JP4995300B2 - Server selection control device, service request device, server selection control method, service request method, server selection control program, service request program, service providing system, and service providing method - Google Patents

Description

  The present invention relates to a server selection control device, a service request device, a server selection control method, a service request method, a server selection control program, a service request program, a service providing system, and a service providing method.

  2. Description of the Related Art Conventionally, a technique for distributing the load of each server device when server devices that provide the same service are installed on a network is known. For example, there are a method using a DNS (Domain Name Service) server (see, for example, Non-Patent Documents 1 and 2), a method using a load distribution transfer device (see, for example, Non-Patent Document 3), and the like.

  In these methods, when receiving a service request from a client device, the DNS server and the load distribution transfer device select a server device that provides a service to the client from a plurality of server devices. At this time, for example, the DNS server and the load distribution transfer device change the server device to be selected every time a service request is received from the client device. Thereby, the load of the server device is distributed.

  Here, for example, when one server device is selected by a DNS server or a load distribution transfer device, each server device is selected with the same probability. In contrast, for example, according to the weighted land robin method used in process scheduling of the operating system, CPUs (Central Processing Units) can be allocated to a plurality of processes at different ratios (for example, non-processing (See Patent Document 4). If this weighted land robin method is applied to the selection of the server device, the server device can be selected with different probabilities.

Cricket Lin, written by Paul Albitz, translated by Nobuo Komine, “DNS & BIND”, 5th edition, published by O'Reilly Japan, 2008. Cricket Lin, translated by Koichi Ito, translated by Nobuo Komine, “DNS & BIND Cookbook”, published by O'Reilly Japan, 2003. By Tony Bruk, translated by Kosho Nabeshima, translated by Haruyo Yokoyama, translated by Kazuya Uetani, "Server Load Balancing Technology", published by O'Reilly Japan, 2001. A. Kuurne, A. P. Miettinen, “Weighted Round Robin Scheduling Strategies in (E) GPRS Radio Interface”, 2004 IEEE 60th Vehicular Technology Conference, Vol 5, pp. 3155-3159, 2004.

  However, generally, a server device that provides a service has an upper limit on the number of services that can be provided. A service request exceeding this upper limit is rejected by the server device. Therefore, when one server device is selected from a plurality of server devices as in the conventional method described above, the selected server device is selected even if the server device is selected with the same probability or a different probability. In some cases, the number of services being provided has already reached the upper limit. In such a case, it is necessary for the client device to send a service request again and select the server device again. As a result, the time until the service is provided to the client may become long.

  The present invention has been made in view of the above, and even when the number of services that can be provided is different for each server device, the load distribution of the server device is reduced while shortening the time until the service is provided to the client. It is an object of the present invention to provide a server selection control device, a service request device, a server selection control method, a service request method, a server selection control program, a service request program, a service providing system, and a service providing method that can be realized.

  In order to solve the above-described problems and achieve the object, the present invention provides a server selection control for selecting a server device as a transfer destination of a service request received by a service request device that transfers a service request from a client device to a server device. Each time a request for a server device priority list indicating the order of server devices to be accessed as a transfer destination of the service request is received from the service request device, the number of services that can be provided simultaneously falls within a predetermined range. A priority list creating means for creating a server device priority list including a plurality of groups obtained by grouping the server devices, and changing the priority ranking of each group and / or the priority ranking of the server devices in each group; The server device priority list created by the priority list creation means is used as the service requirement. Characterized in that a priority list transmitting means for transmitting device.

  In addition, the present invention provides a service request device that transfers a service request received from a client device to a server device, and each time the service request is received, a plurality of servers in which the number of services that can be simultaneously provided falls within a predetermined range A priority list acquisition means for acquiring a server device priority list including a plurality of groups each grouping devices, wherein the priority ranking of each group and / or the priority ranking of the server devices in each group is switched; A service request that attempts to access each server device included in the server device priority list acquired by the priority list acquisition means in the order of priority of the group, and forwards the service request to the server device that has successfully accessed first. Means.

  The present invention also provides a server selection control method applied to a server selection control device that selects a server device that is a transfer destination of a service request received by a service request device that transfers a service request from a client device to the server device. Each time the server selection control device receives a request for a server device priority list indicating the order of server devices to be accessed as a transfer destination of the service request from the service request device, the number of services that can be provided simultaneously is within a predetermined range. Priority list creation for creating a server device priority list that includes a plurality of groups obtained by grouping a plurality of server devices belonging to, and in which the priority order of each group and / or the priority order of server devices in each group is switched Process and server device priority list created by the priority list creation process The characterized in that it contains a priority list transmission step of transmitting to said service requestor.

  The present invention is also a service request method applied to a service request apparatus that transfers a service request received from a client apparatus to a server apparatus, and is provided simultaneously each time the service request apparatus receives the service request. A server that includes a plurality of groups in which a plurality of server devices each having a possible number of services belonging to a predetermined range are grouped, and in which the priority order of each group and / or the priority order of server devices in each group is switched A priority list acquisition step of acquiring a device priority list, and attempts to access each server device included in the server device priority list acquired by the priority list acquisition step in the order of priority of the group. And a service request step for transferring the service request to a successful server device. And wherein the door.

  In addition, the present invention is a server selection control program that causes a computer to function as the server selection control device.

  In addition, the present invention is a service request program that causes a computer to function as the service request device.

  The present invention also provides a service request apparatus that transfers a service request received from a client apparatus to a server apparatus, and a server selection control apparatus that selects a server apparatus that is a transfer destination of the service request received by the service request apparatus. A service that can be provided simultaneously each time the server selection control device receives a request for a server device priority list indicating the order of server devices to be accessed as a transfer destination of the service request from the service request device. A server device priority list including a plurality of groups each of which is a group of a plurality of server devices whose number falls within a predetermined range, and in which the priority order of each group and / or the priority order of server devices in each group is switched Priority list creation means to create and the priority list creation means Priority list transmitting means for transmitting the formed server apparatus priority list to the service request apparatus, and the service request apparatus acquires priority list acquisition means for acquiring the server apparatus priority list transmitted from the server selection control apparatus; The server device tries to access each server device included in the server device priority list acquired by the priority list acquisition means in the order of priority of the group, and transfers the service request to the server device that has successfully accessed first. And a service request means.

  The present invention also provides a service request apparatus that transfers a service request received from a client apparatus to a server apparatus, and a server selection control apparatus that selects a server apparatus that is a transfer destination of the service request received by the service request apparatus. A service providing method applied to a providing system, wherein the server selection control device receives a request for a server device priority list indicating an order of server devices to be accessed as a transfer destination of the service request from the service request device. Including a plurality of groups in which a plurality of server devices whose number of services that can be provided simultaneously falls within a predetermined range are grouped, and the priority order of each group and / or the priority order of server devices in each group A priority list creating step of creating a replaced server device priority list; A priority list transmission step of transmitting the server device priority list created by the priority list creation step to the service request device, and a priority at which the service request device acquires the server device priority list transmitted from the server selection control device. A list acquisition step, and attempts to access each server device included in the server device priority list acquired by the priority list acquisition step in the order of priority of the group, and the service is provided to the server device that has been successfully accessed first. And a service request process for transferring the request.

  According to the present invention, even when the number of services that can be provided is different for each server device, it is possible to achieve load distribution of the server device while reducing the time until the service is provided to the client. .

FIG. 1 is a functional block diagram illustrating the configuration of the service providing system according to the first embodiment. FIG. 2 is a diagram for explaining a load distribution technique using a DNS server. FIG. 3 is a diagram for explaining a load distribution technique using a load distribution transfer apparatus. FIG. 4 is a diagram for explaining the weighted land robin method. FIG. 5 is a diagram showing a configuration of a service providing system having a service requesting device and a server selection control device. FIG. 6 is a functional block diagram illustrating the configuration of the service providing system according to the second embodiment. FIG. 7 is a schematic diagram illustrating a group configuration storage unit according to the second embodiment. FIG. 8 is a diagram illustrating an example of the server device priority list created by the priority list creation unit according to the second embodiment. FIG. 9 is a diagram for explaining a server device priority list creation method by the priority list creation unit according to the second embodiment. FIG. 10 is a diagram for explaining a group priority determination method according to the fourth embodiment. FIG. 11 is a diagram for explaining the group configuration storage unit according to the fifth embodiment. FIG. 12 is a flowchart illustrating an overall flow of the operation of the server selection control apparatus according to the fifth embodiment. FIG. 13 is a flowchart showing a processing procedure of initialization processing of the server selection control apparatus shown in FIG. FIG. 14 is a flowchart of a process procedure of the server apparatus priority list creation process shown in FIG. FIG. 15 is a flowchart of a process procedure of the server apparatus payout position movement process shown in FIG. FIG. 16 is a diagram illustrating an example of the storage order of server devices in the server device priority list when the group in the fifth embodiment is in two stages. FIG. 17 is a flowchart illustrating the operation of the service requesting apparatus according to the fifth embodiment. FIG. 18 is a diagram for explaining rearrangement of the server device priority list shown in FIG. FIG. 19 is a flowchart illustrating the operation of the service requesting apparatus according to the sixth embodiment. FIG. 20 is a flowchart illustrating a flow of a group priority assignment method according to the eighth embodiment. FIG. 21 is a diagram illustrating an example of the storage order of server devices in the server device priority list when the first stage is excluded from the example illustrated in FIG. 16.

