JP5526780B2 - Load distribution system, service processing server, load distribution method, and load distribution program - Google Patents

Description

  The present invention relates to a load distribution system, a load distribution method, and a program therefor, and more particularly, a load distribution system and a load distribution method that distribute and process requests when a large number of clients concentrate on one service processing server. And the program.

  When a large number of requests from multiple clients are concentrated on a service processing server in one location, a method has been used in which multiple service processing servers are prepared and a load balancer is installed in front of the service processing server. ing.

  In such a method, a virtual address is allocated to a load balancer between the client and the service processing server. A plurality of service processing servers are prepared ahead of the load balancer. The client transmits a request to the virtual address, and the load balancer that receives the request transfers the process to the optimum service processing server according to a certain algorithm. Examples of algorithms in this case include a round robin method, a minimum connection method, and a fastest method.

  As a technique for load distribution in a network, for example, there is a technique described in Patent Document 1. In the technique disclosed in Patent Document 1, a node on a network autonomously detects a load, and exchanges the load information with another node to determine a node on which the load is concentrated on the network. Then, by allocating resources for each node based on the determination result, the load is efficiently distributed.

Further, in Patent Document 2, when the load of a service providing apparatus that provides a service in response to a request from a client terminal increases, the operation mode is changed, and a part of predetermined processing is performed between the client terminal and the service. There has been proposed a system that realizes load distribution by acting as a relay device that relays communication with a providing device.
JP 2005-303828 A JP 2006-157601 A

  In the above method, where multiple service processing servers are prepared and a load balancer is installed in front of the service processing server, the number of requests that can be processed is determined by the load balancer's ability when a large number of interactive requests must be processed. There is a problem that.

  The reason is that even if there are many service processing servers connected to the load balancer and the processing capacity is sufficient, the number of sessions that the load balancer can maintain is the limit of a system that handles interactive requests. This is because it becomes.

  In the technique described in Patent Document 1, a node on the network autonomously detects a load, and exchanges the load information with another node to determine a node on which the load is concentrated on the network. Therefore, since it is necessary to exchange the load information as appropriate, there is a problem that the load increases due to the transmission of the load information, and effective load distribution cannot be realized.

  Furthermore, in the technique described in Patent Document 2, since a relay device performs only a part of predetermined processing, the processing that can be performed is limited, and load distribution cannot be realized sufficiently efficiently. There's a problem.

(the purpose)
An object of the present invention is to provide a load distribution system, a service processing server, a load distribution method, and a load distribution program that enable a service processing server to autonomously distribute and process requests without installing a load balancer. It is in.

  The load distribution system according to the present invention connects a plurality of service processing servers through an overlay network using a P2P network, routes a service request to a certain service processing server between the service processing servers according to the P2P algorithm, and connects the service processing of the connection destination. Depending on the load state of the server, the service processing server in the middle of routing executes service request processing on behalf of the connection destination service processing server.

  The load distribution method according to the present invention connects a plurality of service processing servers through an overlay network using a P2P network, and when each service processing server receives a service request message from a client, the service distribution server performs service according to the load status of the service processing server. The first step of determining the service processing server that processes the request and when the service request message is transferred from another service processing server, the service processing server processes the service request according to the load status of the service processing server. A second step of determining whether to perform processing, request substitution of processing, or transfer to another service processing server is executed.

  The service processing server according to the present invention is connected to another service processing server via an overlay network using a P2P network, and receives a service request received from a client in accordance with the load status of the connection destination service processing server. Means for executing service request processing on behalf of the processing server;

  The load distribution program according to the present invention, when receiving a service request message from a client to a plurality of computers functioning as a service processing server connected via an overlay network using a P2P network, sends a service request according to the load status of the service processing server. When the service request message is transferred from the first process for determining the service processing server to be processed and the other service processing server, the service processing server processes the service request according to the load status of the service processing server. Or a second process for determining whether to request a proxy for the process or to transfer the process to another service processing server.

  The first effect is that requests can be distributed based on the load state of the service processing server without installing a device such as a load balancer.

  The reason is that the service processing servers are connected by an overlay network using the P2P technology, and requests to the service processing servers are routed between the service processing servers according to the P2P algorithm. This is because the requests are distributed when the processing server processes the request on behalf of the connection destination service processing server.

  The second effect is that a highly available service system can be provided.

  The reason is that the service processing server is connected by an overlay network using P2P technology, so even if the service processing server is stopped for maintenance, it does not affect other service processing servers and This is because adding a service processing server does not affect other service processing servers.

