JP5100011B2 - Network resource management apparatus, network resource management system, and network resource management method - Google Patents

Description

  The present invention relates to a network resource management apparatus. The present invention particularly relates to a technique of distributed processing in a network resource management apparatus.

  Conventionally, in a technology that uses resources of a computer connected to a network for distributed processing, there is no sufficient mechanism for uniformly managing and controlling resources provided on the network.

  For example, CORBA (Common Object Request Broker Architecture) is known as a scheme for calling distributed objects on a network. However, CORBA only regulates communication between objects distributed over multiple devices, and it is not expected to use it to dynamically change or add to the configuration of devices with such objects. is there.

  For example, clustering is known as a technique for interconnecting a plurality of computers and operating the computers like a single computer. However, in clustering, the hardware configuration, CPU, OS, etc. need to be homogenized or unified among a plurality of computers, and an unspecified number of computer resources are dynamically added or changed. Use is not expected.

  On the other hand, grid computing, which is a system that connects a plurality of computers on a network to virtually construct one computer and extracts and uses resources according to the required processing capacity from the computer, is known. . However, grid computing is only a technology for solving processing problems such as how to connect resources on the network and distribute processing, and issues such as how to promote resource provision and such resources. It cannot be said that the framework construction problem of how to open and promote the use of is not addressed.

  The present inventor has made the present invention based on the above problems, and an object thereof is to construct a mechanism for efficiently and actively using hardware resources on a network.

  In order to solve the above problems, a network resource management apparatus according to an aspect of the present invention is an apparatus including a processor, and includes a system control unit that controls the operation of the entire apparatus, and a program control unit that controls execution of a program. And a usage situation in which at least a part of hardware resources including a processor in the external apparatus is used via a network for executing program processing, and at least a part of hardware resources including a processor in the apparatus Includes a resource management unit that dynamically manages a usage situation used for processing execution by an external device via a network according to the situation.

  Resource mediation that mediates the distribution of resources by transferring resource information indicating that at least a part of hardware resources including a processor in the external device is available to another external device via a network A part may be further provided.

  Another aspect of the present invention is a network resource management system. This network resource management system includes a plurality of resource management terminals each including a processor and a resource mediation server. Each of the plurality of resource management terminals has at least a part of hardware resources including a system control unit that controls the operation of the entire terminal, a program control unit that controls execution of a program, and a processor in another terminal. At least a part of the usage status used via the network for program processing execution and the hardware resources including the processor in the terminal is used for processing execution by another terminal via the network. A resource management unit that dynamically manages the usage status according to the usage status. The resource mediation server receives resource information indicating that at least a part of hardware resources including a processor in the terminal is available from at least one of the plurality of resource management terminals. Is transferred to another resource management terminal via a network to mediate the distribution of resources.

  The resource mediation server records the relationship between use and usage when a hardware resource in at least one of the plurality of resource management terminals is used for processing execution in another resource management terminal. The record may be sent to a server that handles billing for resource usage fees.

  Yet another embodiment of the present invention is a network resource management method. The method includes generating resource information indicating that at least a part of hardware resources including a processor in the first terminal is available from a first terminal including the processor; , To the server via the network, the server receiving the resource information transfers it to the second terminal, and the second terminal to which the resource information has been transferred is based on the resource information. A step of using at least a part of hardware resources including a processor in the terminal for processing execution in the second terminal, and the hardware resource in the first terminal via the network by the first terminal Recording the usage status used for processing execution by the second terminal, and hardware for the second terminal in the first terminal. A source over a network comprising the steps of: recording usage utilized for processing performed in the second terminal.

  It should be noted that any combination of the above-described constituent elements, or the constituent elements and expressions of the present invention may be mutually replaced between methods, apparatuses, systems, computer programs, recording media storing computer programs, data structures, etc. This is effective as an embodiment of the present invention.

  ADVANTAGE OF THE INVENTION According to this invention, the mechanism for enabling it to utilize the hardware resource on a network efficiently and actively can be constructed | assembled.

  In the present embodiment, when a computer executes a specific process, not only the processor in the computer but also the processor of another device in the network to which the computer is connected has secured a scalable environment. Is realized. For this purpose, the provision of hardware resources from each computer is uniformly managed on the network, and a mechanism for controlling allocation of distributed processing and data transfer is constructed. In addition, a mechanism for actively using each resource provided on the network and an incentive trading system for promoting the distribution management, use and use of such resources will be established. Furthermore, the use and provision of hardware resources will be promoted by providing sufficient versatility and openness to the mechanism of use and provision of resources, and the use and provision of dynamically changing resources will be further processed. Promote resource use.