  Hereinafter, embodiments of a server selection control device, a service request device, a server selection control method, a service request method, a server selection control program, a service request program, a service providing system, and a service providing method according to the present invention will be described in detail with reference to the drawings. Explained. In addition, this invention is not limited by the Example shown below.

  First, the configuration of a service providing system according to the first embodiment will be described with reference to FIG. As illustrated in FIG. 1, the service providing system according to the first embodiment includes a server selection control device 10 and a service requesting device 20. The service request device 20 transfers the service request received from the client device 30 to the plurality of server devices 40. The server selection control device 10 selects a server device to which a service request received by the service request device 20 is transferred.

  The server selection control device 10 includes a priority list creation unit 11 and a priority list transmission unit 12. Each time the priority list creation unit 11 receives a request from the service request device 20 for a server device priority list indicating the order of server devices to be accessed as a service request transfer destination, the number of services that can be provided simultaneously falls within a predetermined range. A server device priority list including a plurality of groups obtained by grouping a plurality of server devices and changing the priority order of each group and / or the priority order of server devices in each group is created. The priority list transmission unit 12 transmits the server device priority list created by the priority list creation unit 11 to the service requesting device 20.

  The service request apparatus 20 includes a priority list acquisition unit 21 and a service request unit 22. The priority list acquisition unit 21 acquires the server device priority list transmitted from the server selection control device 10. The service request unit 22 attempts to access each server device included in the server device priority list acquired by the priority list acquisition unit 21 in the order of the priority of the group. Forward.

  Thus, in this embodiment, the server selection control device 10 notifies the service requesting device 20 of information related to a plurality of server devices. Then, the service requesting device 20 attempts to access the plurality of server devices notified from the server selection control device 10 in the order of priority of each server device. Therefore, according to the present embodiment, even when the number of services that can be provided is different for each server device, it is possible to realize load distribution of the server device while reducing the time until the service is provided to the client.

  As described above, in order to provide services on the network to more clients, server devices that provide the same service are installed on the network for the number of clients according to the amount requested by the clients. To be done. As a technique for distributing service requests from clients to these server apparatuses, there are methods described in Non-Patent Documents 1, 2, and 3 described above.

  The methods described in Non-Patent Documents 1 and 2 are methods using a DNS server used on the Internet or the like. In this method, a plurality of IP addresses are assigned to domain names representing server apparatuses that provide the same service. In response to a domain name-to-IP address conversion request sent from the client, a plurality of IP addresses are sequentially sent to the client by round robin. Thereby, the server device accessed by the client is changed, and the load of the server device for providing the service is distributed.

  FIG. 2 is a diagram for explaining a load distribution technique using a DNS server. For example, a domain name for accessing a server device that provides a service is www. service. jp. In this case, first, the client device 105 sends a domain name www. service. A conversion request from jp to IP address is transmitted.

  The DNS server 104 arranges IP addresses corresponding to the server devices 100, 101, 102, and 103 in this order and responds to the client device 105. Thereafter, the client device 105 first transmits a service request to the server device 100. Here, if the server device 100 is not available, the client device 105 performs the same processing in the order of the server devices 101, 102, and 103.

  Subsequently, when the client devices 106 and 107 make domain name conversion requests in this order, the IP addresses corresponding to the server devices 101, 102, 103, and 100 are arranged in this order, and the server The IP addresses corresponding to the devices 102, 103, 100, and 101 are received in this order, and the same processing as that of the client device 105 is performed. In this way, by changing the server devices that are used in order, it is possible to distribute the load on the server devices for providing services.

  Here, the case where different clients make a series of service requests has been described, but when the same client makes a plurality of service requests, the IP addresses received from the DNS server 104 are used in order, The load when a plurality of service requests are transmitted from the same client can be distributed. For example, in the case of the client device 105, by distributing the service requests in the order of the server device 100 → 101 → 102 → 103 → 100 → 101. .

  The method described in Non-Patent Document 3 is a method using a load distribution transfer device. In this method, apart from the individual IP addresses of a plurality of server devices that provide the same service, a representative IP address that represents the entire plurality of server devices is defined. A service request from a client is sent with this representative IP address as a destination. Then, the load distribution transfer device that transfers the service request performs processing such as changing the destination IP address, thereby changing the server device accessed by the client and distributing the load of the server device for providing the service. .

  FIG. 3 is a diagram for explaining a load distribution technique using a load distribution transfer apparatus. For example, the IP addresses of the server apparatuses 110, 111, 112, and 113 are respectively 192.168.1.110, 192.168.1.111, 192.168.1.112, and 192.168 .. It is assumed that 1.113, and the representative IP address of these server apparatuses is 192.168.1.10.

  When the client apparatuses 115, 116, and 117 transmit a service request to the server apparatus, the client apparatuses 115, 116, and 117 transmit the representative IP address 192.168.1.10 as a destination. When receiving the packet, the load distribution transfer device 114 on the path to which the packet including the service request is transferred receives the address of 192.168 .. which is the IP address of the actual server device for each service request. 1.110 → 192.168.1.111 → 192.168.1.112 → 192.168.1.113 → 192.168.1.110 → 192.168.1.111. . Thereby, the load distribution of the server device for providing the service can be realized.

  In the method using the DNS server, when the number of server devices presented to the client by the DNS server is one, and in the method using the load distribution transfer device, the server devices have the same probability, , DNS server, and load distribution transfer device.

  In selecting such a server device, as described in Non-Patent Document 4, by using a weighted land robin method used in process scheduling of an operating system and priority control of a packet to be transferred, with different probabilities, A server device can be selected.

  FIG. 4 is a diagram for explaining the weighted land robin method. For example, it is assumed that there are three processes A, B, and C, and CPU time is allocated by the process scheduling function of the operating system. For example, when allocating CPU time at a ratio of 2 to 2 to 1 for each of processes A, B, and C, 2 times for process A and 2 times for 5 time slots to be allocated Process B is assigned to process C once. By applying this method to a DNS server and a load distribution transfer device, the probability of selection of the server device can be changed when one server device is selected.

  In the load balancing technique described above, when the load on the server device necessary for providing the service is determined by the number of service requests from the client, the load on the server device can be distributed by the existing technology. .

  Here, consider a case where load distribution is performed in the configuration shown in FIG. In FIG. 5, in response to a service request sent from a client device to a service requesting device, the service requesting device makes a service request to the server device on behalf and provides the client device with the service provided from the server device. To do. The service request apparatus shown in FIG. 5 corresponds to the client shown in FIG.

  Further, consider server server load distribution under the following [Condition 1], [Condition 2], and [Condition 3], which are not considered in the method described in the above-described load distribution technique.

  [Condition 1] In response to a service request from the client device, a service is provided from the server device to the client device via the service request device for a period required by the client device, and each server There is an upper limit to the number of services that can be provided from the device. For this reason, a service request exceeding the upper limit is rejected by the server device.

  [Condition 2] There are a plurality of types of server devices, and the upper limit of the number of services that can be provided differs for each type.

  [Condition 3] When the server device is restarted for maintenance work or the like, the service provided to the client provided by the server device is forcibly disconnected.

  In the above [Condition 1], [Condition 2] and [Condition 3], the following [Problem 1] and [Problem 2] occur.

  [Problem 1] When a method using a DNS server or a method using a load distribution transfer device is used, the probability that each server device is used is the same as described above. For this reason, when the above [Condition 1] and [Condition 2] exist, the server apparatuses having different upper limits of the number of services that can be provided are also used with the same probability. Therefore, a situation in which the load is high occurs in the server device that can provide a smaller number of services. Further, when the number of server devices presented in the method using the DNS server is one and the above-described weighted round robin method is used in combination, the service that can be provided by the presented server device according to [Condition 1] If the upper limit of the number is exceeded, an inquiry must be made to the DNS server again, which causes a problem that it takes a longer time to receive the service. The problem of requiring a long time to receive provision of this service occurs for the same reason even when the method using the load distribution transfer apparatus and the weighted round robin method are combined.