It is a block diagram which shows the structure of the service processing server of the load distribution system by the 1st Embodiment of this invention. It is a figure which shows the whole structure of the load distribution system by the 1st Embodiment of this invention. It is a flowchart explaining the operation | movement at the time of receiving the service request from the client in the 1st Embodiment of this invention. It is a flowchart explaining the operation | movement at the time of receiving a service request from the other service processing server in the 1st Embodiment of this invention. It is a flowchart explaining the operation | movement at the time of receiving the P2P management message from the other service processing server in the 1st Embodiment of this invention. It is a flowchart explaining the operation | movement at the time of receiving the P2P search message from the other service processing server in the 1st Embodiment of this invention. It is a block diagram which shows the structure of the service processing server of the load distribution system by the 2nd Embodiment of this invention. It is a block diagram which shows the hardware constitutions of a service processing server.

(First embodiment)
Hereinafter, a first embodiment of the present invention will be described with reference to the accompanying drawings.

  Referring to FIG. 1, a service processing server 1 constituting a load balancing system according to a first embodiment of the present invention includes a transmission / reception unit 10, a P2P function unit 20, a server function unit 30, and a load detection function unit 40. And a service database 50.

  In the load distribution system according to the first embodiment, as shown in FIG. 2, a plurality of service processing servers 1 are connected to each other by a P2P network.

  The transmission / reception unit 10 is connected to an external network, and transmits a message transmitted from the P2P function unit 20 to another service processing server described in the message. When a message is received from another service processing server, the message is notified to the P2P function unit 20.

The P2P function unit 20 includes a P2P management function unit 21, a P2P search function unit 22, and a P2P database 23.
The P2P function unit 20 functions to construct a P2P network by transmitting and receiving P2P management messages to and from other service processing servers, to transmit and receive P2P search messages using the P2P network, and to search for other service processing servers. Have At this time, the server function 30 is notified of temporary route information as a search result by the P2P function unit 20. Further, when a message related to service processing is received, the server function unit 30 is notified. Further, when a message transmission request related to service processing is notified from the server function unit 30, address resolution by P2P is performed and the message is transmitted to another desired service processing server.

  Here, the temporary route information is a service that passes through until reaching a desired other service processing server obtained as a result of searching by using the P2P function unit 20 which service processing server should process the service request message. It is a history of the processing server. In the P2P network, the route to the desired other service processing server is fluid, so the temporary route information is not stored for a long time. For example, it can be deleted when the service request message has been processed.

  The P2P management function unit 21 has a function of maintaining and managing the P2P network by exchanging management messages with other service processing servers.

  The P2P search function unit 22 searches for another desired service processing server by transmitting a search message to the P2P network. The server function unit 30 is notified of the temporary route information obtained at this time.

  The P2P database 23 is a database that stores information for maintaining a P2P network and making it searchable using the P2P network. For example, when Chord is selected as the P2P algorithm, a finger table corresponding to the distributed hash table (DHT) is stored.

  The server function unit 30 includes a proxy function unit 31, a service transmission / reception function unit 32, and a service processing function unit 33, reads a service from a service database, and processes a service request message transmitted from a client.

  However, if the load notified from the load detection function unit 40 is high, the server function unit 30 does not process the service request message, and uses the temporary route information, and the service processing server between the own service processing server and the client Among them, the service processing server 1 also has a function of transmitting a service request message to the service processing server on the one-hop client side and substituting the service processing. Alternatively, the server function unit 30 has a function of receiving a service request message transmitted using the P2P function unit 20 and requesting service processing when another service processing server is requested to perform service processing. Have.

  The proxy function unit 31 uses the load information notified from the load detection function unit 40 to determine what to do with the transmitted service request message.

  When the service request message addressed to the own service processing server is received when the load is low, the proxy function unit 31 causes the service processing function unit 33 to process the service request message.

  When the service request message addressed to the other service processing server is received when the load is low, the proxy function unit 31 uses the temporary route information to provide a service processing server between the own service processing server and the desired service processing server. Among them, the service request message is transmitted from the own service processing server 1 to the service processing server on the one-hop desired service processing server side using the P2P function unit 20 and the transmission / reception unit 10.

  On the other hand, when the service request message addressed to the own service processing server is received when the load is high, the proxy function unit 31 uses the temporary route information to select the own service among the service processing servers between the own service processing server and the client. The service stored in the service database 50 is read from the processing server to the service processing server on the one-hop client side by using the service transmission / reception function unit 32, and the read service is added to the service transmission message to add the P2P function unit 20 and transmission / reception. It transmits using the part 10.