  FIG. 1 schematically shows the configuration of the resource management network 10. The resource management network 10 includes an Internet 14 and computers such as a LAN 12 and a user terminal 16 connected to the Internet 14. Each user terminal 16 may be connected directly to the Internet 14 or may be connected to the Internet 14 via a LAN 12 such as a home network or a corporate network. Connected to the Internet 14 is a network server 20 that manages and controls resource use and provision between user terminals 16. A local server 18 such as a home server that manages and controls resource use and provision between user terminals 16 connected in the LAN 12 may be connected to the LAN 12. Resource utilization and provision may be executed between the user terminals 16 in the LAN 12, and resource utilization and provision may be performed between the user terminals 16 via the Internet 14. The user terminal 16 may execute resource use and provision with the local server 18. As described above, hardware resources are shared among the plurality of user terminals 16 and the local server 18 via the Internet 14 and the LAN 12, and processing is performed independently by using such resources in a complementary manner and performing distributed processing. This makes it possible to improve the processing capacity of each stage.

  FIG. 2 illustrates the hardware configuration of the user terminal 16. The user terminal 16 includes a microprocessor unit (MPU) 22, a graphics processing unit (GPU) 40, a main memory 42, an auxiliary storage device (HDD) 44, and a network control unit 46, which are connected via a main bus 38. The The network control unit 46 transmits / receives data to / from other user terminals 16 via the LAN 12 or the Internet 14.

  The MPU 22 is an asymmetric multiprocessor unit, and includes one main processing unit (PU) 24 and a plurality of sub processing units (SPU) 30, which are connected via an internal bus 36 of the MPU 22. The PU 24 is a unit that processes the control of the OS and each SPU 30, and includes a processor 26 and a local memory 28. The local memory 28 is, for example, a cache memory. Each SPU 30 is a unit that processes operations centering on product-sum calculations, and includes a processor 32 and a local memory 34, respectively. Programs and data read from the main memory 42 are written in the local memory 34 and executed by the processor 32. Each SPU 30 distributes and executes calculations necessary for image processing, data conversion processing, and the like under the control of the PU 24. The PU 24 manages which SPU 30 executes the calculation, the content of the calculation, and the scheduled execution time. Although the user terminal 16 includes one MPU 22, the local server 18 includes a plurality of MPUs 22 and has a higher processing capability than the user terminal 16. Similarly to the user terminal 16, the local server 18 also uses and provides hardware resources with other terminals and servers.

  FIG. 3 is a functional block diagram illustrating processing and operations performed by the user terminal 16, the local server 18, and the network server 20. Hereinafter, the user terminal 16 will be described as an example. The user terminal 16 includes a system control unit 70, a program control unit 72, a resource management unit 74, and a resource mediation unit 76. Each unit is mainly realized by an OS and various programs loaded from the HDD 44 to the main memory 42, but here, functional blocks realized by their cooperation are depicted. Accordingly, those skilled in the art will understand that these functional blocks can be realized in various forms by hardware only, software only, or a combination thereof.

  The system control unit 70 controls the entire user terminal 16 mainly by the OS. The program control unit 72 controls the operation of a program executed on the user terminal 16. The resource management unit 74 is realized mainly by the operation of the PU 24, and dynamically manages which resources can be provided to the outside in the user terminal 16, and further, the resource binding time and adaptable time for the provision. Realize resource management function. The resource mediation unit 76 is also realized mainly by the operation of the PU 24 and executes a resource mediation function for mediating the distribution of resources on the network.

  The hardware resource to be managed by the resource management unit 74 is, for example, a processing unit such as the PU 24 itself or a plurality of SPUs 30 or a storage device such as the main memory 42. Each processing unit is managed in units of internal components such as the processor 26, the processor 32, the local memory 28, and the local memory 34, and may be provided for external use. The resource management unit 74 dynamically manages which SPU 30 processor 32 and local memory 34 can be provided for a plurality of SPUs 30. The resource management unit 74 provides resource information describing which hardware resources can be provided in the user terminal 16 via at least one of the other user terminal 16, the local server 18, and the network server 20 via the network control unit 46. Send to the appropriate time.