  [Problem 2] When a server device is restarted according to the above [Condition 3], the number of services provided by the server device is zero immediately after the restart. When the method using the DNS server and the method using the load distribution transfer device are used, the server devices are selected equally, so that many restarted server devices are not selected. Therefore, load distribution cannot be realized. In addition, when a method using a DNS server, a method using a load distribution transfer device, and a weighted land robin method are used in combination, as described in [Problem 1], a service request from a service request device Since only one server device is presented, if the number of services provided by the presented server device exceeds the upper limit, there is a problem that it takes a longer time to receive the service. appear.

  A server selection control device, a service request device, a server selection control method, a service request method, a server selection control program, a service request program, a service provision system, and a service provision method according to the present invention, for example, solve the above-described problems related to load distribution To do. Hereinafter, with reference to the drawings, the server selection control device, the service request device, the server selection control method, the service request method, the server selection control program, the service request program, the service provision system, and the service provision method according to the present invention will be described in more detail. A specific embodiment will be described.

  In the second embodiment, a service providing system having a service requesting device and a server selection control device will be described. In the present embodiment, an example will be described in which server devices are managed by a one-stage group, and one or more server devices are assigned to each group.

  First, the configuration of a service providing system according to the second embodiment will be described with reference to FIG. As illustrated in FIG. 6, the service providing system according to the second embodiment includes a server selection control device 50 and a service requesting device 60. The server selection control device 50 and the service requesting device 60 are communicably connected via a network 70. In addition, a plurality of client devices 30 and a plurality of server devices 40 are connected to the service request device 60 via a network 70 so that they can communicate with each other.

The server selection control device 50 selects a server device to which a service request received by the service request device 60 is transferred. The server selection control device 50 includes a communication control unit 51, a group configuration storage unit 52, and a control unit 53. The control unit 53 includes a priority list creation unit 53a and a priority list transmission unit 53b. The communication control unit 51 controls transmission / reception of information exchanged with the service requesting device 60. The group configuration storage unit 52, the priority list creation unit 53a, and the priority list transmission unit 53b will be described in detail later.

  The service request device 60 transfers the service request received from the client device 30 to the plurality of server devices 40. The service requesting device 60 includes a communication control unit 61 and a control unit 62. The control unit 62 includes a priority list acquisition unit 62a and a service request unit 62b. The communication control unit 61 controls transmission / reception of information exchanged among the client device 30, the server device 40, and the server selection control device 50. The priority list acquisition unit 62a and the service request unit 62b will be described in detail later.

  Next, the group configuration storage unit 52 according to the present embodiment will be described with reference to FIG. The group configuration storage unit 52 stores information related to the group configuration of a plurality of server devices. In addition, the group in FIG. 7 represents the first stage group, and the priority represents the group priority. As shown in FIG. 7, in this embodiment, the group configuration storage unit 52 has three first-stage groups, and stores information of group 1, group 2, and group 3, respectively.

  Here, for example, the group priorities of group 1, group 2, and group 3 are represented by P (1), P (2), and P (3), respectively. As specific group priorities, P (1) = 4, P (2) = 2, and P (3) = 1. It is assumed that a group having a higher group priority is given priority. Further, in this embodiment, each group priority simultaneously represents the number of times each group is positioned at the top in the server device priority list for every fixed number of times the server device priority list is created.

  Note that the group configuration stored in the group configuration storage unit 52 is set in advance for the apparatus by the system operator. At this time, the groups in the first stage are used by rearranging them in the order from the largest group priority to the smallest group priority. Further, in this embodiment, as shown in FIG. 7, three, three, and two server devices are assigned to group 1, group 2, and group 3, respectively. In addition, the server apparatus grouped may be exclusive between groups, and may overlap between groups.

  Next, the operation of the server selection control device 50 in this embodiment will be described. First, the priority list transmission unit 53b receives a request for the server device priority list from the service requesting device 60, and passes the received request to the priority list creation unit 53a. When receiving a request for a server device priority list via the priority list transmission unit 53b, the priority list creation unit 53a creates a server device priority list based on information stored in the group configuration storage unit 52.

  For example, when creating the server device priority list, the priority list creation unit 53a uses S (g) and H (g) as variables for each group g. S (g) is the number of times group g is already positioned at the top in the server apparatus priority list. H (g) is the first server device in the group g when the server device included in the group g is arranged in the server device priority list in the group g. Represents. Initial values are set for these variables as follows.

S (1) = 0, S (2) = 0, S (3) = 0
H (1) = Server 1, H (2) = Server 4, H (3) = Server 7

  Here, an example of the server device priority list created by the priority list creation unit 53a will be described with reference to FIG. The group in FIG. 8 represents a group in the server priority list and is called a group combination, and the priority represents the group combination priority.

  As shown in FIG. 8, groups are arranged in the server device priority list, and a group combination priority is assigned to each group. This group combination priority is set in advance in the apparatus by the operator. Note that one or more servers are arranged in each group. In this embodiment, it is assumed that the group combination priority is given a smaller value. In FIG. 8, the groups are arranged in order from the highest group combination priority to the lowest group combination priority, but the order of the groups may be arbitrary.

  When the priority list creation unit 53a creates the server device priority list, the priority list transmission unit 53b transmits the created server device priority list to the service requesting device 60.

  Next, a method for creating a server device priority list by the priority list creation unit 53a will be described with reference to FIG. FIG. 9 is a server device priority list created in response to the first to fourteenth server device priority list requests. The group in FIG. 9 represents a group in the server priority list, and the priority represents a group combination priority.

  First, the priority list creation unit 53a identifies the group having the highest group priority among the information stored in the group configuration storage unit 52 in response to the creation request from the service requesting device 60. Then, the priority list creating unit 53a compares the group priority P (g) with the number of times S (g) that has already become the top in the server device priority list for the identified group g. Here, as a result of comparing S (g) and P (g) as a result of comparing S (g) and P (g), the priority list creation unit 53a adds the group g to the server priority list. Position it at the beginning with.

  For example, in the example illustrated in FIG. 7, the priority list creation unit 53a compares S (1) and P (1) for the group 1 having the highest group priority. Since the priority list creation unit 53a has S (1) = 0 and P (1) = 4, as shown in the first time in FIG. 9, the group 1 is positioned at the head in the server device priority list. (Process 1-1).

  Subsequently, the priority list creation unit 53a arranges the server devices included in the group that is positioned at the head of the server device priority list. At this time, the priority list creation unit 53a places the server device set to H (g) first in the group and sequentially arranges the other server devices. Then, after arranging each server device, the priority list creation unit 53a sets the server device next to the server device arranged at the head to H (g).

  For example, since the priority list creation unit 53a is H (1) = server 1 representing the first server apparatus in the group 1, the group 1 is arranged in the server apparatus priority list. The information of each server is stored in the first group in the order of server 1, server 2, and server 3 (processing 1-2), and this group is represented as group 1. In addition, the group here represents that it is the first in the server device priority list, and is different from the group shown in FIG. After storing the information of each server, the priority list creation unit 53a increments S (1) by 1, and moves H (1) from the server 1 to the server 2 (processing 1-3).

  Subsequently, the priority list creation unit 53a arranges all server devices for the group positioned at the head of the server device priority list, and then places the next group in the server device priority list. The priority list creation unit 53a repeats this process until all groups are placed in the server device priority list.

  For example, the priority list creation unit 53a arranges group 2 in the server device priority list because the group 2 that is the group priority next to the group 1 shown in FIG. 7 is H (2) = server 4. Therefore, the information of each server is stored in the second group in the server device priority list in the order of server 4, server 5, and server 6, and this group is represented as group 2 (processing 1-4). After storing the information of each server, the priority list creation unit 53a moves H (2) from the server 4 to the server 5 (processing 1-5).

  Similarly, the priority list creation unit 53a stores the information of each server in the order of the server 7 and the server 8 in the third group in the server device priority list for the group 3, and this group is stored in the group 3 (Processing 1-6). After storing the information of each server, the priority list creation unit 53a moves H (3) from the server 7 to the server 8 (processing 1-7).

  The first result is as described above (Processing 1-1 to 1-7). At this time, the priority list creation unit 53a assigns a predetermined group combination priority to each group in the server device priority list. Each group g in the server device priority list is assigned a group combination priority C (g). In the example shown in FIG. 9, C (1) = 1, C (2) = 5, C (3 ) = 10.

