedure 3.

  Then, a link is formed between the component ai and the component aj (S105).

  However, if the link number dkb of the component bk is smaller than the LNN by deleting the link olb (j, k), the link to be deleted is re-selected except for the link olb (j, k). .

  If there is no link to be deleted, the component ai gives up forming a link with the component aj. The component aj is excluded from the components that are not linked to the component ai.

  Further, when the total link number di of the node i becomes greater than or equal to MLN (S101: Yes) and the link number dia of the component ai is smaller than the expected value E (dia) (S106: No), the above (1) to (3 In the same manner as in the procedure), the links to be deleted are selected in the following procedures (1) to (3), and the selected links are deleted (S107). As a result, the component ai can form a new link.

  (1) A component having the largest number of links is selected from one or more components arranged in the node i (S111).

  (2) Among the one or more components that form a link with the selected link source component, the component with the largest number of links is selected (S112).

  (3) Delete the link formed by the component selected in step 1 and the component selected in step 2 among the one or more links formed by one or more components arranged in node i. (S113).

  Specifically, it is assumed that a component b4 is newly arranged in the node 104 as shown in FIG. In this case, as shown in FIG. 9, the component b4 forms a link olb (4, n) from one or more components arranged in the adjacent node n in the physical network forming the overlay network. Try. At this time, in forming a link with the component b5, the total number of links d5 of the node 105 is equal to or greater than the MLN set in the node 105. Therefore, as shown in FIG. 10, the component b5 having the largest number of links among the components arranged in the node 105 is selected. Further, the component b1 having the largest number of links is selected from the components that form links with the component b5. The link olb (1, 5) between the component b5 and the component b1 is deleted, and then the link olb (4, 5) between the component b4 and the component b5 is added.

  Since the total link number d4 of the node 104 is not equal to or greater than the MLN set in the node 104, the link with other components is continuously formed. In this way, a new overlay network topology is formed as shown in FIG.

<Configuration>
Next, the configuration of the Web service system in the present embodiment will be described.

  FIG. 12 is a diagram showing a configuration of the Web service system in the present embodiment. As shown in FIG. 12, here, as an example, the Web service system 100 includes nodes 101 to 105, an application server 106, and the like.

  Each of the nodes 101 to 105 is a computer system that provides each function of a plurality of components arranged in each node to another node such as the application server 106 via a physical network. Here, as an example, the nodes 101 are arranged with components a1, b1, and c1. In the node 102, components a2, c2, and d2 are arranged. In the node 103, b3, c3, and e3 are arranged. In the node 104, components a4, c4, and d4 are arranged. In the node 105, components c5, d5, and e5 are arranged.

  The components a1, a2, and a4 are components having the same function. The components b1 and b3 are components having the same function. The components c1, c2, c3, and c4 are components having the same function. The components d2, d4, and d5 are components having the same function.

  The application server 106 is a computer system that provides a Web service that combines the functions of the components b1, c4, and e3 to another node such as the user terminal 107 via a physical network.

  When the component used in the Web service (hereinafter referred to as “used component”) becomes unavailable, the application server 106 replaces the component used (hereinafter referred to as “alternative component”). Is searched from among a plurality of components distributed in the overlay network.

<Node>
Next, the configuration of the nodes 101 to 105 in the present embodiment will be described. The nodes 101 to 105 have the same configuration, and only the node 101 will be described here, and the description of the nodes 102 to 105 will be omitted.

  FIG. 13 is a diagram illustrating a configuration of the node 101 in the present embodiment. As illustrated in FIG. 13, the node 101 includes a component holding unit 111, a component database 112, a node database 113, a link database 114, a local node utility information storage unit 115, another node utility information storage unit 116, and a component execution processing unit 120. , Utility information generation processing unit 130, alternative component search processing unit 140, utility information transmission processing unit 150, and the like.

  The component execution processing unit 120, the alternative component search processing unit 140, the utility information transmission processing unit 150, and the like may be realized on a computer system using a distributed object technology such as CORBA, DCOM, or SOAP. It may be realized on a computer system using a distributed object technology other than these. Here, as an example, it is assumed that it is realized on a computer system using SOAP.

<Component holding unit 111>
The component holding unit 111 holds one or more components arranged in its own node.

<Component database 112>
In the component database 112, component information related to components arranged in the own node is registered.

  FIG. 14 is a diagram showing a data structure of component information managed in the component database 112. As shown in FIG. As shown in FIG. 14, here, as an example, the component database 112 includes component identification information (field F111), node identification information (execution node) (field F112), node identification information (search proxy node) (field F113) and the like component information (record R111) is registered (table T111).

  The component identification information (field F111) is information for identifying a component among one or more components constituting the overlay network, such as a component name.

  The node identification information (execution node) (field F112) is information for identifying the node on which the component specified by the component identification information is executed.

  The node identification information (search proxy node) (field F113) is information for identifying a node that searches for a component having the same function as the component specified by the component identification information.

<Node database 113>
In the node database 113, node information related to other nodes adjacent to the own node is registered.

  FIG. 15 is a diagram showing a data structure of node information managed in the node database 113. As shown in FIG. 15, here, as an example, the node database 113 includes node information (record R121) including node identification information (field F121), the number of hops (field F122), RTT (field F123), and the like. Is registered (table T121).

  The node identification information (field F121) is information for identifying each node among one or more nodes constituting the physical network, such as a host name and an IP address.

  The number of hops (field F122) is the number of routers or the like that pass between the own node and the node specified by the node identification information.

  RTT (field F123) is a round trip transmission time between the own node and the node specified by the node identification information.

<Link database 114>
In the link database 114, a component (hereinafter referred to as “link source component”) arranged in the own node and a component (hereinafter referred to as “link destination component”) arranged in another node. Link information related to the link formed between and is registered.

  FIG. 16 is a diagram showing a data structure of link information managed by the link database 114. As shown in FIG. 16, here, as an example, link information (record R131) including link source information (field F131), link destination information (field F132), and the like is registered in the link database 114. (Table T131).

  The link source information (field F131) includes link source component identification information, link source node identification information, and the like. The link destination information (field F132) includes link destination component identification information, link destination node identification information, and the like.

  In this embodiment, a computer system is used in advance to execute a simulation based on the above overlay network construction technique. Assume that link information generated based on a result obtained by execution is registered in the link database 114.

<Own node utility information storage unit 115>
The own node utility information storage unit 115 stores utility information for each component executed in the own node.

  FIG. 17 is a diagram showing a data structure of utility information stored in the own node utility information storage unit 115 in the present embodiment. As shown in FIG. 17, the own node utility information storage unit 115 stores utility information (record R141) for each component executed in the own node (table T141). The utility information (record R141) includes the component identification information (field F141) of the component executed in the own node, the utility value (field F142) of the component, and the like.

<Other node utility information storage unit 116>
The other node utility information storage unit 116 stores utility information for each component executed in the other node.

  FIG. 18 is a diagram showing a data structure of utility information stored in the other node utility information storage unit 116 in the present embodiment. As shown in FIG. 18, the other-node utility information storage unit 116 stores utility information (record R151) for each component executed in the other node (table T151). The utility information (record R151) includes component identification information (field F151) of a component executed by another node, utility value (field F152) of the component, and the like.

<Detailed configuration>
FIG. 19 is a diagram showing a detailed configuration of the node 101 in the present embodiment. As illustrated in FIG. 19, the node 101 includes the following components and the like.

<Component execution processing unit 120>
The component execution processing unit 120 includes a component execution unit 121, an execution request reception unit 122, an execution response transmission unit 123, and the like.

<Component execution unit 121>
The component execution unit 121 receives component identification information from the execution request reception unit 122. The component specified by the received component identification information is read from the component holding unit 111. Executes the read component. The execution result obtained by execution is passed to the execution response transmission unit 123.

<Execution request receiving unit 122>
The execution request receiving unit 122 receives an execution request message transmitted from a transmission source such as the application server 106. The node identification information for identifying the transmission source of the received execution request message is passed to the execution response transmission unit 123. Analyze the received execution request message. Component identification information for identifying the component to be executed is extracted from the analyzed execution request message. The extracted component identification information is passed to the component execution unit 121.

<Execution response transmitter 123>
The execution response transmission unit 123 receives the node identification information from the execution request reception unit 122. An execution result is received from the component execution unit 121. A response message including the received execution result (hereinafter referred to as “execution response message”) is generated. The generated execution response message is transmitted to the transmission destination specified by the received node identification information.

<Utility information generation processing unit 130>
The utility information generation processing unit 130 includes a measurement unit 131, a utility information generation unit 132, and the like.

<Measurement unit 131>
When the component is executed by the component execution unit 121, the measurement unit 131 measures the performance associated with the execution of the component. The component identification information for identifying the component and the measurement result obtained by measurement are passed to the utility information generation unit 132.

<Utility information generation unit 132>
The utility information generation unit 132 receives component identification information and measurement results from the measurement unit 131. The utility value is calculated using the received measurement result. The utility value obtained by calculation is associated with the received component identification information, and utility information including the received component identification information and the utility value obtained by calculation is generated. The generated utility information is written in the own node utility information storage unit 115.

<Alternative component search processing unit 140>
The alternative component search processing unit 140 includes an alternative component search unit 141, a search target selection unit 142, a utility information writing unit 143, a search request reception unit 144, a search response transmission unit 145, a utility information inquiry request transmission unit 146, and a utility information inquiry. A response receiving unit 147 and the like are provided.

<Alternative component search unit 141>
The alternative component search unit 141 receives component identification information from the search request reception unit 144. The received component identification information is passed to the search target selection unit 142. A selection completion notification is received from the search target selection unit 142. Component identification information for identifying a component having a high utility value among one or more components is selected from a plurality of utility information stored in the other node utility information storage unit 116. Component information specified by the selected component identification information is selected from one or more component information registered in the component database 112. The selected component information is passed to the search response transmission unit 145.

<Search target selection unit 142>
The search target selection unit 142 receives component identification information from the alternative component search unit 141. Component information of a component having the same function as the component specified by the received component identification information is selected from one or more component information registered in the component database 112. All of the link information including the selected component information in the link source information is selected from one or more link information registered in the link database 114. The following process (1) is executed for each of the selected one or more link information.

  (1) The search target selection unit 142 passes the node identification information and component identification information included in the link destination information of the link information to the utility information inquiry request transmission unit 146.