  When the resource management unit 74 receives resource information from another user terminal 16 via the network control unit 46, the use of resources of the other user terminal 16 described in the resource information can be determined as appropriate. The resource management unit 74 distributes specific processing in the program executed by the program control unit 72 to a plurality of SPUs 30 and external resources. The resource management unit 74 dynamically manages which internal resources are provided to which external user terminal 16, the processing content to be executed by the provision, processing time, and the like. The resource management unit 74 can also reserve the use of internal resources for provision to the outside. The resource management unit 74 dynamically manages which resource of which external user terminal 16 is used, and the processing content, processing time, and the like to be executed externally.

  The resource management unit 74 assembles and determines the distributed contents such as which resources are used, the calculation time for using the resources, the storage capacity, and the amount of data when the program is executed, and generates call information specifying these. Resources to which processing is assigned by the call information and the processing contents thereof are resources or processing contents for specific uses such as image processing, content generation processing, and data conversion processing. The allocation of which hardware resources are actually used may be left to the local server 18 or the network server 20, but which resource of which user terminal 16 is specifically used in response to a program request. Alternatively, a method may be used in which the resource management unit 74 clearly specifies the resource to be used in the call information. These are determined by the resource management unit 74 according to the response time required for execution of the requested processing, the processing content, the height of the load, or the like, or according to an explicit specification described in the program to be executed. The management unit 74 determines. In the call information, a plurality of resources or candidates for executing the process may be designated. The call information further specifies at least one of data, procedure, and functional process to be used. In the call information, the call source user terminal 16 is specified as the return destination of the processing result. However, in the case of a call for a local resource, specification of the return destination may be omitted. In the call information, designation of resources, processing contents, and the like is described using tags whose structure is defined in advance by a meta language such as XML (Extensible Markup Language). Call information is formed in the form of a collection of such descriptions.

  FIG. 4 logically shows a mechanism for mediating resources on the network. A plurality of resource mediation functions 50 included in the resource management network 10 represent resource mediation functions implemented by the resource mediation unit 76 of the user terminal 16, the local server 18, and the network server 20, respectively. The resource mediation function 50 transmits / receives resource information to / from the resource mediation function 50 executed in the other user terminals 16, the local server 18, and the network server 20 and exchanges the resource information. The resource mediation function 50 mediates resource distribution by sending resource information received from other user terminals 16, the local server 18, and the network server 20 to another user terminal 16, the local server 18, and the network server 20. Connect resource users and providers. Network nodes such as the user terminal 16, the local server 18, and the network server 20 in which the resource mediation function 50 is realized are managed by a system such as DNS (Domain Name System) in the connected network, and are appropriately neighboring nodes from each node. Resource information function 50 can be called to exchange resource information. The resource mediation function 50 of the node that received the call sends resource information to the resource mediation function 50 of the calling node.

  FIG. 5 logically shows a mechanism for overall management of resource mediation on the network. The plurality of resource mediation functions 50 are connected to the resource mediation gateway 54 directly or indirectly. The resource mediation gateway 54 is connected to a service provider 52 that is a provider of a resource distribution service in the resource management network 10. The resource mediation function 50 uses a predetermined resource mediation gateway 54 or a predetermined network server 20 to obtain history information indicating which terminal has used which terminal resource and which terminal resource has been provided to which terminal. To the service provider 52. The history information is transmitted and received using a protocol with high security. The service provider 52 accumulates the received history information and totals the number of times of use and provision, contents, amount, and the like, and processes charging from the user and charging to the provider according to the totaling result. Distribution and billing of resources may be processed by a token issued by the service provider 52 via the service provider 52. The token may be used in another service provided by the service provider 52. The balance between use and provision of resources in the user terminal 16 may be set in advance for each user terminal 16, or may be set at the level of the network server 20 or the like based on an instruction from the service provider 52. Good. The resource information distribution itself does not necessarily go through the predetermined network server 20 or service provider 52, but the resource mediation function 50 mediates peer-to-peer even at the level between the user terminals 16, so that a widely open resource distribution market is formed. Since the history of use and provision is collected in the service provider 52, the distribution status of the entire market is appropriately managed. In the figure, an example in which resource distribution is mediated by the resource mediation function 50 realized by the user terminal 16, the local server 18, and the network server 20 of each node is shown. However, as a modified example, the resource mediation gateway 54 is used. A configuration that mediates resource distribution in a concentrated manner may be used.