  Further, when the service request message addressed to the other service processing server is received when the load is high, the proxy function unit 31 uses the temporary route information to select one of the service processing servers between the own service processing server and the client. A load status notification message is transmitted from the service processing server to the service processing server on the one-hop client side using the P2P function unit 20 and the transmission / reception unit 10.

  Here, the service is a program requested by a service request message from the client, and the service is processed by being read by the service processing function unit 33. A service has an expiration date, and even when a request is made to another service processing server, an expired service is not executed. The service is completed by one or more service request messages.

  When the proxy function unit 31 receives a load state notification message from another service processing server, the proxy function unit 31 notifies the P2P function unit 20 to transmit a service request message to a desired service processing server.

  Further, when receiving the service request message from another service processing server, the proxy function unit 31 reads the service stored in the service database 50 using the service transmission / reception function unit 32, and adds the service to the service transmission message. The P2P function unit 20 and the transmission / reception unit 10 are used for notification.

  Further, when receiving a service transmission message from another service processing server, the proxy function unit 31 stores the service added to the service transmission message in the service database 50 and causes the service processing function unit 33 to read the service. The transmission / reception function unit 32 is notified. As a result, the number of services that can be processed increases, and the proxy processing of the service request message is performed.

  The service transmission / reception function unit 32 reads a service stored in the service database according to an instruction from the proxy function unit 31, and sends a service transmission message to which the read service is added to other service processing servers to the P2P function unit 20 and the transmission / reception unit. 10 is used for transmission. When a service transmission message is received from another service processing server in response to an instruction from the proxy function unit 31, the service added to the service transmission message is stored in the service database 50 and the service processing function unit 33 reads the service. Make it.

  The service processing function unit 33 reads a service from the service database 50 and processes a service request message.

  The load detection function unit 40 manages resources in the service processing server, detects the load state of the service processing server, and notifies the server function unit 30 of the load state. At this time, resources to be managed include CPU usage rate, memory usage rate, hard disk drive usage rate, network usage rate, and the like.

  The service database 50 stores services. The service database 50 monitors the expiration date of the service, and deletes the service when an expired service is found.

Next, the operation of the service processing server 1 according to the first embodiment of the present invention will be described in detail with reference to FIG. 3, FIG. 4, FIG. 5, and FIG.

  The operation when a service request is received from a client will be described with reference to the flowchart of FIG.

(1) Step S-A1
In the initial state (step S-0), a service request message is received from the client.

(2) Step S-A2
The server function unit 30 confirms the service request destination of the service request received in step S-A1. When the service request destination is the own service processing server, the process proceeds to step S-A3. If the service request destination is another service processing server, the process proceeds to step S-A4.

(3) Step S-A3
The server function unit 30 processes the service request received in step S-A1. After the process is completed, the process proceeds to the initial state (step S-0).

(4) Step S-A4
Even for a service request addressed to another service processing server, there is a possibility that the service processing has already been performed. Therefore, it is confirmed whether the service within the expiration date is held for the other service processing server. When the service is held, the process proceeds to step S-A3. If the service is not held, the process proceeds to step S-A5.

(5) Step S-A5
In order to obtain temporary route information for transferring the service request received in step S-A1 to a desired service processing server, the P2P database 23 is referred to perform a route search to another service processing server.

(6) Step S-A6
Using the temporary route information obtained in step S-A5, a service request is transmitted to the next service processing server.

(7) Step S-A7
It is confirmed whether the service request transferred in step S-A6 has been accepted. If it is accepted, the process shifts to the initial state. When the load status notification is received without being accepted, the process proceeds to step S-A8.

(8) Step S-A8
With reference to the temporary route information, a service request is transmitted to a desired service processing server.

(9) Step S-A9
The service requested in step S-A8 is received. After receiving the service, the process proceeds to step S-A3, and the service request is processed using the received service.

  Next, the operation when a service request is received from another service processing server will be described with reference to the flowchart of FIG.

(1) Step S-B1
In the initial state (step S-0), a service request message is received from another service processing server.

(2) Step S-B2
Check the load status of the local service processing server. When the load is low, the process proceeds to step S-B3. When the load is high, the process proceeds to step S-B3.

(3) Step S-B3
Check the service request destination of the service request message. When the load is low and the service request destination is the own service processing server in step S-B2, the process proceeds to step S-A3. When the load is low and the service request destination is another service processing server in step S-B2, the process proceeds to step S-A6. When the load is high and the service request destination is the own service processing server in step S-B2, the process proceeds to step S-B4. When the load is high and the service request destination is another service processing server in step S-B2, the process proceeds to step S-B7.