  The second server device priority list is the same as the first one, but H (1), H (2), and H (3) indicating the first server device in each group are the second in each group. Represents. Therefore, the priority list creation unit 53a stores the server devices included in each group in the server device priority list from the second to the last, returns to the top of the server device, and stores the first server device. That is, in every creation of the server device priority list, all server devices included in each group are stored. Hereinafter, the server device priority list up to the fourth time is created in the same manner as described above.

  Then, as a result of comparing S (g) and P (g) as a result of comparing S (g) and P (g), the priority list creation unit 53a determines the group g having the next highest group priority. When S (g) and P (g) are compared, and S (g) <P (g), the group g is positioned at the head in the server priority list. Thereafter, the priority list creation unit 53a arranges each group and each server device in the server device priority list in the same manner as the first to fourth processing.

  For example, in the example shown in FIG. 9, S (1) = 4 when the fourth server device priority list creation is completed. Here, since S (1) = P (1), the priority list creation unit 53a determines that group 1 cannot be the top in the fifth server device priority list creation, and S (2) of group 2 Since S (2) = 0 is smaller than P (2) = 2, it is determined that group 2 is the head of the server apparatus priority list.

  The server device priority list creation method from here is the same as that described so far. That is, the priority list creation unit 53a stores server device information in the server device priority list in the order of group 2, group 3, and group 1, and the groups in the server device priority list are group 1 and group 2 respectively. Group 3. At this time, the priority list creation unit 53a stores the server devices indicated by H (1), H (2), and H (3) as the head for the order of the server devices in each group as described above. .

  In the sixth server device priority list creation, the priority list creation unit 53a stores the server devices in the order of group 2, group 3, and group 1 in the same manner as described above. In addition, in the seventh server device priority list creation, the priority list creation unit 53a satisfies S (1) = P (1) and S (2) = 2, and P (2) = 2. Therefore, since it is determined that group 2 cannot be placed at the top and S (3) = 0 is smaller than S (3) = 1, the same as above in the order of group 3, group 1, and group 2 The server device information is stored in the server device priority list.

  After the seventh server device priority list is created, S (1) = 4, S (2) = 2, S (1) = 1, and P (1), P (2 ), P (3). Therefore, the priority list creation unit 53a resets S (1) = 0, S (2) = 0, and S (3) = 0, and repeats the above server device priority list creation procedure.

  Thus, for example, in the first to seventh processing, the number of times each group shown in FIG. 7 is positioned at the head in the server device priority list is 4 times for group 1 and 2 times for group 2 respectively. Group 3 is once. Thus, according to the present embodiment, each group can be positioned at the head in the server device priority list at a frequency proportional to the group priority. That is, the priority list creation unit 53a changes the priority order of each group so that the frequency with which the group with the higher group priority is positioned first in the server device priority list becomes higher.

  Next, the operation of the service requesting apparatus 60 according to the present embodiment will be described. First, when receiving a service request from the client device 30, the service request unit 62b requests the server selection control device 50 for a server device priority list including the group combination priority of each group combination g to the priority list acquisition unit 62a. Instruct them to do so. Then, when the priority list acquisition unit 62a succeeds in the request and receives the server device priority list, the service request unit 62b sorts the groups included in the server device priority list in the order of the group combination priority assigned to the group. Rearrange.

  From here, it demonstrates using the example in the case of the 1st payout shown in FIG. In this example, what is stored in the server device priority list is the order of group 1, group 2, and group 3, and the respective group combination priorities are 1, 5, and 10. Therefore, since the group combination priority is given priority to a smaller value, when rearranged in the order of the group combination priority, as in the order originally stored in the server device priority list, group 1, group 2, group It becomes the order of 3.

  Subsequently, the service request unit 62b, for the group combination having the highest group combination priority in the rearranged group combination priority order, that is, the group 1, the server 1, the server 2, and the server 3 included therein. Are sorted in ascending order on the basis of the number of established service provisions in order to provide services to the client apparatus at present with each server apparatus (processing 2-1). At this time, server devices having the same service provision number are arranged in a randomly selected order. The service requesting device 60 stores the number of service provisions established by each server device, and updates the number of service provisions in real time or periodically according to the service provision status. Then, the service request unit 62b makes a service request to the server device in the order in which the sorting is performed (processing 2-2).

  Here, when the service request is successful, the service request unit 62b notifies the client device 30 of the success of the service request and ends the processing (processing 2-3). On the other hand, when the service request fails, the service request unit 62b tries the service requests in the order in which the server devices in the group 1 are sorted (processing 2-4). Then, when there is no server device included in group 1, the service request unit 62b repeats the above operations (processing 2-1 to 2-4) in order of the group combination priority for the groups in the server device priority list. Then, when the service request fails for all the server devices of all groups (group 1, group 2, group 3), the service request unit 62b notifies the client device 30 of the service request failure and performs the processing. finish.

  As described above, in the present embodiment, in the server selection control device 50, the priority list creation unit 53a sends a request for a server device priority list indicating the order of server devices to be accessed as a service request transfer destination from the service request device 60. Each group includes a plurality of groups obtained by grouping a plurality of server devices whose number of services that can be provided simultaneously falls within a predetermined range, and the priority order of each group and / or the priority of server devices in each group Create a server device priority list with the order of Further, the priority list transmission unit 53b transmits the server device priority list created by the priority list creation unit 53a to the service requesting device 60. In the service requesting device 60, the priority list acquisition unit 62a acquires the server device priority list transmitted from the server selection control device 50. In addition, the service request unit 62b tries to access each server device included in the server device priority list acquired by the priority list acquisition unit 62a in the order of priority of the group, and the server device that has been successfully accessed first Forward service request.

  As described above, in the present embodiment, the server selection control device 50 notifies the service requesting device 60 of information related to a plurality of server devices. Then, the service requesting device 60 attempts to access the plurality of server devices notified from the server selection control device 50 in the order of priority of each server device. Therefore, according to the present embodiment, even when the number of services that can be provided is different for each server device, it is possible to realize load distribution of the server device while reducing the time until the service is provided to the client.

  In this embodiment, the priority list creation unit 53a can simultaneously provide a service each time a request for a server device priority list indicating the order of server devices to be accessed as a service request transfer destination is received from the service request device 60. A server device priority list including a plurality of groups each of which is a group of a plurality of server devices whose number falls within a predetermined range, and in which the priority order of each group and / or the priority order of server devices in each group is switched create. Therefore, according to the present embodiment, it is possible to perform load distribution according to the upper limit of the number of services that can be provided by the server device in order to solve [Problem 1].

  Further, according to the present embodiment, the server selection control device 50 can present a plurality of server devices to the service requesting device 60 that is a client. In order to solve the problem of [Problem 2], it takes a long time to provide the service because the service can be provided by sequentially checking whether it can be used or not and communicating with the first available server device. Can do.

  In the present embodiment, the case where the service request unit 62b in the service request device 60 rearranges the groups included in the server device priority list in the order of group combination priority has been described. On the other hand, for example, the server selection control device 50 may rearrange the groups included in the server device priority list in advance.

  In the third embodiment, a modified example of the service requesting device 60 in the second embodiment will be described. The configuration of the service providing system according to the present embodiment is the same as that shown in FIG.

  In the present embodiment, when the service request unit 62b receives a service request from the client device 30, the server selection control device transmits a server device priority list including the group combination priority of each group combination g to the priority list acquisition unit 62a. Instruct 50 to request. Then, when the priority list acquisition unit 62a succeeds in the request and receives the server device priority list, the service request unit 62b sorts the groups included in the server device priority list in the order of the group combination priority assigned to the group. Rearrange.

  From here, it demonstrates using the example in the case of the 1st payout shown in FIG. In this example, what is stored in the server device priority list is the order of group 1, group 2, and group 3, and the respective group combination priorities are 1, 5, and 10. Therefore, since the group combination priority is given priority to a smaller value, when rearranged in the order of the group combination priority, as in the order originally stored in the server device priority list, group 1, group 2, group It becomes the order of 3.

  Subsequently, the service request unit 62b, for the group combination having the highest group combination priority in the rearranged group combination priority order, that is, the group 1, the server 1, the server 2, and the server 3 included therein. In this order, service requests are made (processing 3-1).

  If the service request is successful, the service request unit 62b notifies the client device 30 of the success of the service request and ends the processing (processing 3-2). On the other hand, when the service request fails, the service request unit 62b tries the server 1, the server 2, and the server 3 in order of the server device in the group 1 (processing 3-3). Then, when there are no more server devices included in the group 1, the service request unit 62b repeats the above operations (processing 3-1 to 3-3) for the groups in the server device priority list in the order of group combination priority. Then, when the service request fails for all the server devices of all groups (group 1, group 2, group 3), the service request unit 62b notifies the client device 30 of the service request failure and performs the processing. finish.