  When the search target selection unit 142 executes the above process (1) for all the selected one or more link information, the search target selection unit 142 passes a selection end notification to the alternative component search unit 141.

  Note that the search target selection unit 142 performs the above process (1) for each of one or more link information registered in the link database 114 at predetermined time intervals, and obtains utility information from other nodes. You may exchange regularly. In this case, when the component identification information is received from the alternative component search unit 141, the processing up to the above (1) may be omitted and the selection end notification may be immediately passed to the alternative component search unit 141.

<Utility information writing unit 143>
The utility information writing unit 143 receives the utility information from the utility information inquiry response receiving unit 147. The received utility information is written in the other node utility information storage unit 116.

<Search request receiving unit 144>
The search request receiving unit 144 receives a search request message transmitted from a transmission source such as the application server 106. The node identification information that identifies the transmission source of the received search request message is passed to the search response transmission unit 145. Analyze the received search request message. Component identification information for identifying a component to be searched is extracted from the analyzed search request message. The extracted component identification information is passed to the alternative component search unit 141.

<Search response transmission unit 145>
The search response transmission unit 145 receives node identification information from the search request reception unit 144. Component information is received from the alternative component search unit 141. A response message including the received component information (hereinafter referred to as “search response message”) is generated. The generated search response message is transmitted to the transmission destination specified by the received node identification information.

<Utility information inquiry request transmission unit 146>
The utility information inquiry request transmission unit 146 receives node identification information and component identification information from the search target selection unit 142. A request message including the received component identification information (hereinafter referred to as “utility information inquiry request message”) is generated. The generated utility information inquiry request message is transmitted to the transmission destination specified by the received node identification information.

<Utility information inquiry response receiving unit 147>
The utility information inquiry response receiving unit 147 receives a response message to the utility information query request message transmitted from the utility information query request transmitting unit 146 (hereinafter referred to as “utility information query response message”). Analyze the received utility information inquiry response message. The obtained utility information is extracted from the analyzed utility information inquiry response message. The extracted utility information is passed to the utility information writing unit 143.

<Utility information transmission processing unit 150>
The utility information transmission processing unit 150 includes a utility information reading unit 151, a utility information query request receiving unit 152, a utility information query response transmitting unit 153, and the like.

<Utility information reading unit 151>
The utility information reading unit 151 receives component identification information from the utility information inquiry request receiving unit 152. The utility information specified by the received component identification information is read from the own node utility information storage unit 115. The read utility information is passed to the utility information inquiry response transmission unit 153.

<Utility information inquiry request receiving unit 152>
The utility information query request receiving unit 152 receives a utility information query request message transmitted from a transmission source such as another node. The node identification information specifying the transmission source of the received utility information inquiry request message is passed to the utility information inquiry response transmission unit 153. Analyze the received utility information query request message. Component identification information for identifying the utility information to be acquired is extracted from the analyzed utility information inquiry request message. The extracted component identification information is passed to the utility information reading unit 151.

<Utility information inquiry response transmission unit 153>
The utility information inquiry response transmission unit 153 receives the node identification information from the utility information inquiry request reception unit 152. Receives utility information from the utility information acquisition unit. A utility information inquiry response message including the received utility information is generated. The generated utility information inquiry response message is transmitted to the transmission destination specified by the received node identification information.

<Application server>
Next, the configuration of the application server 106 in the present embodiment will be described.

  FIG. 20 is a diagram showing a configuration of the application server 106 in the present embodiment. As shown in FIG. 20, the application server 106 includes an execution target component database 161, an execution result storage unit 162, a service processing unit 170, a search processing unit 180, and the like.

<Execution target component database 161>
In the execution target component database 161, component information of components to be executed is registered.

  FIG. 21 is a diagram illustrating a data structure of component information managed in the execution target component database 161. As shown in FIG. 21, the data structure is the same as the component information registered in the component database 112.

  The component identification information (field F161) is component identification information for identifying a component to be executed.

  The node identification information (execution node) (field F162) is node identification information for identifying a node that executes the component specified by the component identification information.

  The node identification information (search proxy node) (field F163) is node identification information for identifying a node that searches for a component having the same function as the component specified by the component identification information.

<Execution result storage unit 162>
The execution result storage unit 162 stores execution result information for each executed component.

<Service processing unit 170>
20, the service processing unit 170 includes an execution target selection unit 171, an execution result writing unit 172, a service request reception unit 173, a service response transmission unit 174, an execution request transmission unit 175, an execution response reception unit 176, and the like. .

<Execution Target Selection Unit 171>
The execution target selection unit 171 receives one or more component identification information from the service request reception unit 173. The following process (1) is executed for each of the received one or more component identification information.

  (1) The execution target selection unit 171 selects node identification information specified by the received component identification information from one or more component information registered in the component database 112. The received component identification information and the selected node identification information are passed to the execution request transmission unit 175.

  When the execution target selection unit 171 executes the process (1) for all of the received one or more component identification information, the service response transmission unit 174 sends a response notification and the received one or more component identification information. To pass.

  When the execution target selection unit 171 receives the error notification and the component identification information from the execution response reception unit 176, the execution target selection unit 171 stops the above processing and passes the stop notification to the service response transmission unit 174. Further, the search notification and the component identification information received together with the error notification are passed to the search request transmission unit 182. When a search completion notification is received from the component information registration unit 181, a restart notification is passed to the service response transmission unit 174, and the above process is restarted.

<Execution result writing unit 172>
The execution result writing unit 172 receives the execution result information from the execution response receiving unit 176. The received execution result information is written into the execution result storage unit 162.

<Service request receiving unit 173>
The service request reception unit 173 receives a service request message transmitted from a transmission source such as the user terminal 107. The node identification information that identifies the transmission source of the received service request message is passed to the service response transmission unit 174. Parses the received service request message. One or more components to be executed are identified from the analyzed service request message. One or more component identification information for individually identifying the identified one or more components is passed to the execution target selection unit 171.

<Service response transmission unit 174>
The service response transmission unit 174 receives node identification information from the service request reception unit 173. A response notification and one or more component identification information are received from the execution target selection unit 171. One or more execution results specified by the received one or more component identification information are read from the execution result storage unit 162. A response message including one or more read execution results (hereinafter referred to as “service response message”) is generated. The generated service response message is transmitted to the transmission destination specified by the received node identification information.

  When receiving a stop notification from the execution target selection unit 171, the service response transmission unit 174 generates a message notifying that the service is stopped as a service response message until receiving a restart notification from the execution target selection unit 171. To do. Send the generated service response message.

<Execution Request Transmitter 175>
The execution request transmission unit 175 receives node identification information and component identification information from the execution target selection unit 171. A request message including the received component identification information (hereinafter referred to as “execution request message”) is generated. The generated execution request message is transmitted to the transmission destination specified by the received node identification information.

<Execution response receiver 176>
The execution response receiving unit 176 receives a response message to the execution request message transmitted from the execution request transmitting unit 175 (hereinafter referred to as “execution response message”). Analyze the received execution response message. The execution result information for the executed component is extracted from the analyzed execution response message. The extracted execution result information is passed to the execution result writing unit 172.

  When the execution response message is an error message, the execution response reception unit 176 passes the error notification and component identification information for identifying the identified component to the execution target selection unit 171.

<Search processing unit 180>
The search processing unit 180 includes a component information registration unit 181, a search request transmission unit 182, a search response reception unit 183, and the like.

<Component information registration unit 181>
The component information registration unit 181 receives component information from the search response reception unit 183. The received component information is registered in the execution target component database 161. The search completion notification is passed to the execution target selection unit 171.

  As shown in FIG. 21, here, as an example, it is assumed that component information of the component c4 is registered in the execution target component database 161 before searching for an alternative component of the component c4 (T161). . In this case, when the component c2 is selected as an alternative component of the component c4, the component information registration unit 181 receives the component information of the component c2. Instead of the component information of the component c4, the received component information of the component c2 is registered in the execution target component database 161 (T162).

<Search request transmission unit 182>
The search request transmission unit 182 receives node identification information and component identification information from the execution target selection unit 171. A request message including the received component identification information (hereinafter referred to as “search request message”) is generated. The generated search request message is transmitted to the transmission destination specified by the received node identification information.

<Search response receiving unit 183>
The search response receiving unit 183 receives a response message to the search request message transmitted from the search request transmitting unit 182 (hereinafter referred to as “search response message”). Analyze the received search response message. Component information related to the alternative component is extracted from the analyzed search response message. The extracted component information is passed to the component information registration unit 181.

<Operation>
Next, processing when searching for an alternative component in the Web service system according to the present embodiment will be described.

<When using>
FIG. 22 is a diagram showing a sequence when the Web service is used in the Web service system according to the present embodiment. As shown in FIG. 22, the user terminal 107 transmits a service request regarding the use of the Web service provided by the application server 106 to the application server 106 (S121).

  Here, it is assumed that the component b1, the component c4, and the component e4 are used for the Web service provided by the application server 106. In this case, when receiving the service request from the user terminal 107, the application server 106 individually executes the following processes (1) to (3).

  (1) The application server 106 transmits an execution request message for requesting execution of the component b1 to the node 101 where the component b1 is arranged (S122). In response to this, when the node 101 receives an execution request message requesting execution of the component b1 from the application server 106, the node 101 executes the component b1 and transmits an execution response message including the execution result of the component b1 to the application server 106. (S125). The application server 106 receives an execution response message including the execution result of the component b1 from the node 101.

  (2) The application server 106 transmits an execution request message for requesting execution of the component c4 to the node 104 where the component c4 is arranged (S123). On the other hand, when the node 104 receives an execution request message requesting execution of the component c4 from the application server 106, the node 104 executes the component c4 and transmits an execution response message including the execution result of the component c4 to the application server 106. (S126). The application server 106 receives an execution response message including the execution result of the component c4 from the node 104.

  (3) The application server 106 transmits an execution request message for requesting execution of the component e3 to the node 103 where the component e3 is arranged (S124). On the other hand, when the node 103 receives an execution request message requesting execution of the component e3 from the application server 106, the node 103 executes the component e3 and transmits an execution response message including the execution result of the component e3 to the application server 106. (S127). The application server 106 receives an execution response message including the execution result of the component e3 from the node 103.

  Then, the application server 106 transmits a service response message generated using the execution result of the component b1, the execution result of the component c4, and the execution result of the component e3 to the user terminal 107 (S128).