(First application example)
As a first application example using the resource management network 10 in the present embodiment, there is a distributed processing role playing game (RPG). Assume that RPG is executed by the user terminal 16 in the LAN 12. When it comes to a situation where rain is drawn on the RPG screen, the user terminal 16 requests other terminals or servers to perform physical calculations necessary for the rain drawing process. When requesting the calculation of drawing processing that requires such time responsiveness to the outside, it is not desirable to request with unspecified resources and indeterminate processing time. Specify whether to use hardware resources and specify response time. Here, for example, the resource use of the local server 18 connected to the same LAN 12 is designated. The user terminal 16 is appropriately notified of information indicating available resources among the hardware resources of the local server 18. When a calculation is requested from the user terminal 16, the local server 18 performs a specified calculation and sends the result back to the user terminal 16.

  In the user terminal 16, for example, when a hero in the RPG encounters a crowd and specifies a person in a predetermined field from the crowd, a calculation request for person recognition or face recognition for the identification is received from the user terminal 16. Sent. The terminal requested by the user terminal 16 executes a designated person recognition (face recognition) function. Specifically, matching of an image with a person database is executed to identify a person, and the identification result is sent back to the user terminal 16.

  FIG. 6 shows an example of a program description for transmitting resource information from the local server 18 to the user terminal 16 in response to a call from the user terminal 16. In this example, the execution of the method “Register” in the resource information transmission program “Resource” and the name “Console 1” are described as the destination user terminal 16. As a transmission source, the name “HomeServer” of the local server 18 is described, and “profileA” in which a resource that can be provided is defined is described. Furthermore, information such as the number, calculation time, and storage capacity of the MPU 22 included in the local server 18 is described. Resource information indicating these contents is sent from the local server 18 to the user terminal 16.

  FIG. 7 shows a program description example for requesting calculation from the user terminal 16 to the local server 18. In this example, the execution of a method “rain_fall” for rain image processing in a physical calculation program “Physics” and the name “HomeServer” of the local server 18 that is a request destination are described. In addition, “profileA” in which the resource to be used is defined, “land_area1” indicating the place where the rain is drawn, and “condition_profile1” indicating the rain condition are described.

  FIG. 8 shows an example of a program description for calculation execution and result return by the local server 18. In this example, the execution of the method “rain_fall_body” for image processing of rain in the physical calculation program “Physics” and the execution of the method “physics_report” in the result return program “Report” are appropriately processed in parallel. is described.

  FIG. 9 shows an example of a program description for transmitting resource information from the network server 20 to the user terminal 16 in response to a call from the user terminal 16. In this example, the execution of the method “Register” in the resource information transmission program and the name “Console 1” of the destination user terminal 16 are described. As a transmission source, the name “NetworkServerX” of the network server 20 is described, and “profileX1” in which a resource that can be provided is defined is described. Furthermore, information such as the number, calculation time, and storage capacity of the MPU 22 included in the network server 20 is described. Resource information indicating these contents is sent from the network server 20 to the user terminal 16.

  FIG. 10 shows a program description example for requesting calculation from the user terminal 16 to the network server 20. In this example, the execution of a method “identify” for authentication processing in the face authentication calculation program “FaceRecognition” and “NetworkServerX” as a request destination are described. In addition, “profileX1” in which resources to be used are defined and “face_image” indicating an authentication target are described.

  FIG. 11 shows an example of a program description for calculation execution and result return by the network server 20. In this example, it is described that the execution of the authentication process method “identify_body” in the face authentication calculation program “FaceRecognition” and the execution of the method “id_report” in the result return program “Report” are appropriately performed in parallel.

(Second application example)
As a second application example using the resource management network 10 in the present embodiment, there is a distributed processing type search engine. Each node on the resource management network 10 has a search engine function such as text search, image search, person search based on face image matching authentication, database search by field or region, and the like. Each node holds search results at least temporarily, shares resources and data with other search engines, and processes search in cooperation.

  Each node has a preferred query list that lists search queries that each node is good at. The criteria for whether you are good are, for example, whether you can return results with a short response time, whether you can return results with high search accuracy, or whether you can return many results. . Each node is requested to search from a large number of other nodes, and the preferred query list is dynamically and subsequently formed as the search is repeatedly processed.

  FIG. 12 schematically shows the mechanism of the distributed processing type search engine in the application example. Each of a plurality of nodes connected to the resource management network 10 serves as a distributed search engine. In this figure, search engines 60A to 60H are connected to each other, and preferred query lists 62A to 62H are held as respective preferred query lists.