  In the fourth embodiment, an example in which the group priority of each group is determined based on the total number of services that can be simultaneously provided by the server apparatuses included in the group in the first and second embodiments will be described. The configuration of the service providing system according to the present embodiment is the same as that shown in FIG.

  In the present embodiment, the group configuration storage unit 52 stores information related to the group configuration of a plurality of server devices, as in the first and second embodiments. Also in the present embodiment, the group configuration stored in the group configuration storage unit 52 is set in advance for the apparatus by the system operator. However, the group priority is determined by the following method. Here, for a group that directly includes server devices, the total number of services that can be simultaneously provided by the server devices included in the group is referred to as “the number of services that can be provided by the group”.

  Here, with reference to FIG. 10, the determination method of the group priority in a present Example is demonstrated. First, it is assumed that the number of services that can be provided simultaneously in each server device included in group 1 is 2000. Since group 1 includes three server devices, the number of services that can be provided by group 1 is 6000. Subsequently, it is assumed that the number of services that can be simultaneously provided in each server device included in the group 2 is 500. Since group 2 includes three server devices, the number of services that can be provided by group 2 is 1500. Finally, in each server device included in group 3, the number of services that can be provided simultaneously is 150. Since group 3 includes two server devices, the number of services that can be provided by group 3 is 300.

  That is, in the group 1, group 2, and group 3, the ratio of the number of services that can be provided is 6000: 1500: 300, and the greatest common divisor is 300. This ratio is therefore 20: 5: 1. In this embodiment, the priority list creation unit 53a sets group priorities of group 1, group 2, and group 3 as 20, 5, and 1, respectively, based on this value. As a result, the frequency at which group 1, group 2, and group 3 are stored at the top of the server device priority list is 20 times, 5 times, and 1 time out of 26 times.

  Here, a case has been described in which server devices having the same number of services that can be provided are grouped into the same group. On the other hand, for example, the range of the number of services that can be provided (for example, 800 to 1000) may be set for each group, and the server devices that provide the number of services within each range may be grouped. .

  Here, the case has been described in which the priority list creation unit 53a sets the group priority of each group based on the “number of services that can be provided by the group” of each group. On the other hand, for example, the group priority of each group may be set for the apparatus in advance by the system operator.

  As described above, in this embodiment, the priority list creation unit 53a determines the group priority of each group based on the total number of services that can be provided simultaneously by the server apparatuses included in the group. Therefore, according to the present embodiment, since the ratio of the total number of services that can be provided by each group is equal in each group, load distribution is appropriately performed from the viewpoint of the group. Moreover, since the frequency with which each service apparatus is located at the head in each group is equal, load distribution is appropriately performed also from the viewpoint of a server apparatus.

  Further, in this embodiment, the server selection control device 50 can present a plurality of server devices to the service requesting device that is a client, and the server device is selected according to the number of services that can be provided simultaneously. It is positioned at the head of the server device priority list. As a result, the server device can be controlled to be used according to the number of services that can be provided. From this, [Problem 1] described above can be solved.

  In the fifth embodiment, an example in which server devices are grouped hierarchically will be described. The configuration of the service providing system according to the present embodiment is the same as that shown in FIG.

  First, the group configuration storage unit 52 in the present embodiment will be described with reference to FIG. The group configuration storage unit 52 stores information related to the group configuration of a plurality of server devices. As shown in FIG. 11, in this embodiment, the group configuration storage unit 52 stores information obtained by grouping server devices in a hierarchical manner. In addition, the server apparatus grouped may be exclusive between groups, and may overlap between groups.

  In this embodiment, there are N (0) groups in the first stage, and the first, second,..., N (0) th groups in the first stage are G (1), G, respectively. (2), ..., G (N (0)).

Subsequently, each group G (1), G (2),..., G (N (0)) in the first stage has N (1), N (2), and N (N (0)) There are second stage groups. Furthermore, let i _{1 be} an integer greater than or equal to 1 and less than or equal to N (0), and each of the _first , second,..., N (i ₁ ) th groups of the second stage group of G (i ₁ ), , G (i ₁ , 1), G (i ₁ , 2),..., G (i ₁ , N (i ₁ )).

Continuing this, i _1st 1st stage group i _2nd 2nd stage group i _3rd 3rd stage group ... i _m-1 th m-1 stage group G (i ₁ , i ₂ , i ₃ ,..., I _m−2 , i _m−1 ). Here, the · · · _{i m-2} th m-2 stage Group _{i 3} th third stage group of _{i 1} th _{i 2} th second stage group of the first stage group, the m It is assumed that there are N (i ₁ , i ₂ , i ₃ ,..., I _m−2 ) −1 stage groups.

In addition, i _1st first stage group i _2nd second stage group i _3rd third stage group ... i _jth jth stage group G (i ₁ , i ₂ , i ₃ ,..., I _j ) is represented by P (i ₁ , i ₂ , i ₃ ,..., I _j ), and the larger one is given priority.

Subsequently, _i is the _first · · · _{i m-1} th m-1 stage Group _{i 3} th third step Group _{i 2} th second stage group of the first stage group G (i ₁ , i ₂ , i ₃ ,..., Im ₂ , im ₁ ) include N (i ₁ , i ₂ , i ₃ ,..., Im ₁ ) server devices. _{, G (i 1, i 2} , i 3, ..., i m-2, i m-1) i m -th server apparatus _E of _{(i 1, i 2, i} 3, ..., i m-1, i _m ).

  Note that the group configuration stored in the group configuration storage unit 52 is set in advance for the apparatus by the system operator.

  Next, the operation of the priority list creation unit 53a in the present embodiment will be described. In this embodiment, when the priority list creation unit 53a pays out the hierarchically grouped server devices, the server device group included under each group is a group that is one higher than the group. Is counted in the server device priority list when compared with the group of server devices under the group included in.

  Further, the priority list creation unit 53a, for the first stage group, when the server device group included under each group is compared with the server device group under the group included in the other first stage group, In the server device priority list, count the number of times that became the top. The payout of the server device here means that the server device is placed in the server device priority list.

Then, the priority list creating unit 53a is, i ₁ th first stage group becomes the top in the first phase group, i ₁ th i ₂ th second stage group of the first stage group, i ₁ th becomes the first of the group included in the first stage group, i ₃ th third stage group of i ₁ th i ₂ th second stage group of the first stage group, i ₁ th ... becomes the first in the one-step group of i _{the second} of the second stage group, i ₁ th of _{the second} i of the first stage group of i _3-th of the second stage group of - of the third stage group · · _{i m-1} th m-1 stage group, the _{i 1-th} first stage group of _{i 2} th _{i 3} th third stage group of the second-stage group · · · _{i m-2} The top of the m-2 stage group _{, G (i 1), G} (i 1, i 2), G (i 1, i 2, i 3), ··· G (i 1, i 2, i 3, ..., i m-2), G (i ₁ , i ₂ , i ₃ ,..., I _m−2 , i _m−1 ) represents the number of times that the head is reached, S (i ₁ ), S (i ₁ , i ₂ ), S (I ₁ , i ₂ , i ₃ ),... S (i ₁ , i ₂ , i ₃ ,..., Im ₂ ), S (i ₁ , i ₂ , i ₃ ,..., Im ₂ , I _m-1 ) is incremented by 1, and the number of times the head is reached is counted.

In the present embodiment, each group priority simultaneously represents the number of times each group is positioned at the top in the server device priority list for every fixed number of times the server device priority list is created. Also, any _{i 1, i 2, i 3} , ..., the _{i m-2, i m-} 1, i m, H (i 1, i 2, i 3, ..., i m-1) is, G In (i ₁ , i ₂ , i ₃ ,..., I _m−2 , i _m−1 ), it is assumed that this is the first server device when the next server device is paid out.

  Moreover, the priority list creation unit 53a is grouped in the server device priority list for each combination of the first stage group, the second stage group,. Hold C (g) as a degree array. That is, the priority list creation unit 53a holds the g-th group combination priority on the server device priority list as an array. C (g) is set in advance in the apparatus by the operator.

  Next, the overall flow of the operation of the server selection control device 50 in this embodiment will be described with reference to FIG. In this embodiment, the priority list creation unit 53a first executes an initialization process (step S90). When the priority list transmission unit 53b receives a request for the server device priority list from the service requesting device 60 (step S91), the priority list creation unit 53a executes server device priority list creation processing (step S92).