(4) Step S-B4
In step S-B1, a load status notification message is transmitted to the other service processing server that has transferred the service request message.

(5) Step S-B5
In step S-B4, the service request message is received from the other service processing server that has transmitted the load status notification message.

(6) Step S-B6
In step S-B5, the service is transmitted to the other service processing server that has transmitted the service request message.

(7) Step S-B7
In step S-B1, a load status notification message is transmitted to the other service processing server that has transferred the service request message. After that, the initial state is entered.

  The operation when a P2P management message is received from another service processing server will be described with reference to the flowchart of FIG.

(1) Step S-C1
A P2P management message is received from another service processing server.

(2) Step S-C2
The P2P management message including the information in the P2P database is transmitted to the other service processing server that has transmitted the P2P management message in step S-C1.

The operation when a P2P search message is received from another service processing server will be described with reference to the flowchart of FIG.
(1) Step S-D1
A P2P search message is received from another service processing server.

(2) Step S-D2
In step S-C1, the P2P search message is transmitted with reference to the information in the P2P database to the other service processing server that has transmitted the P2P search message.

(Effects of the first embodiment)
Next, the effect of the first embodiment of the present invention will be described.

  In this embodiment, service processing servers are connected by an overlay network using P2P technology, and a request to a certain service processing server is routed between the service processing servers according to a P2P algorithm. When the load on the connection destination service processing server is high, the service processing server in the middle of routing processes the request on behalf of the connection destination service processing server. As described above, since the routing route of the request varies depending on the place where the request is transmitted, the request can be distributed without installing a load balancer.

  Furthermore, in this embodiment, since the service processing server group is connected by the overlay network using the P2P technology, even if the service processing server is stopped for maintenance, the other service processing servers are not affected. It is possible to provide a highly available service system that does not affect other service processing servers even if a service processing server for service enhancement is added.

(Second Embodiment)
Next, a second embodiment of the present invention will be described in detail with reference to the drawings.

  Referring to FIG. 7, the second embodiment of the present invention is similar to the first embodiment of the present invention in that the transceiver 10, the P2P function unit 20, the server function unit 30, and the load detection function unit. 40 and a service database 50. In FIG. 7, the parts denoted by the same reference numerals as those of the constituent elements of the first embodiment are the same as those of the first embodiment, and thus the description thereof is omitted.

  The second embodiment is different from the first embodiment in having a load distribution program 60.

  The load distribution program 60 is read into the transmission / reception unit 10, the P2P function unit 20, the server function unit 30, the load detection function unit 40, and the service database 50, and the transmission / reception unit 10, the P2P function unit 20, and the server This is a program that controls the operation of the function unit 30, the load detection function unit 40, and the service database 50.

  The transmission / reception unit 10, the P2P function unit 20, the server function unit 30, the load detection function unit 40, and the service database 50 are processed by the load distribution system according to the first embodiment under the control of the load distribution program 60. Execute the same process as

  Here, the hardware configuration of the service processing server 1 according to the present embodiment will be described with reference to FIG.

  FIG. 8 is a block diagram showing a hardware configuration of the service processing server 1 according to the present embodiment.

  Referring to FIG. 8, the service processing server 1 according to the present embodiment can be realized by a hardware configuration similar to that of a general computer device, such as a CPU (Central Processing Unit) 301, a RAM (Random Access Memory), and the like. Main memory 302 used for a data work area and a temporary data save area, a communication control part (transmission / reception part 10) 303 for transmitting / receiving data via a network, and data connected to a peripheral device The interface unit 304 that transmits and receives data, the auxiliary storage unit 305 that is a hard disk device composed of a nonvolatile memory such as a ROM (Read Only Memory), a magnetic disk, and a semiconductor memory, and the above-described components of the information processing apparatus are mutually connected. Sys to connect It has a Mubasu 306 or the like.

  In the service processing server 1 according to the present embodiment, the operation is performed by mounting a circuit component which is a hardware component such as an LSI (Large Scale Integration) in which the load distribution program 60 described above is incorporated in the server. Of course, the load distribution program 60 that provides each function of each component described above can be realized by software by the CPU 301 executing it.

  When implemented in software, the CPU 301 loads the load distribution program 60 stored in the auxiliary storage unit 305 into the main storage unit 302 and executes it, thereby executing the functions of the P2P function unit 20 and the server function unit 30 described above. Run.

(Effects of the second embodiment)
According to the second embodiment, the computer can be operated as a load distribution system by using the load distribution system program, and the same effect as the effect of the first embodiment can be obtained.