  The user terminal 107 receives the service response message transmitted from the application server 106.

<When searching>
FIG. 23 is a diagram showing a sequence when searching for an alternative component in the Web service system according to the present embodiment. As shown in FIG. 23, the application server 106 stops the service when any one or more components used in the Web service becomes unavailable.

  Here, it is assumed that the component c4 cannot be used. In this case, the application server 106 transmits a search request message for requesting a search for an alternative component of the component c4 to the node 105 associated with the component c4 (S131). On the other hand, when the node 105 receives the search request message from the application server 106, the component c4 is selected from the component c5 having the same function as the component c4 and the component c2 and the component c3 linked in the overlay network. Select the best component as an alternative component. At this time, the following processes (1) and (2) are executed individually.

  (1) The node 105 transmits an inquiry request message for inquiring the utility information of the component c2 to the node 102 in which the component c2 having the same function as the component c4 is arranged (S132). In response to this, when receiving the inquiry request message for inquiring the utility information of the component c2, the node 102 transmits an acquisition response message including the utility information of the component c2 to the node 105 (S134). The node 105 receives the acquisition response message including the utility information of the component c2.

  (2) The node 105 transmits an inquiry request message for inquiring the utility information of the component c3 to the node 103 where the component c3 having the same function as the component c4 is arranged (S133). On the other hand, when the node 103 receives the inquiry request message for inquiring the utility information of the component c3, the node 103 transmits an acquisition response message including the utility information of the component c3 to the node 105 (S135). The node 105 receives the acquisition response message including the utility information of the component c3.

  Then, the node 105 selects the component c2 having a higher utility value as the alternative component of the component c4 from the components c2 and c3 based on the utility information of the component c2 and the utility information of the component c3. A search response message including the identification information of the component c2 is transmitted (S136).

  When the application server 106 receives the search response message including the identification information of the component c2 from the node 105, the application server 106 resumes the service using the component c2 instead of the component c4.

<When resuming>
FIG. 24 is a diagram showing a sequence when the Web service is restarted in the Web service system according to the present embodiment. As shown in FIG. 24, the user terminal 107 transmits a service request regarding the use of the Web service provided by the application server 106 to the application server 106 (S141).

  Here, it is assumed that the component b1 and the component e4 are used in the Web service provided by the application server 106. Further, it is assumed that the component c2 is used instead of the component c4. In this case, when receiving the service request from the user terminal 107, the application server 106 individually executes the following processes (1) to (3).

  (1) The application server 106 transmits an execution request message for requesting execution of the component b1 to the node 101 where the component b1 is arranged (S142). In response to this, when the node 101 receives an execution request message requesting execution of the component b1 from the application server 106, the node 101 executes the component b1 and transmits an execution response message including the execution result of the component b1 to the application server 106. (S145). The application server 106 receives an execution response message including the execution result of the component b1 from the node 101.

  (2) The application server 106 transmits an execution request message for requesting execution of the component c2 to the node 102 where the component c2 is arranged (S143). On the other hand, when the node 102 receives an execution request message requesting execution of the component c2 from the application server 106, the node 102 executes the component c2 and transmits an execution response message including the execution result of the component c2 to the application server 106. (S146). The application server 106 receives an execution response message including the execution result of the component c2 from the node 102.

  (3) The application server 106 transmits an execution request message for requesting execution of the component e3 to the node 103 where the component e3 is arranged (S144). On the other hand, when the node 103 receives an execution request message requesting execution of the component e3 from the application server 106, the node 103 executes the component e3 and transmits an execution response message including the execution result of the component e3 to the application server 106. (S147). The application server 106 receives an execution response message including the execution result of the component e3 from the node 103.

  Then, the application server 106 transmits a service response message generated using the execution result of the component b1, the execution result of the component c2, and the execution result of the component e3 to the user terminal 107 (S148).

  The user terminal 107 receives the service response message transmitted from the application server 106.

<Summary>
As described above, according to the Web service system 100 according to the present embodiment, even if one of one or more components used in the Web service cannot be used, a replacement component can be found immediately, and early. The web service can be resumed. At this time, the search proxy node searches for a component constituting the overlay network, and selects a component having high utility based on the utility information of the component having the same function as the component that cannot be used. As a result, it is possible to efficiently select a component that replaces a component that cannot be used immediately. Further, by notifying the application server 106 of the component information of the selected component, the application server 106 can identify the optimum component instead of the component that can no longer be used, and can restart the Web service at an early stage. it can.

(Embodiment 2)
Embodiment 2 according to the present invention will be described below with reference to the drawings.

<Overview>
The Web service system in the present embodiment has the characteristics shown in (c) below.

  (C) Each of the plurality of nodes is a second component having the same function as the first component among the first component arranged in the own node and one or more components arranged in the other nodes. An overlay network construction processing unit is provided that forms a logical network with a component and registers link information indicating a link of the formed logical network in a link database.

  Based on the above points, the Web service system in the present embodiment will be described. Note that the same components as those of the Web service system according to the first embodiment are denoted by the same reference numerals and description thereof is omitted.

<Configuration>
First, the configuration of the Web service system in the present embodiment will be described.

  The Web service system according to the present embodiment is different from the Web service system 100 according to the first embodiment in that the nodes 101 to 105 have a function for constructing a new overlay network. Here, only the node in the present embodiment will be described by taking as an example a case where the node 101 has a function of newly constructing an overlay network.

  Next, the configuration of the node in the present embodiment will be described.

  FIG. 25 is a diagram illustrating a configuration of a node in the present embodiment. As shown in FIG. 25, the node 201 is different from the node 101 in the first embodiment in that an overlay network construction processing unit 210 is newly provided.

  The overlay network construction processing unit 210 may be realized on a computer system using a distributed object technology such as CORBA, DCOM, or SOAP, or may be realized on a computer system using a distributed object technology other than these. It may be realized as follows. Here, as an example, it is assumed that it is realized on a computer system using SOAP.

<Overlay network construction processor>
FIG. 26 is a diagram showing a detailed configuration of a node in the present embodiment. As shown in FIG. 26, the overlay network construction processing unit 210 includes an MLN holding unit 211, an LLN holding unit 212, a link number information storage unit 213, a link number rank storage unit 214, a network distance rank storage unit 215, and an expected value calculation. Unit 216, overlay network construction unit 217, and the like.

  Furthermore, the overlay network construction processing unit 210 includes a deletion target selection unit 231, a deletion target link source selection unit 232, a link source component selection unit 233, a deletion target link destination selection unit 234, and the like.

  Furthermore, the overlay network construction processing unit 210 includes a link source deletion unit 241, a link destination deletion unit 242, a link deletion determination unit 243, a link deletion request transmission unit 244, a link deletion response reception unit 245, a link deletion request reception unit 246, a link A deletion response transmission unit 247 and the like are provided.

  Furthermore, the overlay network construction processing unit 210 includes an addition target selection unit 251, an addition target link destination selection unit 252, a link destination node selection unit 253, and the like.

  Furthermore, the overlay network construction processing unit 210 includes a link source addition unit 261, a link destination addition unit 262, a link addition determination unit 263, a link addition request transmission unit 264, a link addition response reception unit 265, a link addition request reception unit 266, and a link. An additional response transmission unit 267 and the like are provided.

  Furthermore, the overlay network construction processing unit 210 includes a link number information reading unit 271, a link number information writing unit 272, a link number information inquiry request transmission unit 273, a link number information inquiry response reception unit 274, and a link number information inquiry request reception unit. 275, a link number information inquiry response transmission unit 276, and the like.

<MLN holding unit 211>
The MLN holding unit 211 holds an MLN that is the maximum value of the total number of links of one or more components arranged in its own node.

<LLN holding unit 212>
The LLN holding unit 212 holds an LLN that is the minimum value of the number of links of one or more components arranged in the own node.

<Link Number Information Storage Unit 213>
The link number information storage unit 213 stores link number information individually for each link destination.

<Link number rank storage unit 214>
The link number rank storage unit 214 stores link number rank information that specifies the rank when one or more components arranged in the own node are arranged in descending order of the number of links.

<Network Distance Order Storage Unit 215>
The network distance rank storage unit 215 stores network distance rank information in which the rank is specified when one or more other nodes constituting the network are arranged in order from the own node.

<Expected value calculation unit 216>
The expected value calculation unit 216 calculates an expected value of the average number of links of components arranged in the own node.

<Overlay network construction unit 217>
When a new component is placed at a node, the overlay network construction unit 217 sets the network distance rank information to a value indicating “closest to the first”. The link number rank information is set to a value indicating “most frequently”. The following link formation process is executed.

  Based on the MLN held by the MLN holding unit 211 and one or more link information registered in the link database 114, the overlay network construction unit 217 determines whether the total number of links of the node is greater than or equal to the MLN. Determine. One of the following processes (1) and (2) is executed according to the determined result.

  (1) When the total number of links of the node is greater than or equal to MLN, the overlay network construction unit 217 passes the component identification information for identifying the component newly arranged in the node to the deletion target selection unit 231. The deletion target link information is received from the deletion target selection unit 231. One of the following processes (1a) and (1b) is executed according to the received deletion target link information.

  (1a) When the received deletion target link information is null information, that is, when there is no link to be deleted, the link formation process is terminated.

  (1b) When the received deletion target link information is not null information, that is, when there is a link to be deleted, the overlay network construction unit 217 passes the received deletion target link information to the link source deletion unit 241. When a deletion failure notification is received from the link source deletion unit 241, the link formation process is terminated. When a deletion success notification is received from the link source deletion unit 241, the link formation process is executed again.

  (2) When the total number of links of the node is not equal to or greater than MLN, the overlay network construction unit 217 passes component identification information for identifying the component newly arranged at the node to the addition target selection unit 251. The addition target link information is received from the addition target selection unit 251. One of the following processes (2a) and (2b) is executed according to the received addition target link information.

  (2a) When the received addition target link information is null information, that is, when there is no link to be added, the link formation process is terminated.

  (2b) When the received addition target link information is not null information, that is, when there is a link to be added, the overlay network construction unit 217 passes the received addition target link information to the link source adding unit 261. When an addition failure notification is received from the link source adding unit 261, the link forming process is terminated. When an addition success notification is received from the link source adding unit 261, the link formation process is executed again.