  Subsequently, the priority list creation unit 53a executes server device payout position movement processing (step S93). Finally, the priority list transmission unit 53b transmits the server device priority list including the group combination priority (step S94). Thereafter, when the priority list transmission unit 53b receives the request for the server device priority list, the server selection control device 50 repeats the processes of steps S92 to S94.

Next, with reference to FIG. 13, the processing procedure of the initialization process of the server selection control apparatus shown in FIG. 12 will be described. In this embodiment, the group included in the · · · _{i j} th j stage group of _{i 1-th} second stage group _{i 2} th of the first stage group _G (i _1, i 2, _{i 3} , ..., i _j , 1), G (i ₁ , i ₂ , i ₃ , ..., i _j , 2), G (i ₁ , i ₂ , i ₃ , ..., i _j , 3), ... In G (i ₁ , i ₂ , i ₃ ,..., I _j , N (i ₁ , i ₂ , i ₃ ,..., I _j )), P (i ₁ , i ₂ , i ₃ ,..., I _j , 1), P (i ₁ , i ₂ , i ₃ ,..., I _j , 2), P (i ₁ , i ₂ , i ₃ ,..., I _j , 3), ... P (i ₁ , i ₂ , i ₃ ,..., i _j , N (i ₁ , i ₂ , i ₃ ,..., i _j )) is P (i ₁ , i ₂ , i ₃ ,..., i _j , 1) ≧ P _{(i 1, i 2, i} 3, ..., i j, 2) ≧ P (i 1, _{2, i 3, ..., i} j, 3) ≧ ··· ≧ P (i 1, i 2, i 3, ..., i j, N (i 1, i 2, i 3, ..., i j)) It shall be arranged.

  As shown in FIG. 13, in this initialization process, the priority list creation unit 53a sets the number of times that the group has become the head for all groups as 0 by the processes of steps S100 to S111. In addition, the priority list creation unit 53a sets the first server device to be paid out in the group for all the (m−1) -th stage groups by the process of step S112.

  Next, with reference to FIG. 14, the processing procedure of the server apparatus priority list creation process shown in FIG. 12 will be described. In the server device priority list, groups are arranged as shown in FIG. 8, and server devices are arranged in each group. That is, the information of the server device in each group can be represented as a two-dimensional array. For example, the l-th server device in the group g is represented as L (g, l), and the server device information is stored therein.

  As shown in FIG. 14, the priority list creation unit 53a first initializes the group number g in the server device priority list to 1 (step S119), and calculates the first group number in the first stage group (step S119). Step S120). The process shown in step S120 is a process for obtaining a group to be stored first when storing in the server device priority list in the first stage group. In this process, the priority list creation unit 53a satisfies the condition of S (i) <P (i) for the first-stage group in order from the group with the highest group priority to the lowest group. In this process, the first satisfied group is substituted into the variable T, and then the value of S (T) is increased by 1. At this time, the priority list creation unit 53a determines that all the first stage groups i do not satisfy S (i) <P (i) (that is, when S (i) = P (i)). For all first stage groups i, S (i) is set to 0, T = 1, and the value of S (T) is set to 1.

Subsequently, the priority list creating unit 53a substitutes the T calculated in step S120 to i _1, sets the variable k ₁ for counting the number of first phase group stored in the server device priority list 1 ( Step S121). The priority list creation unit 53a repeatedly executes the above processing on the second stage group included in the first stage group G (i ₁ ), and G (i ₁ , i ₂ , i ₃ ,... I _m-2 ) are executed up to the group included in step (S122 to S129).

Subsequently, the priority list creation unit 53a stores the server devices included in G (i ₁ , i ₂ , i ₃ ,... Im ₁ ) in the server device priority list (step S130). A method for storing the server devices included in this group in the server device priority list is as follows. That is, the priority list creation unit 53a includes H (i ₁ , i ₂ , i ₃ ,... Im ₁ ), H (i ₁ , i ₂ , i ₃ ,... Im ₁ ) +1, ... the server devices are stored in the server device priority list in the order of N (i ₁ , i ₂ , i ₃ , ... i _m-1 ) th, 1, 2, ... H (i ₁ , I ₂ , i ₃ ,..., I _m−1 ) −1.

  After all the server devices included in this group are stored in the server device priority list, the priority list creation unit 53a increments the group number g in the server device priority list by 1 (step S131). Thereafter, the priority list creation unit 53a executes server device payout position movement processing (step S132).

Here, with reference to FIG. 15, the server apparatus payout position movement process shown in FIG. 14 will be described. As illustrated in FIG. 15, the priority list creation unit 53a increases H (i ₁ , i ₂ , i ₃ ,..., I _m-1 ) by 1 (step S190). At this time, if H (i ₁ , i ₂ , i ₃ ,... Im ₁ ) exceeds N (i ₁ , i ₂ , i ₃ ,... Im ₁ ) (step S191). , Yes), the priority list creation unit 53a sets H (i ₁ , i ₂ , i ₃ ,..., I _m−1 ) to 1 (step S192), and ends the server apparatus payout position movement process. . On the other hand, if H (i ₁ , i ₂ , i ₃ ,... Im ₁ ) is N (i ₁ , i ₂ , i ₃ ,... Im ₁ ) or less (step S191, No), the priority list creation unit 53a ends the server device payout position movement process without setting H (i ₁ , i ₂ , i ₃ ,... Im ₁ ) to 1.

Returning to the description of FIG. 14, the priority list creation unit 53a then sends G (i ₁ , i ₂ , i ₃ ,... Im ₁ + 1), G (i ₁ , i ₂ , i ₃ ,. ..I _m-1 +2)... G (i ₁ , i ₂ , i ₃ ,... N (i ₁ , i ₂ , i ₃ ,... Im ₂ )), G (i _{1 , i 2, i 3, ···} 1) ··· G (i 1, i 2, i 3, in the order of ··· _{i m-1} -1), the same processing, the server device server priority The list is stored (steps S133 to S136). The priority list creation unit 53a performs this process for all groups at all stages (steps S137 to S152).

  Next, with reference to FIG. 16, an example of the storage order of server devices in the server device priority list when the group is in two stages will be described. In the example shown in FIG. 16, first, there are three groups of G (1), G (2), and G (3) in the first stage. G (1) includes G (1,1), G (1,2), and G (1,3), and G (2) includes G (2,1), G ( 2, 2) and G (2, 3), and G (3) includes two groups G (3, 1) and G (3, 2).

  In the case of this example, the group priority in the first stage is in the order of G (1) → G (2) → G (3). When S (1) and P (1), S (2) and P (2), S (3) and P (3) are compared in this order, S (1) <P (1), not S (2 ) <P (2), G (2) is the first storage target in the first stage. As a result, in the first stage, server devices included in each group are stored in the server device priority list in the order of G (2) → G (3) → G (1).

  Subsequently, the group priority of the second stage group included in G (2) is in the order of G (2,1) → G (2,2) → G (2,3). If the same processing is performed here, G (2, 1) becomes the first storage target. Therefore, first, the server devices included in each group are stored in the server device priority list in the order of G (2,1) → G (2,2) → G (2,3).

  Subsequently, since storage in the server device priority list of the server device of the second stage group included in G (2) is completed, the server included in the group under G (3) which is the next group of the first stage The device is stored in the server device priority list. Since the group priority of the group under G (3) is G (3,1) → G (3,2), the server devices included in each group are stored in the server device priority list in this order.

  Finally, the same processing is performed for the group G (1) which is the final storage order in the first stage, so that G (1,1) → G (1,2) → G (1,3). In order, the server devices included in each group are stored in the server device priority list.

  Next, with reference to FIG. 17, the operation of the service requesting device 60 in the present embodiment will be described. As shown in FIG. 17, in this embodiment, when the service request unit 62b receives a service request from the client device (step S220), the service request unit 62b indicates the group combination priority of the group combination g to the priority list acquisition unit 62a. The server selection control device 50 is requested to request the server device priority list L (g, l) including C (g) (step S221).

  When the priority list acquisition unit 62a succeeds in the request and receives L (g, l) and C (g) (step S222, Yes), the service request unit 62b returns L (g, l). Rearrangement is performed in the order of group combination priorities included in C (g) (step S223).

  Here, the rearrangement of the server device priority list L (g, l) shown in FIG. 17 will be described with reference to FIG. As shown in FIG. 18, first, L (g, l) is a two-dimensional array. That is, there is an array of one or more server devices for each group combination g. This is shown in the table “before arrangement based on group combination priority” shown in FIG. The service request unit 62b sorts in order from the higher group combination priority based on the group combination priority C (g) assigned to each row of this table. At this time, the service request unit 62b replaces each row of the “before arrangement based on group combination priority” table shown in FIG. 18 in the order of priority of the group combination. This is shown in the table “After Arrangement Based on Group Combination Priority” shown in FIG.