  Although the present invention has been described with reference to the preferred embodiments and examples, the present invention is not necessarily limited to the above-described embodiments and examples, and various modifications can be made within the scope of the technical idea. Can be implemented.

  This application claims the priority on the basis of Japanese application Japanese Patent Application No. 2007-237779 for which it applied on September 13, 2007, and takes in those the indications of all here.

Claims (44)

A load balancing system for connecting a plurality of service processing servers through an overlay network using a P2P network ,
In load balancing system Ru is routed between the service processing server a service request to a service processing server according P2P algorithm,
The service processing server is
A server function unit for determining whether or not to execute the service related to the service request message on the service processing server;
The server function unit of the service processing server that has received a service request message from a client,
When the service request message is addressed to the own service processing server and the load of the own service processing server is low, and when the destination is another service processing server and the service related to the service request message is held Run the service,
When the service request message is addressed to the own service processing server and the load on the own service processing server is high, and when the destination is the other service processing server and the service request message is not held Sending the service request message to the other service processing server;
The server function unit of the service processing server that has received a service request message from another service processing server,
If the load on the own service processing server is low, and the service request message is addressed to the own service processing server, execute the service,
When the load of the service processing server is low and the service request message is addressed to another service processing server, the service request message is further transmitted to another server,
When the load on the service processing server is high and the service request message is addressed to the service processing server, the service processing server at the transmission source is notified of the load status notification message, and the service at the transmission source is transmitted. Send the service in response to the service request message from the processing server,
When the load of the service processing server is high and the service request message is addressed to another service processing server, the load status notification message is notified to the service processing server of the transmission source. A load balancing system characterized by that. The service processing server
A transmission / reception unit connected to an external network and transmitting / receiving messages to / from other service processing servers;
A P2P function unit that constructs a P2P network by exchanging P2P management messages and P2P search messages with other service processing servers, and searches for other service processing servers using the P2P network ;
A load detection function unit for detecting the load of the own service processing server;
The load distribution system according to claim 1, further comprising a service database that stores and manages services. The transmission / reception unit is connected to an external network and includes a P2P management message, a P2P search message, a load state notification message, a service request message, and a service transmission message that are transmitted from the P2P function unit. When a P2P management message, a P2P search message, a load status notification message, a service request message, and a service transmission message are received from another service processing server, the P2P function unit The load distribution system according to claim 2 , further comprising a function of notifying a message and a function of transmitting the message to a P2P function unit when a service request message is received from a client. The P2P function unit transmits / receives a P2P management message to / from another service processing server to construct a P2P network, and transmits / receives a P2P search message using the P2P network to search for another service processing server. load balancing system according to claim 2 or claim 3 characterized in that it has a P2P database storing the P2P search function section, the information for searching using the P2P network. 5. The load distribution system according to claim 4 , wherein the P2P management function unit has a function of maintaining and managing the P2P network by exchanging management messages with other service processing servers. The P2P search function unit searches a desired other service processing server by transmitting a search message to the P2P network, and notifies the server function unit of temporary route information obtained from the search result. The load distribution system according to claim 4 or 5 , characterized by comprising: The load according to any one of claims 4 to 6 , wherein the P2P database has a function of storing information for maintaining the P2P network and making it searchable using the P2P network. Distributed system. The server function unit includes a proxy function unit that manages a request and reception of a service proxy, a service transmission / reception function unit that transmits / receives a service to / from another service processing server, and a service processing function unit that processes a service stored in the service database load balancing system according to claim 7 claim 2, characterized in that it comprises and. The proxy function unit of the service processing server that has received a service request message from another service processing server ,
When executing a service on its own service processing server, a function for causing the service processing function unit to execute the service ;
When the load on the own service processing server is low and the service request message is destined for another service processing server, the service between the own service processing server and the desired service processing server using temporary route information Among the processing servers, a function of transmitting the service request message from the own service processing server to the service processing server on the 1-hop service processing server side;
Even if the load of the self-service processing server is high, if the service request message is addressed to its own service processing server, among the service processing server between a self service processing server and a client using a temporary route information, A function of reading a service stored in the service database from the own service processing server to a service processing server on the one-hop client side, adding the service to a service transmission message, and transmitting the service;