<Delete Target Selection Unit 231>
The deletion target selection unit 231 receives component identification information from the overlay network construction unit 217. All the link information in which the received component identification information is included in the link source information is selected from one or more link information registered in the link database 114. Based on the selected one or more link information and the expected value calculated by the expected value calculation unit 216, it is determined whether or not the number of links of the component specified by the received component identification information is greater than or equal to the expected value. . As a result of the determination, if the number of links is greater than or equal to the expected value, null information is passed to the overlay network construction unit 217 as deletion target link information. On the other hand, if the expected value is not exceeded, the following deletion target link selection process is executed.

  The deletion target selection unit 231 passes the node identification information for identifying the own node to the deletion target link source selection unit 232. The link source information is received from the deletion target link source selection unit 232. The received component identification information is passed to the deletion target link destination selection unit 234. Link destination information is received from the deletion target link destination selection unit 234. Link information including the received link source information and the received link destination information is generated. Depending on the received link source information, one of the following processes (1) and (2) is executed.

  (1) When the received link source information is null information, that is, when there is no link source, the deletion target selection unit 231 sets the link number rank information to a value indicating “most frequently”. The null information is passed to the overlay network construction unit 217 as deletion target link information.

  (2) If the received link source information is not null information, that is, if there is a link source, the deletion target selection unit 231 sets the link number rank information to a value indicating “next highest”. One of the following processes (2a) and (2b) is executed according to the received link destination information.

  (2a) When the received link destination information is null information, that is, when there is no link destination, the deletion target selection unit 231 executes again from the deletion target link selection process.

  (2b) If the received link destination information is not null information, that is, if there is a link destination, the deletion target selection unit 231 passes the generated link information to the overlay network construction unit 217 as deletion target link information.

  When receiving the component identification information from the link destination adding unit 262, the deletion target selecting unit 231 performs the same process as the above process, and passes the deletion target link information to the link destination adding unit 262.

<Deleted Link Source Selection Unit 232>
The deletion target link source selection unit 232 receives node identification information from the deletion target selection unit 231. The selection notification is passed to the link source component selection unit 233. Component identification information is received from the link source component selection unit 233. When the received component identification information is null information, the null information is passed to the deletion target selection unit 231 as link source information. On the other hand, if the received component identification information is not null information, link source information including the received component identification information and the received node identification information is generated. The generated link source information is passed to the deletion target selection unit 231.

<Link source component selection unit 233>
The link source component selection unit 233 receives the selection notification from the deletion target link source selection unit 232. Based on one or more pieces of link information registered in the link database 114, the number of links of each component at the link source is calculated. Based on the result obtained by the calculation, the component of the rank specified by the link number rank information stored in the link number rank storage unit 214 is specified. The component identification information for identifying the identified component is passed to the deletion target link source selection unit 232.

<Deleted Link Target Selection Unit 234>
The deletion target link destination selection unit 234 receives the component identification information from the deletion target selection unit 231. All the link information in which the received component identification information is included in the link source information is selected from one or more link information registered in the link database 114. The following process (1) is executed for each of the selected one or more link information.

  (1) The deletion target link destination selection unit 234 passes the node identification information and component identification information included in the link destination information of the selected link information to the link number information inquiry request transmission unit 273.

  The deletion target link destination selection unit 234 executes the process (1) for all of the selected one or more link information, and if the link number information is present in the link number information storage unit 213, the number of links Based on the one or more link number information stored in the information storage unit 213, the link destination having the largest number of links is selected from one or more link destinations specified by the selected one or more link information. The link destination information for specifying the selected link destination is passed to the deletion target selection unit 231.

  On the other hand, when the link number information does not exist in the link number information storage unit 213, the deletion target link destination selection unit 234 passes null information as link destination information to the deletion target selection unit 231.

<Link source deletion unit 241>
The link source deletion unit 241 receives the deletion target link information from the overlay network construction unit 217. The node identification information included in the link destination information of the received deletion target link information and the received deletion target link information are passed to the link deletion request transmission unit 244. The deletion confirmation link information is received from the link deletion response receiving unit 245. Depending on the received deletion confirmation link information, one of the following processes (1) and (2) is executed.

  (1) When the received deletion confirmation link information is null information, that is, when there is no link to be deleted, the link source deletion unit 241 passes a deletion failure notification to the overlay network construction unit 217.

  (2) If the received deletion confirmation link information is not null information, that is, if there is a link to be deleted, the link source deletion unit 241 registers the received deletion confirmation link information in the link database 114. Delete one or more existing link information. A deletion success notification is passed to the overlay network construction unit 217.

  When the link source deletion unit 241 receives the deletion target link information from the link destination addition unit 262, the link source deletion unit 241 executes the same processing as the above processing, and adds either the deletion failure notification or the deletion success notification to the link destination. To the unit 262.

<Link destination deletion unit 242>
The link destination deletion unit 242 receives the deletion target link information from the link deletion request reception unit 246. The received deletion target link information is passed to the link deletion confirmation unit 243. Deletion confirmation link information is received from the link deletion confirmation unit 243. When the received deletion confirmation link information is null information, the received deletion confirmation link information is passed to the link deletion response transmission unit 247. On the other hand, if the received deletion confirmation link information is not null information, one of one or more components arranged in the link destination node based on the component identification information included in the link destination information of the received deletion confirmation link information. From this, a link destination component having the same function as the link source component is specified. Based on the LLN held in the LLN holding unit 212 and one or more link information registered in the link database 114, it is determined whether or not the number of links of the identified link destination component is LLN or less. . One of the following processes (1) and (2) is executed according to the determined result.

  (1) When the number of links of the identified link destination component is LLN or less, the link destination deletion unit 242 passes null information to the link deletion response transmission unit 247 as deletion confirmed link information.

  (2) If the number of links of the identified link destination component is not less than or equal to LLN, the link destination deletion unit 242 deletes the received deletion confirmation link information from one or more link information registered in the link database 114. . The received deletion confirmation link information is passed to the link deletion response transmission unit 247.

<Link deletion confirmation unit 243>
The link deletion confirmation unit 243 receives the deletion target link information from the link destination deletion unit 242. Component information of a component having the same function as the component specified by the component identification information included in the link source information of the received link information to be deleted is selected from one or more component information registered in the component database 112 To do. One of the following processes (1) and (2) is executed according to the selected component information.

  (1) If the selected component information is null information, the link deletion confirmation unit 243 passes the null information to the link destination deletion unit 242 as deletion confirmation link information.

  (2) If the selected component information is not null information, the link deletion confirmation unit 243 generates link destination information including the component identification information and the node identification information (execution node) in the selected component information. Link information including link source information included in the received deletion target link information and the generated link destination information is generated. The generated link information is transferred to the link destination deleting unit 242 as deletion confirmed link information.

<Link deletion request transmission unit 244>
The link deletion request transmission unit 244 receives node identification information and deletion target link information from the link source deletion unit 241. A request message including the received deletion target link information (hereinafter referred to as “link deletion request message”) is generated. The generated link deletion request message is transmitted to the transmission destination specified by the received node identification information.

<Link deletion response receiver 245>
The link deletion response receiving unit 245 receives a response message to the link deletion request message transmitted from the link deletion request transmitting unit 244 (hereinafter referred to as “link deletion response message”). Analyze the received link deletion response message. Delete confirmed link information is extracted from the analyzed link deletion response message. The extracted deletion confirmation link information is passed to the link source deletion unit 241.

<Link deletion request receiving unit 246>
The link deletion request receiving unit 246 receives a link deletion request message from a transmission source such as another node. The node identification information for identifying the transmission source of the received link deletion request message is passed to the link deletion response transmission unit 247. Analyzes the received link deletion request message. The deletion target link information is extracted from the analyzed link deletion request message. The extracted deletion target link information is passed to the link destination deletion unit 242.

<Link deletion response transmission unit 247>
The link deletion response transmission unit 247 receives node identification information from the link deletion request reception unit 246. Deletion confirmation link information is received from the link destination deletion unit 242. A link deletion response message including the received deletion confirmation link information is generated. The generated link deletion response message is transmitted to the transmission destination specified by the received node identification information.

<Additional selection section 251>
The addition target selection unit 251 receives component identification information from the overlay network construction unit 217. Link source information including the received component identification information and node identification information for identifying the own node is generated. The received component identification information is passed to the addition target link destination selection unit 252. Link destination information is received from the addition target link destination selection unit 252. Link information including the generated link source information and the received link destination information is generated. Depending on the received link destination information, one of the following processes (1) and (2) is executed.

  (1) When the received link destination information is null information, that is, when there is no link destination, the addition target selection unit 251 sets the network distance rank information to a value indicating “closest to the first”. Null information is passed to the overlay network construction unit 217 as additional target link information.

  (2) When the received link destination information is not null information, that is, when a link destination exists, the addition target selection unit 251 sets the network distance rank information to a value indicating “closest to next”. The generated link information is passed to the overlay network construction unit 217 as addition target link information.

<Additional link destination selection unit 252>
The addition target link destination selection unit 252 receives component identification information from the addition target selection unit 251. The selection notification is passed to the link destination node selection unit 253. Node identification information is received from the link destination node selection unit 253. When the received node identification information is null information, the null information is passed to the addition target selection unit 251 as link destination information.

  On the other hand, if the received node identification information is not null information, the addition target link destination selection unit 252 sets the received component identification information as the link destination component identification information. Link destination information including the set link destination component identification information and the received node identification information is generated. The generated link destination information is passed to the addition target selection unit 251.

<Link destination node selection unit 253>
The link destination node selection unit 253 receives the selection notification from the addition target link destination selection unit 252. Based on one or more pieces of node information registered in the node database 113, the network distance of each node at the link destination is calculated. Based on the result obtained by the calculation, the node having the rank specified by the network distance rank information stored in the network distance rank storage unit 215 is specified. The node identification information for identifying the identified node is passed to the addition target link destination selection unit 252.

  Here, in order to simplify the story, RTT of the node information is a network distance. If the RTT value is large, the network distance is long. If the RTT value is small, the network distance is short.

<Link source addition unit 261>
The link source addition unit 261 receives the addition target link information from the overlay network construction unit 217. The node identification information included in the link destination information of the received addition target link information and the received addition target link information are passed to the link addition request transmission unit 264. The addition confirmed link information is received from the link addition response receiving unit 265. One of the following processes (1) and (2) is executed according to the received additional confirmation link information.