  Returning to the description of FIG. 17, after step S223, the service request unit 62b prioritizes the server device for the group combination having the highest group combination priority in the rearranged group combination priority order, that is, the group combination 1. The list is stored in the array T (l) (steps S224 to S228).

  Thereafter, the service request unit 62b further increases T (l) in ascending order based on the number of service provisions that are currently established in order to provide services to client apparatuses with individual server apparatuses. Sorting is performed (step S229). Here, server devices having the same number of service provisions are arranged in a randomly selected order. Then, the service request unit 62b makes a service request to the server device in the sorted order (steps S230 to S234).

  If the service request is successful (step S232, Yes), the service request unit 62b notifies the client device 30 of the success of the service request (step S238), and the process ends. If the service request fails in the determination in step S232 (step S232, No), the service request unit 62b tries the service requests in the sorted order to the server devices in the same group combination g ( Steps S231 to S234).

  Here, when there is no server device included in the group combination g (No in step S234), the service request unit 62b repeats the above operation in the order of group combination priority of the group combination g (steps S225 to S236). . If the service request fails for all server devices of all group combinations g (No in step S236), the service request unit 62b notifies the client device 30 of the service request failure (step S237). The process is terminated.

  The service request unit 62b notifies the client device of the service request failure even when acquisition of the server device priority list and the group combination priority of each group combination fails in step S221 (step S222, No). (Step S237), and the process ends.

  As described above, in the present embodiment, the priority list creation unit 53a uses the information related to a plurality of server devices grouped hierarchically, or information on server devices directly assigned to the group, or others. A server device priority list including information on server devices indirectly assigned through the group of layers is created. Therefore, according to the present embodiment, it is possible to distribute the load among the server devices having the relationship represented by the hierarchical structure.

  In the sixth embodiment, a modified example of the service requesting device 60 in the fifth embodiment will be described. The configuration of the service providing system according to the present embodiment is the same as that shown in FIG.

  Here, with reference to FIG. 19, the operation of the service requesting device 60 in this embodiment will be described. As shown in FIG. 19, in this embodiment, when the service request unit 62b receives a service request from the client device (step S200), the service request unit 62b indicates the group combination priority of the group combination g to the priority list acquisition unit 62a. The server selection control device 50 is requested to request the server device priority list L (g, l) including C (g) (step S201).

  When the priority list acquisition unit 62a succeeds in the request and receives L (g, l) and C (g) (Yes in step S202), the service request unit 62b determines L (g, l). Rearrangement is performed in the order of the group combination priorities included in C (g) (step S203). This rearrangement is the same as the operation described in the process of step S223 in the fifth embodiment.

  After step S203, the service request unit 62b stores the group combination with the highest group combination priority in the rearranged group combination priority order, that is, the group combination 1 in L (g, l). In order, a service request is made to the server device (steps S204 to S209).

  If the service request is successful (step S207, Yes), the service request unit 62b notifies the client device 30 of the success of the service request (step S213), and the process ends. If the service request fails in the determination in step S207 (No in step S207), the service request unit 62b tries service requests to the server devices in the same group combination g in order (steps S206 to S206). S209).

  Here, when there are no more server devices included in the group combination g (No in step S209), the service request unit 62b increases the group combination g by 1 (step S210), and then the group of the group combination g The above operations are repeated in the order of combination priority (steps S205 to S211). If the service request fails for all server devices of all group combinations g (No in step S211,), the service request unit 62b notifies the client device 30 of the service request failure (step S212). The process is terminated.

  The service request unit 62b notifies the client device of the service request failure even when acquisition of the server device priority list and the group combination priority of each group combination fails in step S202 (step S202, No). (Step S212), and the process ends.

  In the seventh embodiment, various modifications regarding the fifth and sixth embodiments will be described.

  For example, in the fifth and sixth embodiments, when server devices are grouped hierarchically, an example is shown in which all server devices are grouped the same number of times. It may be different for each group in the stage of direct accommodation.

  In addition, for example, in the fifth and sixth embodiments, the server device does not explicitly describe whether or not the server device can belong to different groups that directly accommodate the server device. It may belong to different groups at the stage of directly accommodating the server device or may not belong to the group at the same time.

  In addition, for example, in the above-described fifth and sixth embodiments, all the groups included in each group and the server devices included in the group are all included in the server device priority list presented from the server selection control device to the service requesting device. Although shown, a predetermined number of groups among the groups included in each group may be used when creating the server device priority list, or for each group, the server devices included in the server device priority list may be used. Of these, a predetermined number of server devices may be used when creating the server device priority list. Further, for example, all server device priority lists created by the methods shown in the fifth and sixth embodiments may be presented to the server selection control device, or may be limited to a predetermined number.

  In the eighth embodiment, an example in which the group priority of each group is determined based on the total number of services that can be simultaneously provided by the server apparatuses included in the group in the fifth, sixth, and seventh embodiments will be described. The configuration of the service providing system according to the present embodiment is the same as that shown in FIG.

  In the present embodiment, the group configuration storage unit 52 stores information related to the group configurations of a plurality of server devices, as in the fifth, sixth, and seventh embodiments. Also in the present embodiment, the group configuration storage unit 52 stores information in which server devices are grouped hierarchically. Here, first, for a group that directly includes server devices, the total number of services that can be simultaneously provided by the server devices included in the group is referred to as “the number of services that can be provided by the group”. Further, regarding the group G that does not directly include the server device, the total number of services that can be simultaneously provided by the server devices included in the group that is hierarchically included in the group G in the hierarchical grouped structure. Similarly, for the group G, it is referred to as “the number of services that can be provided by the group”.

  Here, with reference to FIG. 20, the group priority assignment method in the present embodiment will be described. First, when a group to which a group priority is assigned is a group G, the priority list creation unit 53a determines whether the group G is a first stage group (step S240).

  When the group G is the first stage group (step S240, Yes), the priority list creation unit 53a determines the number of services that can be provided by the group G as the number of services that can be provided by each group in the first stage. An integer multiple of the quotient divided by the greatest common divisor is set as the group priority of the group (step S241), and the process is terminated.

  If it is determined in step S240 that the group G is not the first stage group (No in step S240), the priority list creation unit 53a increases the number of services that can be provided by the group G by one stage including the group G. An integer multiple of the quotient divided by the greatest common divisor of the number of services that can be provided by each group included in the group is set as the group priority of the group (step S242), and the process is terminated.

  In steps S241 and S242, the integer used as an integer multiple is set to 1 in the case of the final stage group, and in other cases, the quotient calculated in each step and the one level included in the group The value obtained by dividing the least common multiple of the sum of the group priorities by the quotient calculated in each step. Moreover, the integer used for integer multiples is not limited to this example, and may be an arbitrary integer.

  Next, with reference to FIG. 16, group priority assignment when the group hierarchy is in two stages will be described. In the example shown in FIG. 16, the ratio of the number of services that can be provided by the group of G (1,1), G (1,2), and G (1,3) is 6000: 4500: 3000 = 4: 3: 2. It becomes. Further, the ratio of the number of services that can be provided by the group of G (2, 1), G (2, 2), and G (2, 3) is 1500: 1200: 900 = 5: 4: 3. The ratio of the number of services that can be provided by the group of G (3, 1) and G (3, 2) is 600: 300 = 2: 1. Thereby, the group priority in the second stage group shown in FIG. 16 is obtained.

  Subsequently, when comparing the first stage, the ratio of the number of services that can be provided by the group of G (1), G (2), and G (3) is 13500: 3600: 900 = 15: 4: 1. Here, if the group priority of the first stage group is set to be an integral multiple of the total group priority of the subordinate second stage group, the groups described in the first stage group of this figure It becomes priority.

  Here, with reference to FIG. 21, consider an example of the storage order of the server devices in the server device priority list when the first stage is removed from the example shown in FIG. The group priority of each group is as shown in FIG. Consider the first nine times for the first server device stored in the server device priority list in FIGS. 16 and 21. In the example shown in FIG. 16, G (1,1) is 4 times → G (1,2) is 3 times → G (1,3) is 2 times. On the other hand, in the example shown in FIG. 21, G (1) is 9 times. As can be seen from this, by increasing the number of stages in the group, it is possible to distribute the load within a shorter time span.