  Hereinafter, a search example will be described. First, the search engine 60A receives a search query from the user, and determines whether or not the search condition included in the query is included in its own preferred query list 62A. The received search query includes a plurality of words such as “Tokyo Taisai” as a search condition. The search engine 60A outputs a corresponding search result for the search condition included in the preferred query list 62A, and the search condition not included is unique to the adjacent search engines 60B, 60C, 60J. The search ID and the query transfer path history are transferred in parallel as a search query. The search engines 60B, 60C, and 60J to which the search query has been transferred also determine whether or not the contents of the query are included in their own preferred query lists 62B, 62C, and 62J. The search result is returned in reverse, and the search condition not included is transferred to another adjacent search engine. For example, since the search condition “Tokyo” is included in its own preferred query list 62B, the search engine 60B returns search results that match “Tokyo” to the search engine 60A. The search engine 60C transfers the search query to the search engines 60D, 60E, and 60F because the content of the search query is not included in its preferred query list 62C. Since the search condition “Tie” is included in its preferred query list 62D, the search engine 60D returns the search result matching “Tie” to the search engine 60A via the search engine 60C. The search engine 60E returns the search result that matches “spicy” to the search engine 60A via the search engine 60C because the search condition “spicy” is included in its preferred query list 62E. The search engine 60A totals the search results received from the search engine 60B, the search engine 60D, and the search engine 60E, calculates the AND set, and presents it to the user.

  A limit may be set on the number of nodes or the number of hops added to the transfer path history so that the search query transfer is properly converged. Although an example has been described in which the search results by each search engine are transferred so as to go back in the transfer node history, a method may be used in which the search results are directly returned to the search engine that is the starting point of the search.

  Each search engine may maintain a preferred query list of search engines that are adjacent or close together on the network. In the example of this figure, the search engine 60A holds, for example, suitable query lists 62B, 62C, and 62J. Each search engine is useless by transferring the corresponding search condition only to the search engine if the search condition included in the search query is included in the preferred query list of the other search engine held. Transfer and unnecessary distribution can be reduced.

  Each search engine not only transfers the search result but also stores it as a cache on the route for returning the search result to the search engine at the starting point. When a similar search query is sent later, the search result is extracted from the cache and returned. This can reduce the circulation of search queries more than necessary, reduce the amount of search processing for the entire system, and increase the response speed.

  Each search engine supports not only text search but also search such as image search and music search. In the case of a search engine specialized in such special search or a search engine specializing in special search, This is described in the preferred query list. In particular, when each search engine holds a preferred query list of another engine, the query transfer route is easily optimized for a special search query, and distribution efficiency and responsiveness can be improved.

  Each search engine predicts queries that are likely to be requested in the future based on past search history and external conditions such as time, weather, temperature, and stock prices, and collects search results in advance. Therefore, the responsiveness may be further improved.

  Any of the search engines may have a function of generating a search query from an image. For example, an image feature amount is extracted from one or more designated image data, and the image feature amount is transferred to another search engine as a search query. When the search engine is a terminal whose network to be frequently connected is changed, such as a mobile terminal, the search query transfer destination may be switched to a nearby search engine each time the connection is made. Each search engine may preferentially set a search engine with a relatively high return ratio of search results as the next transfer destination among search query transfer destinations, or transfer a search engine with a low return ratio. You may perform the process removed from a previous candidate.

  FIG. 13 shows a program description example for providing a search function to other nodes. The process described in this figure is executed each time a value indicating another node is designated, and by executing the search function “Search” using the resource specified by “profile2” for the designated node. , Let the specified node use the resource.

  FIG. 14 shows an example of a program description for using the search function of another node. In the processing described in this figure, resources of other nodes are used by causing the search function “Search” to be executed using resources of other nodes specified by “profile2”.

  FIG. 15 shows an example of a program description for executing the search function. In the example described in this figure, execution of the search program “Search” in its own node, execution of the search program “Search” in each of a plurality of other nodes, and execution of the result return program “search_report” are performed in parallel. Process them. In addition, a value for limiting propagation of search processing in other nodes is added.

  FIG. 16 shows an example of a program description for requesting the resource mediation function to search for another node. In the example described in this figure, a node having a resource mediation function is caused to execute a method “find_resource” for resource search in the resource mediation program “Mediation”.

  FIG. 17 shows a program description example for linking a search function with another node. In the example described in this figure, the execution of the resource search method “find_resource” of the resource mediation program “Mediation” in its own node, and the execution of the resource search method “find_resource” of the resource mediation program “Mediation” in the neighboring nodes; The execution of the result reply program “search_report” is processed in parallel. As a result return destination, the node of the resource mediation gateway 54 is designated in addition to the search requesting node. In addition, a value for limiting propagation of search processing in other nodes is added.