  (1) When the received addition confirmed link information is null information, that is, when there is no link to be added, the link source adding unit 261 passes an addition failure notification to the overlay network construction unit 217.

  (2) When the received additional confirmed link information is not null information, that is, when there is a link to be added, the link source adding unit 261 registers the received additional confirmed link information in the link database 114. To one or more existing link information. An addition success notification is passed to the overlay network construction unit 217.

<Link destination adding unit 262>
The link destination adding unit 262 receives the addition target link information from the link addition request receiving unit 266. The received addition target link information is passed to the link addition confirmation unit 263. The addition confirmation link information is received from the link addition confirmation unit 263. If the received additional confirmed link information is null information, the received additional confirmed link information is passed to the link addition response transmission unit 267. On the other hand, if the received additional confirmed link information is not null information, the total number of links of the node based on the MLN held in the MLN holding unit 211 and one or more link information registered in the link database 114 Is determined to be greater than or equal to MLN. One of the following processes (1) and (2) is executed according to the determined result.

  (1) When the total number of links of the node is greater than or equal to MLN, the link destination adding unit 262 passes the component identification information included in the link destination information of the received addition confirmed link information to the deletion target selecting unit 231. The deletion target link information is received from the deletion target selection unit 231. One of the following processes (1a) and (1b) is executed according to the received deletion target link information.

  (1a) When the received deletion target link information is null information, the link destination adding unit 262 passes the null information to the link addition response transmitting unit 267 as additional confirmed link information.

  (1b) If the received deletion target link information is not null information, the link destination adding unit 262 passes the received deletion target link information to the link source deleting unit 241. When a deletion failure notification is received from the link source deletion unit 241, null information is passed to the link addition response transmission unit 267 as additional confirmed link information. When a deletion success notification is received from the link source deletion unit 241, the received addition confirmation link information is added to one or more link information registered in the link database 114. The received addition confirmation link information is passed to the link addition response transmission unit 267.

  (2) When the total number of links of the node is not equal to or greater than MLN, the link destination adding unit 262 adds the received addition confirmed link information to one or more link information registered in the link database 114. The received addition confirmation link information is passed to the link addition response transmission unit 267.

<Link addition confirmation unit 263>
The link addition confirmation unit 263 receives the addition target link information from the link destination addition unit 262. The component information of the component having the same function as the component specified by the component identification information included in the link source information of the received addition target link information is selected from one or more component information registered in the component database 112 To do. One of the following processes (1) and (2) is executed according to the selected component information.

  (1) When the selected component information is null information, the link addition confirmation unit 263 passes the null information to the link destination addition unit 262 as addition confirmation link information.

  (2) When the selected component information is not null information, the link addition determination unit 263 generates link destination information including the component identification information and the node identification information (execution node) in the selected component information. Link information including link source information included in the received addition target link information and the generated link destination information is generated. The generated link information is passed to the link destination adding unit 262 as additional confirmed link information.

<Link addition request transmission unit 264>
The link addition request transmission unit 264 receives node identification information and addition target link information from the link source addition unit 261. A request message including the received addition target link information (hereinafter referred to as “link addition request message”) is generated. The generated link addition request message is transmitted to the transmission destination specified by the node identification information included in the link destination information of the received addition target link information.

<Link addition response receiving unit 265>
The link addition response receiving unit 265 receives a response message to the link addition request message transmitted from the link addition request transmitting unit 264 (hereinafter referred to as “link addition response message”). Analyze the received link addition response message. Add confirmed link information is extracted from the analyzed link addition response message. The extracted addition confirmed link information is passed to the link source adding unit 261.

<Link addition request receiving unit 266>
The link addition request receiving unit 266 receives a link addition request message from a transmission source such as another node. The node identification information for identifying the transmission source of the received link addition request message is passed to the link addition response transmission unit 267. Analyze the received link addition request message. Add target link information from the analyzed link addition request message. The extracted addition target link information is passed to the link destination adding unit 262.

<Link addition response transmission unit 267>
The link addition response transmission unit 267 receives node identification information from the link addition request reception unit 266. The addition confirmed link information is received from the link destination adding unit 262. A link addition response message including the received additional confirmed link information is generated. The generated link addition response message is transmitted to the transmission destination specified by the received node identification information.

<Link Number Information Reading Unit 271>
The link number information reading unit 271 receives component identification information from the link number information inquiry request receiving unit 275. Based on one or more component information registered in the component database 112, a component having the same function as the component specified by the received component identification information is selected from the one or more components arranged in the own node. Identify. Based on one or more pieces of link information registered in the link database 114, the number of links of the specified component is calculated. Link number information including component identification information for identifying the calculated component, node identification information for identifying the own node, and the identified number of links is generated. The generated link number information is passed to the link number information inquiry response transmission unit 276.

<Link number information writing unit 272>
The link number information writing unit 272 receives the link number information from the link number information inquiry response receiving unit 274. The received link number information is written into the link number information storage unit 213.

<Link Number Information Inquiry Request Transmitter 273>
The link number information inquiry request transmission unit 273 receives node identification information and component identification information from the deletion target link destination selection unit 234. A request message including the received component identification information (hereinafter referred to as “link number inquiry request message”) is generated. The generated link number inquiry request message is transmitted to the transmission destination specified by the received node identification information.

<Link Number Information Query Response Reception Unit 274>
The link number information inquiry response reception unit 274 receives a response message to the link number inquiry request message transmitted from the link number information inquiry request transmission unit 273 (hereinafter referred to as “link number inquiry response message”). Analyze the received link count inquiry response message. Link number information is extracted from the analyzed link number inquiry response message. The extracted link number information is passed to the link number information writing unit 272.

<Link Number Information Inquiry Request Receiving Unit 275>
The link number information inquiry request receiving unit 275 receives a link number inquiry request message from a transmission source such as another node. The node identification information for identifying the transmission source of the received link number inquiry request message is passed to the link number information inquiry response transmission unit 276. Analyzes the received link count query request message. Component identification information is extracted from the analyzed link number inquiry request message. The extracted component identification information is passed to the link number information reading unit 271.

<Link Number Information Query Response Transmitter 276>
The link number information inquiry response transmission unit 276 receives node identification information from the link number information inquiry request reception unit 275. Link number information is received from the link number information reading unit 271. A link number inquiry response message including the received link number information is generated. The generated link number inquiry response message is transmitted to the transmission destination specified by the received node identification information.

<Operation>
Next, the operation of the Web service system in this embodiment will be described.

<Overlay network construction processing>
FIG. 27 is a diagram showing overlay network construction processing in the present embodiment. As shown in FIG. 27, when a new component is placed at a node (S201: Yes), the overlay network construction unit 217 sets the network distance rank information to a value indicating “closest to 1” (S202). ). The link number order information is set to a value indicating "most" (S203). The following link formation processing (S204) to (S212) is executed.

  Based on the MLN held by the MLN holding unit 211 and one or more link information registered in the link database 114, the overlay network construction unit 217 determines whether the total number of links of the node is greater than or equal to the MLN. Determine. Depending on the result of the determination, one of the following processes (1) and (2) is executed (S204).

  (1) When the total number of links of the node is greater than or equal to MLN (S204: Yes), the overlay network construction unit 217 sends the component identification information for identifying the component newly arranged in the node to the deletion target selection unit 231. The deletion target selection process is executed (S205). The deletion target link information is received from the deletion target selection unit 231. Depending on the received deletion target link information, one of the following processes (1a) and (1b) is executed (S206).

  (1a) When the received deletion target link information is null information, that is, when there is no deletion target link (S206: Yes), the link formation process is terminated.

  (1b) If the received deletion target link information is not null information, that is, if there is a link to be deleted (S206: No), the overlay network construction unit 217 deletes the received deletion target link information as the link source deletion Then, the link source deletion process is executed (S207). When a deletion failure notification is received from the link source deletion unit 241 (S208: Yes), the link formation process is terminated. When a deletion success notification is received from the link source deletion unit 241 (S208: No), the link formation process is executed again (S204).

  (2) When the total link number of the node is not equal to or greater than MLN (S204: No), the overlay network construction unit 217 passes the component identification information for identifying the component newly arranged in the node to the addition target selection unit 251 Then, the addition target selection process is executed (S209). The addition target link information is received from the addition target selection unit 251. One of the following processes (2a) and (2b) is executed according to the received addition target link information (S210).

  (2a) When the received addition target link information is null information, that is, when there is no link to be added (S210: Yes), the link formation process is terminated.

  (2b) When the received addition target link information is not null information, that is, when there is a link to be added (S210: No), the overlay network construction unit 217 adds the received addition target link information to the link source. Then, the link source addition process is executed (S211). When an addition failure notification is received from the link source adding unit 261 (S212: Yes), the link formation process is terminated. When an addition success notification is received from the link source adding unit 261 (S212: No), the link formation process is executed again (S204).

<Selection process for deletion>
Next, the deletion target selection process executed in the link source node in the present embodiment will be described.

  FIG. 28 is a diagram showing a deletion target selection process executed at the link source node in the present embodiment. As illustrated in FIG. 28, the deletion target selection unit 231 receives component identification information from the overlay network construction unit 217. All the link information in which the received component identification information is included in the link source information is selected from one or more link information registered in the link database 114 (S221). Based on the selected one or more link information and the expected value calculated by the expected value calculation unit 216, it is determined whether or not the number of links of the component specified by the received component identification information is greater than or equal to the expected value. (S222). As a result of the determination, if the number of links is greater than or equal to the expected value (S222: Yes), the null information is passed to the overlay network construction unit 217 as deletion target link information (S223). On the other hand, if the expected value is not exceeded (S222: No), the following deletion target link selection processing (S224) to (S231) is executed.

  The deletion target selection unit 231 passes the node identification information for identifying the own node to the deletion target link source selection unit 232, and selects a link source to be deleted (S224). The link source information is received from the deletion target link source selection unit 232. The received component identification information is transferred to the deletion target link destination selection unit 234, and a link destination to be deleted is selected (S225). Link destination information is received from the deletion target link destination selection unit 234. Link information including the received link source information and the received link destination information is generated (S226). Depending on the received link source information, one of the following processes (1) and (2) is executed (S227).

  (1) When the received link source information is null information, that is, when the link source does not exist (S227: Yes), the deletion target selection unit 231 indicates that the number-of-links rank information is “the most”. A value is set (S228). Null information is transferred to the overlay network construction unit 217 as deletion target link information (S223).