Even if the load of the self-service processing server is high, if the service request message is addressed to another of the service processing server, the service processing server between a self service processing server and a client using a temporary route information among them, the service processing server 1 hop client side from the own service processing server load balancing system according to claim 8 claims 2, characterized in that it comprises a function of transmitting a state notification message load . The proxy function unit, when receiving the load state notification message from another service processing server according to claim 9, characterized in that it has a function of notifying the P2P functional unit to transmit a service request message to the desired service processing server Load balancing system described in 1. The proxy function unit, when receiving a service request message from another service processing server, reads the service stored in the service database to the service transmission / reception function unit, and notifies the service transmission message to add the service and transmit it The load distribution system according to claim 9 or 10 , characterized by comprising: In the proxy function unit, when a service transmission message is received from another service processing server, the service added to the service transmission message is stored in the service database, and the transmission / reception is performed so that the service processing function unit reads the service. The load distribution system according to any one of claims 9 to 11 , which has a function of notifying a functional unit. The transmission / reception function unit includes:
A function of reading a service stored in the service database according to an instruction from the proxy function unit, and transmitting a service transmission message to which the service is added to another service processing server;
A function that, when a service transmission message is received from another service processing server in response to an instruction from the proxy function unit, stores the service added to the service transmission message in the service database and causes the service processing function unit to read the service load balancing system according to claim 11 claim 9, characterized in that it comprises and. The load distribution system according to any one of claims 2 to 13 , wherein the service processing function unit has a function of reading a service from the service database and executing a service related to a service request message. The load detection function unit manages the resources of the service processing server, service processing to detect the load state of the server, the claims from claim 2, characterized in that it has a function of notifying the server function unit 14 The load distribution system according to any one of the above. The service database, claim claim 2 in which the function of storing service to monitor the expiration date of the service, and having a function of deleting the service in case of finding a service expired The load distribution system according to any one of 15 . A plurality of service processing servers are connected by an overlay network using a P2P network.
When a service request message is received from a client,
When the service request message is addressed to the own service processing server and the load of the own service processing server is low, and when the destination is another service processing server and the service related to the service request message is held Run the service,
When the service request message is addressed to the own service processing server and the load on the own service processing server is high, and when the destination is the other service processing server and the service request message is not held Sending the service request message to the other service processing server; and
When a service request message is received from another service processing server,
If the load on the own service processing server is low, and the service request message is addressed to the own service processing server, execute the service,
When the load of the service processing server is low and the service request message is addressed to another service processing server, the service request message is further transmitted to another server,
When the load on the service processing server is high and the service request message is addressed to the service processing server, the service processing server at the transmission source is notified of the load status notification message, and the service at the transmission source is transmitted. Send the service in response to the service request message from the processing server,
When the load of the own service processing server is high and the service request message is addressed to another service processing server, a load status notification message is notified to the service processing server of the transmission source , Steps ,
The load distribution method characterized by performing. The service processing server is
A third step of returning a P2P management message when a P2P management message is received from another service processing server;
The load balancing method according to claim 17 , further comprising a fourth step of returning a P2P search message when a P2P search message is received from another service processing server. The first step,
When a service request message is received from a client, a step of confirming a service request destination of the received service request ;
As a result of confirming the service request destination, if the service request destination is addressed to another service processing server, confirming whether the service within the expiration date is retained for the other service processing server ;
As a result of confirming the service retention, if the service within the validity period is not retained,
To obtain a temporary route information for forwarding received service request to a predetermined service processing server, the steps of the reference route search to the other service processing server P2P database,
A step of using the acquired temporary route information, and transfers the service request to the next service processing server,
A step of transferring the service request was to verify if it were accepted,
As a result of confirming acceptance of the service request, if the forwarded service request is not accepted,
Referring to the temporary route information, and sending a service request for a given service processing server,
And Luz step is to receive the requested service,
The load distribution method according to claim 17 or 18 , characterized by comprising: In the second step, when a service request message is received from another service processing server ,
A step of checking the load status of the own service processing server ;
Confirming the service request destination of the service request message ;
As a result of the confirmation, when the service request destination is the own service processing server and the load on the own service processing server is high,
To other service processing server which has to forward the service request message, sending a load status notification message,
Receiving a service request message from the other service processing server destination which transmitted the load status notification message,
To other service processing server having transmitted the service request message, and sending the service,
As a result of the confirmation, when the service request destination is another service processing server and the load of the own service processing server is high,
Sending a load status notification message to another service processing server that has transferred the service request message ;
Load balancing method according to claim 19 claim 17, characterized in that it comprises a. The third step,
Receiving a P2P management message from another service processing server ;
Transmitting a P2P management message including information in a P2P database to the other service processing server that has transmitted the P2P management message ;
Load balancing method according to any one of claims 20 to claim 18, characterized in that it comprises a. Said fourth step,
Receiving a P2P search message from another service processing server ;