  The present invention has been described based on the embodiments. This embodiment is an exemplification, and it is understood by those skilled in the art that various modifications can be made to each component and combination of processing processes, and such modifications are within the scope of the present invention.

The structure of a resource management network is shown typically. It is a figure which illustrates the hardware constitutions of a user terminal. It is a functional block diagram which shows the process and operation | movement by a user terminal or a local server. Logically shows a mechanism for mediating resources on a network. Logically shows a mechanism for overall management of resource mediation on a network. It is a figure which shows the example of a program description for transmitting the resource information from a local server to a user terminal in response to the call from a user terminal. It is a figure which shows the example of a program description for requesting calculation with respect to a local server from a user terminal. It is a figure which shows the example of a program description for the calculation execution by a local server, and a result reply. It is a figure which shows the example of a program description for transmitting the resource information from a network server to a user terminal in response to the call from a user terminal. It is a figure which shows the example of a program description for requesting calculation with respect to a network server from a user terminal. It is a figure which shows the example of a program description for the calculation execution by a network server, and a result reply. A mechanism of a distributed processing type search engine in an application example is schematically shown. It is a figure which shows the example of a program description for providing the search function to another node. It is a figure which shows the example of a program description for utilizing the search function of another node. It is a figure which shows the example of a program description for performing a search function. It is a figure which shows the example of a program description for requesting | requiring the search of another node with respect to a resource mediation function. It is a figure which shows the example of a program description for linking a search function between other nodes.

Explanation of symbols

  10 resource management network, 16 user terminal, 18 local server, 20 network server, 22 MPU, 24 PU, 30 SPU, 48 resource management function, 50 resource mediation function, 52 service provider, 54 resource mediation gateway, 70 system control unit, 72 Program control unit, 74 Resource management unit, 76 Resource mediation unit.

Claims (5)

A device comprising a multiprocessor,
A system control unit for controlling the operation of the entire apparatus;
A program control unit for controlling execution of the program;
A usage situation in which at least a part of hardware resources including the multiprocessor in the external device is used via the network for processing execution of the program, and at least one of hardware resources including the multiprocessor in the device A resource management unit that dynamically manages the usage status used for execution of processing by an external device via a network according to the status;
With
The resource management unit, whether to use any of the hardware resources in the external device on the network, and a reference whether explicitly specified can be processed with the response time to perform the processing of their A network resource management device characterized by determining.   Resources that mediate the distribution of resources by transferring resource information indicating that at least some of the hardware resources including the multiprocessor in the external device are available to another external device via the network The network resource management apparatus according to claim 1, further comprising an intermediary unit. A plurality of resource management terminals each including a multiprocessor and a resource mediation server,
Each of the plurality of resource management terminals is
A system control unit for controlling the operation of the entire terminal;
A program control unit for controlling execution of the program;
A usage situation in which at least a part of hardware resources including a multiprocessor in another terminal is used via a network for processing execution of the program, and at least hardware resources including a multiprocessor in the terminal A resource management unit that dynamically manages a usage situation that is partly used for processing execution by another terminal via a network, and according to the situation,
The resource management unit, another or to use any of the hardware resources in the terminal, the reference whether explicitly specified can be processed with the response time to perform the processing of that on the network determined by the,
The resource mediation server receives resource information indicating that at least a part of hardware resources including a multiprocessor in the terminal is available from at least one of the plurality of resource management terminals. A network resource management system that mediates the distribution of resources by transferring the data to another resource management terminal via a network.   The resource mediation server records a relationship between use and usage when the hardware resource in at least one of the plurality of resource management terminals is used for processing execution in another resource management terminal. 4. The network resource management system according to claim 3, wherein the record is transmitted to a server that processes billing of the resource use fee. Generating resource information indicating that at least a part of hardware resources including the multiprocessor in the first terminal is available from the first terminal including the multiprocessor;
Transmitting the resource information to a server via a network;
The server receiving the resource information transfers it to the second terminal;
Second terminal the resource information has been transferred, or can be processed by explicitly specified response time to perform the processing of a program in the terminal hardware resources the second specified by the resource information Determining whether or not,
Utilizing the hardware resource specified by the resource information for execution of processing in the second terminal when it is determined by the determining step that processing is possible in the response time;
The first terminal recording a usage situation in which the hardware resource in the first terminal is used for processing execution by the second terminal via a network;
Recording a usage situation in which the second terminal uses the hardware resource in the first terminal for processing execution in the second terminal via a network;
A network resource management method comprising:

Images