  (2) When the received link source information is not null information, that is, when there is a link source (S227: No), the deletion target selection unit 231 indicates that the link number rank information is “next highest”. (S229). Depending on the received link destination information, one of the following processes (2a) and (2b) is executed (S230).

  (2a) When the received link destination information is null information, that is, when there is no link destination (S230: Yes), the deletion target selection unit 231 executes again from the deletion target link selection process (S224). ).

  (2b) When the received link destination information is not null information, that is, when there is a link destination (S230: No), the deletion target selection unit 231 uses the generated link information as the deletion target link information as an overlay network construction unit. It is passed to 217 (S231).

<Link source deletion processing>
Next, the link source deletion process executed in the link source node in the present embodiment will be described.

  FIG. 29 is a diagram showing link source deletion processing executed at the link source node in the present embodiment. As illustrated in FIG. 29, the link source deletion unit 241 receives deletion target link information from the overlay network construction unit 217. The received deletion target link information is passed to the link deletion request transmission unit 244, and the link destination deletion process is executed by the link destination node to be deleted (S241). The deletion confirmation link information is received from the link deletion response receiving unit 245. Depending on the received deletion confirmation link information, one of the following processes (1) and (2) is executed (S242).

  (1) When the received deletion confirmation link information is null information, that is, when there is no link to be deleted (S242: Yes), the link source deletion unit 241 sends a deletion failure notification to the overlay network construction unit 217. (S243).

  (2) When the received deletion confirmation link information is not null information, that is, when there is a link to be deleted (S242: No), the link source deletion unit 241 stores the received deletion confirmation link information in the link database. It deletes from one or more link information registered into 114 (S244). A deletion success notification is passed to the overlay network construction unit 217 (S245).

<Link destination deletion processing>
Next, the link destination deletion process executed at the link destination node in the present embodiment will be described.

  FIG. 30 is a diagram showing link destination deletion processing executed at the link destination node in the present embodiment. As illustrated in FIG. 30, the link destination deletion unit 242 receives the deletion target link information from the link deletion request reception unit 246. The received deletion target link information is passed to the link deletion determination unit 243, and the link to be deleted is determined (S251). Deletion confirmation link information is received from the link deletion confirmation unit 243. When the received deletion confirmation link information is null information (S252: Yes), the received deletion confirmation link information is passed to the link deletion response transmission unit 247 (S253). On the other hand, if the received deletion confirmation link information is not null information (S252: No), 1 is arranged in the link destination node based on the component identification information included in the link destination information of the received deletion confirmation link information. Among the above components, a link destination component having the same function as the link source component is specified (S254). Based on the LLN held in the LLN holding unit 212 and one or more link information registered in the link database 114, it is determined whether or not the number of links of the identified link destination component is LLN or less. . One of the following processes (1) and (2) is executed according to the determined result (S255).

  (1) When the number of links of the identified link destination component is LLN or less (S255: Yes), the link destination deletion unit 242 passes the null information to the link deletion response transmission unit 247 as deletion confirmed link information (S253). ).

  (2) When the number of links of the identified link destination component is not less than or equal to LLN (S255: No), the link destination deletion unit 242 uses the received deletion confirmation link information as one or more registered in the link database 114. It deletes from link information (S256). The received deletion confirmation link information is transferred to the link deletion response transmission unit 247 (S257).

<Additional selection process>
Next, an addition target selection process executed at the link source node in the present embodiment will be described.

  FIG. 31 is a diagram showing an addition target selection process executed by the link source node in the present embodiment. As illustrated in FIG. 31, the addition target selection unit 251 receives component identification information from the overlay network construction unit 217. Link source information including the received component identification information and node identification information for identifying the own node is generated, and a link source to be added is set (S261). The received component identification information is passed to the addition target link destination selection unit 252 to select a link target to be added (S262). Link destination information is received from the addition target link destination selection unit 252. Link information including the generated link source information and the received link destination information is generated (S263). Depending on the received link destination information, one of the following processes (1) and (2) is executed (S264).

  (1) When the received link destination information is null information, that is, when there is no link destination (S264: Yes), the addition target selection unit 251 indicates that the network distance rank information is “closest to the first”. A value is set (S265). Null information is passed to the overlay network construction unit 217 as additional target link information (S266).

  (2) When the received link destination information is not null information, that is, when a link destination exists (S264: No), the addition target selection unit 251 indicates that the network distance rank information is “closest next”. (S267). The generated link information is transferred as addition target link information to the overlay network construction unit 217 (S268).

<Link source addition processing>
Next, the link source addition process executed at the link source node in the present embodiment will be described.

  FIG. 32 is a diagram showing link source addition processing executed at the link source node in the present embodiment. As illustrated in FIG. 32, the link source adding unit 261 receives the addition target link information from the overlay network construction unit 217. The received addition target link information is passed to the link addition request transmission unit 264, and the link destination node to be added is caused to execute link destination addition processing (S281). The addition confirmed link information is received from the link addition response receiving unit 265. One of the following processes (1) and (2) is executed according to the received additional confirmation link information (S282).

  (1) When the received addition confirmation link information is null information, that is, when there is no link to be added (S282: Yes), an addition failure notification is passed to the overlay network construction unit 217 (S283).

  (2) When the received additional confirmed link information is not null information, that is, when there is a link to be added (S282: No), the received additional confirmed link information is registered in the link database 114 1 It adds to the above link information (S284). The addition success notification is passed to the overlay network construction unit 217 (S285).

<Link destination addition processing>
Next, link destination addition processing executed at the link destination node in the present embodiment will be described.

  FIG. 33 is a diagram showing link destination addition processing executed at the link destination node in the present embodiment. As illustrated in FIG. 33, the link destination adding unit 262 receives the addition target link information from the link addition request receiving unit 266. The received addition target link information is passed to the link addition confirmation unit 263, and the link to be added is confirmed (S291). The addition confirmation link information is received from the link addition confirmation unit 263. When the received additional confirmed link information is null information, that is, when there is no link to be added (S292: Yes), the null information is passed to the link addition response transmission unit 267 as additional confirmed link information (S293). . On the other hand, when the received addition confirmed link information is not null information, that is, when there is a link to be added (S292: No), the MLN held in the MLN holding unit 211 and the link database 114 are registered. Whether or not the total number of links of the node is greater than or equal to MLN is determined based on the one or more link information. One of the following processes (1) and (2) is executed according to the determined result (S294).

  (1) When the total number of links of the node is greater than or equal to MLN (S294: Yes), the link destination adding unit 262 deletes the component identification information included in the link destination information of the received addition confirmed link information as the deletion target selecting unit 231. The deletion target selection process is executed (S295). The deletion target link information is received from the deletion target selection unit 231. Depending on the received deletion target link information, one of the following processes (1a) and (1b) is executed (S296).

  (1a) When the received deletion target link information is null information, that is, when there is no link to be deleted (S296: Yes), the link destination adding unit 262 links the null information as additional confirmed link information. It passes to the additional response transmission part 267 (S293).

  (1b) When the received deletion target link information is not null information, that is, when there is a link to be deleted (S296: No), the link destination adding unit 262 deletes the received deletion target link information as the link source deletion The link source deletion process is executed (S297). When a deletion failure notification is received from the link source deletion unit 241 (S298: Yes), null information is passed to the link addition response transmission unit 267 as additional confirmed link information (S293). When a deletion success notification is received from the link source deletion unit 241 (S298: No), the received addition confirmation link information is added to one or more link information registered in the link database 114 (S299). The received addition confirmation link information is passed to the link addition response transmission unit 267 (S300).

  (2) When the total number of links of the node is not greater than or equal to MLN (S294: No), the link destination adding unit 262 adds the received addition confirmed link information to one or more link information registered in the link database 114 (S299). The received addition confirmation link information is passed to the link addition response transmission unit 267 (S300).

<Sequence>
Next, the sequence between nodes at the time of construction of the overlay network in the present embodiment will be described. Here, as an example, the interaction between nodes will be described in the case where a component b4 is newly arranged in the node 4 as shown in FIGS.

  In the physical network, it is assumed that node 1, node 2, node 5, and node 3 are arranged in the order closer to node 4.

  34 to 36 are diagrams showing a sequence between nodes at the time of constructing the overlay network in the present embodiment. As shown in FIG. 34, in this case, the node 104 executes the following processes (1) to (4) during the link addition process.

  (1) The node 104 transmits a link addition request message for requesting to form a link between the component b4 and the component b1 to the node 101 where the component b1 is arranged (S311). In contrast, when the node 101 receives the link addition request message from the node 104, the node 101 executes link destination addition processing. As a result of the execution, since the upper limit value of the total number of links set in the node 101 is not exceeded, a link addition response message indicating that a link is formed between the component b4 and the component b1 is transmitted to the node 104 ( S312). The node 104 receives the link addition response message from the node 101.

  (2) The node 104 transmits a link addition request message requesting to form a link between the component b4 and the component b2 to the node 102 where the component b2 is arranged (S313). In contrast, when the node 102 receives the link addition request message from the node 104, the node 102 executes link destination addition processing. As a result of the execution, since the upper limit value of the total number of links set in the node 102 is not exceeded, a link addition response message indicating that a link has been formed between the component b4 and the component b2 is transmitted to the node 104 ( S314). The node 104 receives the link addition response message from the node 102.

  (3) The node 104 transmits a link addition request message for requesting to form a link between the component b4 and the component b5 to the node 105 where the component b5 is arranged (S315). On the other hand, when the node 105 receives the link addition request message from the node 104, the node 105 executes link destination addition processing. As a result of the execution, the upper limit of the total number of links set in the node 105 is exceeded, so an attempt is made to delete another link.

  At this time, as shown in FIG. 35, the node 105 individually executes the following processes (3a) to (3c) during the deletion target selection process.

  (3a) The node 105 transmits a link number information inquiry request message for requesting an inquiry about the number of links of the component b1 to the node 101 where the component b1 is arranged (S321). On the other hand, when the node 101 receives the link number information inquiry request message from the node 105, the node 101 calculates the link number of the component b1, and sends a link number information inquiry response message including the calculated link number (four) to the node 105. (S324). The node 105 receives the link number information inquiry response message from the node 101.

  (3b) The node 105 transmits a link number information inquiry request message for requesting an inquiry about the number of links of the component b2 to the node 102 in which the component b2 is arranged (S322). On the other hand, when the node 102 receives the link number information inquiry request message from the node 105, the node 102 calculates the link number of the component b2, and sends a link number information inquiry response message including the calculated link number (three) to the node 105. (S325). The node 105 receives the link number information inquiry response message from the node 102.