With respect to the P2P search message other service processing server that sent, and sending the P2P search message referring to the information in the P2P database,
The load balancing method according to any one of claims 18 to 21 , further comprising: A service processing server connected to another service processing server via an overlay network using a P2P network ,
A server function unit for determining whether or not to execute a service related to a service request received from a client on the own service processing server;
The server function unit
When the service request is addressed to the own service processing server and the load of the own service processing server is low, and when the destination is the other service processing server and holds the service related to the service request, Run the service,
Other when the service request is addressed to the own service processing server and the load on the own service processing server is high, and the destination is the other service processing server and the service related to the service request is not held. Sending the service request message to the service processing server of
The server function unit of the service processing server that has received a service request from another service processing server,
If the load on the own service processing server is low, and the service request message is addressed to the own service processing server, execute the service,
If the load on the service processing server is low and the service request is addressed to another service processing server, the service request is further transmitted to another server,
When the load on the service processing server is high and the service request is addressed to the service processing server, the service processing server at the transmission source is notified of a load status notification message, and the service processing at the transmission source is performed. In response to the service request message from the server, send the service,
If the load of the service processing server is high and the service request is addressed to another service processing server, a load status notification message is notified to the service processing server of the transmission source.
A service processing server characterized by the above. A transmission / reception unit connected to an external network and transmitting / receiving messages to / from other service processing servers;
A P2P function unit that constructs a P2P network by exchanging P2P management messages and P2P search messages with other service processing servers, and searches for other service processing servers using the P2P network ;
A load detection function unit for detecting the load of the own service processing server;
The service processing server according to claim 24 or claim 23, characterized in that it comprises a service database that stores and manages service. The transmission / reception unit is connected to an external network, and includes a P2P management message, a P2P search message, a load state notification message, a service request message, and a service transmission message transmitted from the P2P function unit. When a function to transmit to a service processing server, a P2P management message, a P2P search message, a load status notification message, a service request message, and a service transmission message are received from another service processing server, the message is sent to the P2P function unit 25. The service processing server according to claim 24 , further comprising: a function for notifying a client, and a function for transmitting a message to a P2P function unit when a service request message is received from a client. The P2P function unit transmits / receives a P2P management message to / from another service processing server to construct a P2P network, and transmits / receives a P2P search message using the P2P network to search for another service processing server. the service processing server according to claim 24 or claim 25 characterized in that it has a P2P database storing the P2P search function section, the information for searching using the P2P network. 27. The service processing server according to claim 26 , wherein the P2P management function unit has a function of maintaining and managing the P2P network by exchanging management messages with other service processing servers. The P2P search function unit searches a desired other service processing server by transmitting a search message to the P2P network, and notifies the server function unit of temporary route information obtained from the search result. The service processing server according to claim 26 or claim 27 , characterized by comprising: The service according to any one of claims 26 to 28 , wherein the P2P database has a function of storing information for maintaining the P2P network and making it searchable using the P2P network. Processing server. The server function unit includes a proxy function unit that manages a request and reception of a service proxy, a service transmission / reception function unit that transmits / receives a service to / from another service processing server, and a service processing function unit that processes a service stored in the service database the service processing server according to claim 29 claim 24, characterized in that it comprises and. The proxy function unit that has received a service request message from another service processing server ,
When executing a service on its own service processing server, a function for causing the service processing function unit to execute the service ;
When the load on the own service processing server is low and the service request message is destined for another service processing server, the service between the own service processing server and the desired service processing server using temporary route information Among the processing servers, a function of transmitting the service request message from the own service processing server to the service processing server on the 1-hop service processing server side;
Even if the load of the self-service processing server is high, if the service request message is addressed to its own service processing server, among the service processing server between a self service processing server and a client using a temporary route information, A function of reading a service stored in the service database from the own service processing server to a service processing server on the one-hop client side, adding the service to a service transmission message, and transmitting the service;
Even if the load of the self-service processing server is high, if the service request message is addressed to another of the service processing server, the service processing server between a self service processing server and a client using a temporary route information among them, the service processing server 1 hop client side from the own service processing server, the service processing server according to any one of claims 30 to claim 24, characterized in that it comprises a function of transmitting a state notification message load . The proxy function unit, when receiving the load state notification message from another service processing server according to claim 31, characterized in that it has a function of notifying the P2P functional unit to transmit a service request message to the desired service processing server Service processing server described in 1. The proxy function unit, when receiving a service request message from another service processing server, reads the service stored in the service database to the service transmission / reception function unit, and notifies the service transmission message to add the service and transmit it 33. The service processing server according to claim 31 or claim 32 , comprising: In the proxy function unit, when a service transmission message is received from another service processing server, the service added to the service transmission message is stored in the service database, and the transmission / reception is performed so that the service processing function unit reads the service. The service processing server according to any one of claims 31 to 33 , which has a function of notifying a function unit. The transmission / reception function unit includes:
A function of reading a service stored in the service database according to an instruction from the proxy function unit, and transmitting a service transmission message to which the service is added to another service processing server;
A function that, when a service transmission message is received from another service processing server in response to an instruction from the proxy function unit, stores the service added to the service transmission message in the service database and causes the service processing function unit to read the service 36. The service processing server according to any one of claims 24 to 35 , comprising: The service processing server according to any one of claims 24 to 35 , wherein the service processing function unit has a function of reading a service from the service database and executing a service related to a service request message. The load detection function unit manages the resources of the service processing server, service processing to detect the load state of the server, the claims from claim 24, characterized in that it has a function of notifying the server function unit 36 The service processing server according to any one of the above. 25. The service database according to claim 24 , wherein the service database has a function of storing a service and a function of monitoring an expiration date of the service and deleting the service when an expired service is found. 37. The service processing server according to any one of 37 . To a computer functioning as a service processing server connected via an overlay network using a P2P network,
When receiving a service request message from the client,
When the service request message is addressed to the own service processing server and the load of the own service processing server is low, and when the destination is another service processing server and the service related to the service request message is held Run the service,
When the service request message is addressed to the own service processing server and the load on the own service processing server is high, and when the destination is the other service processing server and the service request message is not held A first process for transmitting the service request message to another service processing server ;
When a service request message is received from another service processing server,
If the load on the own service processing server is low, and the service request message is addressed to the own service processing server, execute the service,
When the load of the service processing server is low and the service request message is addressed to another service processing server, the service request message is further transmitted to another server,
When the load on the service processing server is high and the service request message is addressed to the service processing server, the service processing server at the transmission source is notified of the load status notification message, and the service at the transmission source is transmitted. Send the service in response to the service request message from the processing server,
When the load on the service processing server is high and the service request message is addressed to another service processing server, a second process of notifying the service processing server of the transmission source of a load status notification message When,
A load distribution program characterized by causing In the computer,
A third process for returning a P2P management message when a P2P management message is received from another service processing server;
40. The load distribution program according to claim 39 , wherein when a P2P search message is received from another service processing server, a fourth process of returning the P2P search message is executed. The first process,
When a service request message is received from a client , processing for confirming the service request destination of the received service request ,
As a result of confirming the service request destination, when the service request destination is addressed to another service processing server, a process of confirming whether the service within the expiration date is retained for the other service processing server ;
As a result of confirming the service retention, if the service within the validity period is not retained,
To obtain a temporary route information for forwarding received service request to a predetermined service processing server, a process of the reference route search to the other service processing server P2P database,
Using the acquired temporary route information, the process of transferring the service request to the next service processing server,
A process to check if the forwarded service request has been accepted ,
As a result of confirming acceptance of the service request, if the forwarded service request is not accepted,
Referring to the temporary route information, a process of transmitting a service request for a given service processing server,
And processing that will receive the requested service,
41. The load distribution program according to claim 39 or claim 40 , comprising: When the second process receives a service request message from another service processing server ,
A process of checking the load status of the own service processing server ;
Checking the service request destination of the service request message ,
As a result of the confirmation, when the service request destination is the own service processing server and the load on the own service processing server is high,
To other service processing server which has to forward the service request message, the process of transmitting the load status notification message,
A process of receiving a service request message from the other service processing server destination which transmitted the load status notification message,
To other service processing server having transmitted the service request message, a process of transmitting a service,
As a result of the confirmation, when the service request destination is another service processing server and the load of the own service processing server is high,
A process of sending a load status notification message to another service processing server that has transferred the service request message ;
The load distribution program according to any one of claims 39 to 41 , further comprising: The third process,
A process of receiving a P2P management message from another service processing server;
A process of transmitting a P2P management message including information in a P2P database to the other service processing server that has transmitted the P2P management message ;
Load balancing program according to claim 42 claim 40, characterized in that it comprises a. The fourth process,
A process of receiving a P2P search message from another service processing server ;
A process of referring to information in the P2P database and transmitting a P2P search message to another service processing server that has transmitted the P2P search message ;
Load balancing program according to claim 43 claim 40, characterized in that it comprises a.

Images