  (3c) The node 105 transmits a link number information inquiry request message for requesting an inquiry about the number of links of the component b3 to the node 103 where the component b3 is arranged (S323). On the other hand, when the node 103 receives the link number information inquiry request message from the node 105, the node 103 calculates the link number of the component b3, and sends a link number information inquiry response message including the calculated link number (two) to the node 105. (S326). The node 105 receives the link number information inquiry response message from the node 103.

  The node 105 stores the number of links of the component b1 (4), the number of links of the component b2 (3), and the number of links of the component b3 (2) in the link number information storage unit 213 for each link destination. The component b1 with the largest number of links is selected as a deletion target.

  Further, as shown in FIG. 36, the node 105 sends a link deletion request message requesting the node 101 where the component b1 is arranged to delete the link formed between the component b5 and the component b1. Transmit (S331). In contrast, when the node 101 receives a link deletion request message from the node 105, the node 101 executes link destination deletion processing. As a result of the execution, the link deletion response message indicating that the link formed between the component b5 and the component b1 has been deleted is transmitted to the node 105 because it does not fall below the lower limit of the number of links set in the node 101. (S332). The node 105 receives the link deletion response message from 101.

  Accordingly, since the node 105 does not exceed the upper limit value of the total number of links set in the node 105, a link addition response message indicating that a link is formed between the component b4 and the component b5 is sent to the node 104. (S316). The node 104 receives the link addition response message from the node 105.

  (4) The node 104 transmits a link addition request message requesting to form a link between the component b1 and the component b3 to the node 103 where the component b3 is arranged (S317). On the other hand, when the node 103 receives the link addition request message from the node 104, the node 103 executes link destination addition processing. As a result of the execution, since the upper limit value of the total number of links set in the node 103 is not exceeded, a link addition response message indicating that a link has been formed between the component b4 and the component b3 is transmitted to the node 104 ( S318). The node 104 receives the link addition response message from the node 103.

<Summary>
As described above, according to the Web service system in the present embodiment, even if components are added or deleted, an overlay network can be dynamically constructed in cooperation with each node. Moreover, since it is based on autonomous distributed processing, there is no single point of failure and it has resistance against a plurality of failures.

(Other)
The node 101 or the node 201 may include a CPU (Central Processing Unit), a RAM (Random Access Memory), a ROM (Read Only Memory), a HDD (Hard Disk Drive), a network adapter, and the like. Further, a program for controlling the node 101 or the node 201 (hereinafter referred to as “node control program”) is installed in the HDD or the like, and the node 101 or the Each function such as the node 201 may be realized.

  Note that the node 101, the node 201, or the like may be an embedded system such as a digital video camera, a digital recorder, a digital television, a game machine, a home server, an IP phone, or a mobile phone.

  In addition, the node 101, the node 201, or the like may be configured by a plurality of computer systems. In this case, message communication may be performed between each function such as the node 101 or the node 201 distributed in a plurality of computer systems based on CORBA, DCOM, SOAP, and the like.

  Further, the node 101, the node 201, or the like may be configured by a Web server and a database server. In this case, at least the component execution processing unit 120, the utility information generation processing unit 130, the alternative component search processing unit 140, the utility information transmission processing unit 150, the overlay network construction processing unit 210, and the like may be realized in the Web server. Further, the component database 112, the node database 113, the link database 114, and the like may be realized in the database server. Further, an ODBC (Open DataBase Connectivity) or a JDBC (Java (registered trademark) DataBase Connectivity) may be used for accessing the database server from the Web server.

  Further, based on TLS (Transport Layer Security) defined in RFC (Request for Comments) 2246, security communication may be performed between the node 101 or the node 201 and each application server. In addition, based on XML encryption (http://www.w3.org/TR/2002/REC-xmlenc-core-20021210/), data is encrypted, and the encrypted data is sent and received for security communication. May be performed.

  The application server 106 may include a CPU, RAM, ROM, HDD, network adapter, and the like. Further, a program for controlling the application server 106 (hereinafter referred to as “application server control program”) is installed in the HDD or the like, and each function of the application server 106 is executed by executing the application server control program. May be realized.

  Note that the application server 106 may be an embedded system such as a digital video camera, a digital recorder, a digital TV, a game machine, a home server, an IP phone, a mobile phone, or the like.

  The application server 106 may be configured by a plurality of computer systems. In this case, message communication may be performed between the functions of the application server 106 distributed in a plurality of computer systems based on CORBA, DCOM, SOAP, and the like.

  Further, the application server 106 may be composed of a Web server and a database server. In this case, at least the service processing unit 170, the search processing unit 180, and the like may be realized on the Web server. Moreover, the execution target component database 161 may be realized in a database server. Furthermore, ODBC or JDBC may be used for accessing the database server from the WEB server.

  Further, security communication may be performed between the application server 106 and each user terminal based on TLS (Transport Layer Security) defined in RFC2246. In addition, based on XML encryption (http://www.w3.org/TR/2002/REC-xmlenc-core-20021210/), data is encrypted, and the encrypted data is sent and received for security communication. May be performed.

  Note that the node control program and the application server control program may be recorded on a recording medium readable by a hardware system such as a computer system or an embedded system. Furthermore, the program may be read and executed by another hardware system via the recording medium. Thereby, each function of the node 101 or the node 201 or each function of the application server 106 can be realized in another hardware system. Here, as a computer system-readable recording medium, an optical recording medium (for example, a CD-ROM), a magnetic recording medium (for example, a hard disk), a magneto-optical recording medium (for example, an MO), or a semiconductor memory. (For example, a memory card).

  Further, the node control program and the application server control program may be held in a hardware system connected to a network such as the Internet or a local area network. Furthermore, it may be downloaded to another hardware system via a network and executed. Thereby, each function of the node 101 or the node 201 and each function of the application server 106 can be realized in another hardware system. Here, as a network, a terrestrial broadcasting network, a satellite broadcasting network, a PLC (Power Line Communication), a mobile telephone network, a wired communication network (for example, IEEE802.3), and a wireless communication network (for example, IEEE802.11). There is.

(Appendix 1)
In the following, supplementary explanations of Web services and distributed Web services will be provided. Web services mainly include elemental technologies such as XML, SOAP (Simple Object Access Protocol), WSDL (Web Services Description Language), and UDDI (Universal Description, Discovery, and Integration). However, since these elemental technologies are widely known, description thereof is omitted here.

<Web service>
A web service is a software system that can be identified by a URI (Uniform Resource Identifier), and whose interface and binding are defined and described in XML (eXtensible Markup Language). It is a software system that can communicate with other software systems by using XML-based messages transmitted by the Internet protocol. It is also one of the important technologies for realizing SOA (Service Oriented Architecture).

  In the Web service, a service provider registers and publishes a service in a service directory. When the service user receives the service, the application server searches the service directory, obtains the location of the service, and connects to the server providing the service. On the other hand, in a service using the conventional Internet, a service user searches for a service using a Web browser or the like, and uses the searched service.

  The SOA is a software architecture that defines a system structure that can be freely linked and used through a common interface without depending on hardware, OS, or language.

  In the SOA, the entire information system is regarded as a collection of services.

  Conventional methods for realizing the SOA include CORBA (Common Object Request Broker Architecture) of OMG (Object Management Group) and DCOM (Distributed Component Object Model) of Microsoft. These methods are difficult to replace and improve because the components are tightly coupled.

  On the other hand, since the architecture of Web services is loosely coupled, even if components cooperate with each other, they are less dependent on each other, and each component can be easily exchanged and improved. High adaptability. Furthermore, an existing service can be used as a component, and a service can be constructed by combining components. For this reason, it is not necessary to construct a service from scratch as in the prior art, and the cost required for constructing the service can be reduced. It is also easier to build a more advanced service than existing Web services.

<Distributed Web Service>
On the other hand, a distributed Web service is not officially defined. Therefore, in this case, in particular, a component having one component per type is a Web service, and a component having a plurality of duplicate components per component is a distributed Web service.

  In the Web service, there are no components having the same function. For this reason, the load concerning a component increases because the provider of the service using the same component increases. The user satisfaction of the service decreases due to a decrease in the processing speed of the service using the component.

  On the other hand, in the distributed Web service, since components are duplicated and distributed, there are a plurality of components having the same function. By providing services using these components, the load on one component can be distributed. Therefore, even when a service using a component having a specific function is provided, a decrease in service processing speed can be suppressed, and a satisfactory service can be provided to service users.

  In the Web service, when a failure occurs in a certain component, a plurality of application servers stop providing services using the component.

  On the other hand, in the distributed Web service, there is a component having the same function as the component in which the failure has occurred. Therefore, it is desirable to restart a service by searching for a component having the same function among a plurality of components arranged in a distributed manner as an alternative component and instantaneously switching the component to the alternative component.

(Appendix 2)
In the following, the explanation of the alternative component search described in the background art will be supplemented.

<Search alternative components>
In the distributed Web service, when a component used in the Web service becomes unusable, it is possible to find an alternative component and continue providing the service using the alternative component. However, the search method is not defined. Therefore, here, a reactive search method and a proactive search method will be described.

<Reactive search method>
The reactive search method is a method of searching for an alternative component after a failure occurs in the component.

  The reactive search method does not hold the position information of the alternative component in advance. For this reason, flooding is used to search for an alternative component when a failure occurs. Flooding is a technique for sequentially copying the same message and transmitting it to the entire network. For this reason, it is possible to surely find a substitute component with high utility. However, since the number of messages increases, the load on the network increases. Therefore, in the reactive search method, messages are not exchanged until a failure occurs, so that no load is applied to the network. However, when a failure occurs, the load on the network increases in a short period of time by exchanging a large number of search messages by flooding. In addition, since the search is started after a failure occurs, it takes time to find an alternative component.

  There are roughly two methods for searching for alternative components by flooding. The first is First-Reply, and the second is Best-Reply.

<First-Reply>
First-Reply is a method in which a component first found among a plurality of components having the same function as a component in which a failure has occurred is used as an alternative component.

  In First-Reply, since the component that is the first alternative candidate discovered is used, the search itself is completed immediately, and the search time can be shortened. The delay before service restart is short. However, even if there is a component on the network that is more effective than the first discovered alternative component, it cannot be selected. In addition, a component with low utility may be used by another application server. In the worst case, an application server that newly uses the component receives the service due to overload by using the component. It may not be possible.

<Best Reply>
Best-Reply is a technique in which a component that has a response within a predetermined time and has high utility among a plurality of components having the same function as the component in which a failure has occurred is used as a substitute component.

  Best-Reply searches for alternative component candidates by flooding, and can select high-utility components from the alternative candidate components that have responded to the application server within a certain period of time. It is a technique. Here, it is assumed that a response from the alternative candidate component is waited for a sufficient period of time so that it can be checked whether there are alternative component candidates in all nodes. Also, if the time for waiting for a response (response waiting time) is too long, another application server uses the component during the response waiting time after the utility information of the alternative candidate component obtained first is returned. It has the disadvantages that it may change and the reliability of information will be reduced. For this reason, it is not suitable to wait for responses from all the alternative candidate components in a large-scale network, and it is difficult to set a response waiting time when it is not known where the alternative candidate components are. However, the search message does not remain on the network at the end of the search by setting the message waiting time as the response waiting time.

<Proactive alternative component search method>
The proactive search method is a method for determining which component to use before a failure occurs. In the proactive search method, the position information and utility information of alternative components are stored in the database in advance. In addition, information is updated by constantly exchanging messages, and a decrease in reliability of information is suppressed.

  In the proactive search method, it is possible to search for an optimal substitute component immediately when a failure occurs. However, since messages are always exchanged, a constant load is applied to the network.

  Note that database management methods include integrated database management in which all data is collectively managed in one database, and distributed database management in which all data is divided and managed in a plurality of distributed databases. .

<Integrated database management>
Integrated database management is a technique for collectively managing information such as the positions and utilities of all components in one database. By using this method, if the database is accessed, the optimum location information of the database can be obtained immediately.

  However, in the integrated database, the amount of information to be managed becomes enormous in a large-scale network. In addition, in order to grasp dynamically changing information in real time, it is necessary to frequently exchange a large number of control messages, and traffic on the network increases. For this reason, it is difficult to always grasp dynamic information in a large-scale network. In addition, since all information is collectively managed in the database, the inquiries to the database are concentrated, which may overload the database and make it impossible to provide information. In such a case, integrated database management makes it impossible to search for alternative components. Furthermore, the application server that has used the component in which the failure has occurred must stop the service.

<Distributed database management>
Distributed database management is a technique for managing information that has been managed in an integrated database by dividing it into a plurality of databases that are distributed in a plurality of locations.

  In distributed database management, not all data is managed. For this reason, the information of the optimal alternative component is not necessarily obtained as compared with the integrated database. However, the vulnerability to the failure of the database existing in the integrated database can be alleviated, and an increase in traffic in the surrounding network can be suppressed. In distributed database management, since each node does not need to hold information on all alternative components, control messages can be reduced compared to integrated database management. However, in distributed database management, the location of the database and the type and number of information managed by the database are important factors. For this reason, it is considered that by appropriately determining these elements, it is possible to reduce the time required for searching and to efficiently search for alternative components.

  The present invention is a Web service system that provides a Web service that uses a plurality of components, and in particular, among the plurality of components that are distributed and arranged in an overlay network as alternative components for the components that are used in the Web service. It can be used as a Web service system that searches from the Internet.

The figure which shows the outline | summary of the Web service system in Embodiment 1 concerning this invention. The figure which shows the outline | summary at the time of utilization in the alternative component search method in Embodiment 1 concerning this invention. The figure which shows the outline | summary at the time of a failure in the alternative component search method in Embodiment 1 concerning this invention The figure which shows the outline | summary at the time of a search in the alternative component search method in Embodiment 1 concerning this invention. The figure which shows the outline | summary at the time of restart in the alternative component search method in Embodiment 1 concerning this invention FIG. 1 is a first diagram showing the procedure of an overlay network construction method according to the first embodiment of the present invention; FIG. 2 is a second diagram showing the procedure of the overlay network construction method according to the first embodiment of the present invention. The figure which shows the outline | summary before construction in the overlay network construction method in Embodiment 1 concerning this invention The figure which shows the outline | summary at the time of addition in the overlay network construction method in Embodiment 1 concerning this invention The figure which shows the outline | summary at the time of deletion in the overlay network construction method in Embodiment 1 concerning this invention The figure which shows the outline | summary after construction | assembly in the overlay network construction method in Embodiment 1 concerning this invention. The figure which shows the structure of the Web service system in Embodiment 1 concerning this invention. The figure which shows the structure of the node in Embodiment 1 concerning this invention. The figure which shows the data structure of the component information managed by the component database in Embodiment 1 concerning this invention. The figure which shows the data structure of the node information managed by the node database in Embodiment 1 concerning this invention. The figure which shows the data structure of the link information managed by the link database in Embodiment 1 concerning this invention. The figure which shows the data structure of the utility information memorize | stored in the own node utility information storage part in Embodiment 1 concerning this invention. The figure which shows the data structure of the utility information memorize | stored in the other node utility information storage part in Embodiment 1 concerning this invention. The figure which shows the detailed structure of the node in Embodiment 1 concerning this invention. The figure which shows the structure of the application server in Embodiment 1 concerning this invention. The figure which shows the data structure of the component information managed by the execution object component database in Embodiment 1 concerning this invention. The figure which shows the sequence at the time of utilization of a web service in the web service system in Embodiment 1 concerning this invention. The figure which shows the sequence at the time of the search of an alternative component in the Web service system in Embodiment 1 concerning this invention. The figure which shows the sequence at the time of resumption of a web service in the web service system in Embodiment 1 concerning this invention. The figure which shows the structure of the node in Embodiment 2 concerning this invention. The figure which shows the detailed structure of the node in Embodiment 2 concerning this invention. The figure which shows the overlay network construction process in Embodiment 2 concerning this invention. The figure which shows the deletion object selection process performed in the node of the link origin in Embodiment 2 concerning this invention The figure which shows the link origin deletion process performed in the node of the link origin in Embodiment 2 concerning this invention The figure which shows the link destination deletion process performed in the link destination node in Embodiment 2 concerning this invention. The figure which shows the addition object selection process performed in the node of the link origin in Embodiment 2 concerning this invention The figure which shows the link source addition process performed in the node of the link source in Embodiment 2 concerning this invention The figure which shows the link destination addition process performed in the link destination node in Embodiment 2 concerning this invention. First diagram showing a sequence between nodes at the time of construction of an overlay network according to the second embodiment of the present invention. 2nd figure which shows the sequence between nodes at the time of construction of the overlay network in Embodiment 2 concerning this invention 3rd figure which shows the sequence between nodes at the time of construction of the overlay network in Embodiment 2 concerning this invention

Explanation of symbols

100 Web service systems 101 to 105, 201 Node 106 Application server 111 Component holding unit 112 Component database 113 Node database 114 Link database 115 Local node utility information storage unit 116 Other node utility information storage unit 120 Component execution processing unit 121 Component execution unit 122 Execution request reception unit 123 Execution response transmission unit 130 Utility information generation processing unit 131 Measurement unit 132 Utility information generation unit 140 Alternative component search processing unit 141 Alternative component search unit 142 Search target selection unit 143 Utility information writing unit 144 Search request reception unit 145 Search response transmission unit 146 Utility information inquiry request transmission unit 147 Utility information inquiry response reception unit 150 Utility information transmission processing unit 151 Utility information reading unit 152 Information query request receiving unit 153 Utility information query response sending unit 161 Execution target component database 162 Execution result storage unit 170 Service processing unit 171 Execution target selection unit 172 Execution result writing unit 173 Service request receiving unit 174 Service response sending unit 175 Execution Request transmission unit 176 Execution response reception unit 180 Search processing unit 181 Component information registration unit 182 Search request transmission unit 183 Search response reception unit 210 Overlay network construction processing unit 211 MLN storage unit 212 LLN storage unit 213 Link number information storage unit 214 Number of links Rank storage unit 215 Network distance rank storage unit 216 Expected value calculation unit 217 Overlay network construction unit 231 Deletion target selection unit 232 Deletion target link source selection unit 233 Link source component selection unit 234 Deletion target phosphorus Destination selection unit 241 link source deletion unit 242 link destination deletion unit 243 link deletion confirmation unit 244 link deletion request transmission unit 245 link deletion response reception unit 246 link deletion request reception unit 247 link deletion response transmission unit 251 addition target selection unit 252 addition Target link destination selection unit 253 Link destination node selection unit 261 Link source addition unit 262 Link destination addition unit 263 Link addition confirmation unit 264 Link addition request transmission unit 265 Link addition response reception unit 266 Link addition request reception unit 267 Link addition response transmission unit 271 Link number information reading unit 272 Link number information writing unit 273 Link number information inquiry request transmission unit 274 Link number information inquiry response reception unit 275 Link number information inquiry request reception unit 276 Link number information inquiry response transmission unit

Claims (3)

A web service system comprising: an application server that provides a web service using a plurality of components; and a plurality of nodes that are connected to the application server via a physical network and in which the plurality of components are distributed. Because
Each of the plurality of nodes is
A link database in which link information indicating a link of a logical network formed between one or more components arranged in the own node and one or more components arranged in another node is registered;
When requested by the application server to search for a component having the same function as a predetermined component used in the Web service, the logical network is formed based on link information registered in the link database. An alternative component search processing unit that searches for one or more components and selects an optimum component instead of the predetermined component. A Web service system, comprising: Each of the plurality of nodes is
Execution processing means for executing a component used in the Web service among one or more components arranged in the own node;
Utility information generation processing means for measuring the processing capability of a component being executed by the execution processing means and generating utility information that is an index of utility obtained by executing the component using a result obtained by measurement. And
In the case where a node that searches for the optimum component among the plurality of nodes is a search proxy node, and a node in which a component having the same function as the predetermined component is arranged is a search target node, the search The substitute component search processing means in the proxy node selects the component having the highest utility as the optimum component based on the utility information generated by the utility information generation processing means in the search target node. The web service system according to claim 1. Each of the plurality of nodes includes a first component arranged in its own node and a second component having the same function as the first component among one or more components arranged in other nodes. The web service system according to claim 1, further comprising: an overlay network construction processing unit that forms a logical network with the link database and registers link information indicating a link of the formed logical network in the link database.

Images