  In the ninth embodiment, a group combination priority assignment method will be described with respect to the second, third, fifth, sixth, and seventh embodiments.

  As described in the second, third, fifth, sixth and seventh embodiments, when the server device priority list is created, the server device / group combination to be used more is stored near the top. Therefore, for example, the group combination priority is set so that the group combination priority decreases in the order from the first group combination to the last group combination in accordance with the order in the server device priority list.

  In the examples of Examples 2, 3, 5, 6, and 7, the priority of a smaller value represents a higher priority. Therefore, for example, the group combination priority is set so that the value increases in the order from the first group combination to the last group combination in accordance with the order in the server device priority list.

  The groups (= group combinations) included in the server apparatus priority list to which this group combination priority is assigned may be transmitted to the service request apparatus in the order created in the second, third, fifth, sixth and seventh embodiments. . Alternatively, transmission may be performed after rearrangement in the order of the group combination having a higher group combination priority to the lower group combination, or may be performed by rearranging in an arbitrary order.

  Furthermore, although the other embodiments are described as being unrelated to the group priority regarding the group combination priority, the group combination priority may be determined based on the group priority. That is, the product of the group priority assigned to the group included in each group combination is obtained and rearranged from a small value to a large value. This rearranged group priority product sequence may be a group combination priority sequence.

  All or part of the functions of the server selection control device and the service requesting device described in the above embodiments can be realized by software. That is, the function of each device is realized by causing a computer to execute a server selection control program and a service request program that define the processing procedure of processing performed by the server selection control device and the service request device described in each embodiment.

  At this time, the server selection control program and the service request program may be preinstalled in the computer, recorded on a removable recording medium such as a magnetic disk, a magneto-optical disk, an optical disk, or a semiconductor memory, or via a network. It may be distributed and appropriately installed on the computer device.

  As mentioned above, although the Example of this invention was described, this invention may be implemented by various different Example besides the Example mentioned above.

  For example, in the second embodiment, the third embodiment, the fifth embodiment, the sixth embodiment, the seventh embodiment, and the ninth embodiment, the group priority becomes the head of the server device priority list with a high probability that the larger group priority is high. However, it may be as follows. That is, the group priority is assigned so that a smaller group priority is used with a high probability, and when the head is the first stage group, all the groups in the first stage group and the other groups have the same higher priority. The assigned group priorities may be used in reverse order for all the groups included in this group. For example, in the case of the example in FIG. 7, group priority 4, 2, and 1 assigned to group 1, group 2, and group 3, respectively, group priority 1, When 2 and 4 are assigned and used, they may be used in reverse order, and may follow the second embodiment, the third embodiment, the fifth embodiment, the sixth embodiment, the seventh embodiment, and the ninth embodiment.

  Further, for example, in the fourth and eighth embodiments, the method of determining the group priority of each group based on the total number of services that can be simultaneously provided by the server devices included in the group has been described. It may be as follows. That is, after determining the group priority based on the fourth embodiment and the eighth embodiment, when assigning each group priority to each group in the reverse order and creating the server device priority list, the above is described. As such, they may be used in the reverse order. For example, the first stage group in FIG. 16 calculates group priority 45, 12, and 3 for G (1), G (2), and G (3). To P (1) = 3, P (2) = 12, and P (3) = 45 in reverse order, and when creating a server device priority list, use them in reverse order as described above. Also good.

10, 50 Server selection control device 11, 53a Priority list creation unit 12, 53b Priority list transmission unit 20, 60 Service request device 21, 62a Priority list acquisition unit 22, 62b Service request unit 30, 105-107, 115-117, 128 to 130 Client device 40, 100 to 103, 110 to 113, 120 to 123 Server device 51, 61 Communication control unit 52 Group configuration storage unit 53, 62 Control unit 70 Network 104 DNS server 114 Load distribution transfer device

Claims (11)

A server selection control device that selects a server device that is a transfer destination of a service request received by a service request device that transfers a service request from a client device to a server device,
Each time a request for a server device priority list indicating the order of server devices to be accessed as a transfer destination of the service request is received from the service request device, a plurality of server devices in which the number of services that can be provided simultaneously falls within a predetermined range, respectively A priority list creating means for creating a server device priority list that includes a plurality of groups, and in which the priority ranking of each group and / or the priority ranking of the server devices in each group is switched;
A server selection control device comprising: a priority list transmission unit configured to transmit the server device priority list created by the priority list creation unit to the service requesting device.   The priority list creating means changes the priority order of the groups so that the higher the priority of the group is, the higher the frequency of being placed at a higher priority position in the server device priority list. The server selection control apparatus according to claim 1.   The priority list creating means determines the priority of each group based on the total number of services that can be simultaneously provided by the server device so that the priority of the group having the larger total number of services becomes higher. The server selection control apparatus according to claim 1 or 2, wherein A service request device that transfers a service request received from a client device to a server device,
Each time the service request is received, it includes a plurality of groups each grouping a plurality of server devices whose number of services that can be provided simultaneously falls within a predetermined range, and the priority ranking of each group and / or the servers in each group Priority list acquisition means for acquiring a server device priority list in which the priority order of the devices has been exchanged;
A service that attempts to access each server device included in the server device priority list acquired by the priority list acquisition means in the order of priority of the group, and forwards the service request to the server device that has successfully accessed first. And a service requesting device.   The service request means sorts the server devices in ascending order for each group based on the number of service provisions established between the plurality of server devices and the client devices, and sets each server device in the sorted order. 5. The service requesting apparatus according to claim 4, wherein access to the service is attempted. A server selection control method applied to a server selection control device that selects a server device that is a transfer destination of a service request received by a service request device that transfers a service request from a client device to a server device,
The server selection control device is
Each time a request for a server device priority list indicating the order of server devices to be accessed as a transfer destination of the service request is received from the service request device, a plurality of server devices in which the number of services that can be provided simultaneously falls within a predetermined range, respectively A priority list creating step of creating a server device priority list including a plurality of groups, and changing the priority order of each group and / or the priority order of server devices in each group;
A server selection control method, comprising: a priority list transmission step of transmitting the server device priority list created by the priority list creation step to the service request device. A service request method applied to a service request device that transfers a service request received from a client device to a server device,
The service requesting device is
Each time the service request is received, it includes a plurality of groups each grouping a plurality of server devices whose number of services that can be provided simultaneously falls within a predetermined range, and the priority ranking of each group and / or the servers in each group A priority list acquisition step of acquiring a server device priority list in which the priority order of the devices is replaced;
A service that attempts to access each server device included in the server device priority list acquired by the priority list acquisition step in the order of priority of the group, and forwards the service request to the server device that has successfully accessed first. A service request method characterized by including a request process.   A server selection control program for causing a computer to function as the server selection control device according to claim 1, 2 or 3.   A service request program for causing a computer to function as the service request device according to claim 4. A service providing system comprising: a service request device that transfers a service request received from a client device to a server device; and a server selection control device that selects a server device that is a transfer destination of the service request received by the service request device,
The server selection control device is
Each time a request for a server device priority list indicating the order of server devices to be accessed as a transfer destination of the service request is received from the service request device, a plurality of server devices in which the number of services that can be provided simultaneously falls within a predetermined range, respectively A priority list creating means for creating a server device priority list that includes a plurality of groups, and in which the priority ranking of each group and / or the priority ranking of the server devices in each group is switched;
Priority list transmission means for transmitting the server device priority list created by the priority list creation means to the service request device;
The service requesting device is
Priority list acquisition means for acquiring a server device priority list transmitted from the server selection control device;
A service that attempts to access each server device included in the server device priority list acquired by the priority list acquisition means in the order of priority of the group, and forwards the service request to the server device that has successfully accessed first. A service providing system comprising: a request means. The present invention is applied to a service providing system having a service request device that transfers a service request received from a client device to a server device, and a server selection control device that selects a server device that is a transfer destination of the service request received by the service request device. A service providing method comprising:
The server selection control device is
Each time a request for a server device priority list indicating the order of server devices to be accessed as a transfer destination of the service request is received from the service request device, a plurality of server devices in which the number of services that can be provided simultaneously falls within a predetermined range, respectively A priority list creating step of creating a server device priority list including a plurality of groups, and changing the priority order of each group and / or the priority order of server devices in each group;
A priority list transmission step of transmitting the server device priority list created by the priority list creation step to the service request device;
The service requesting device is
A priority list acquisition step of acquiring a server device priority list transmitted from the server selection control device;
A service that attempts to access each server device included in the server device priority list acquired by the priority list acquisition step in the order of priority of the group, and forwards the service request to the server device that has successfully accessed first. A service providing method characterized by including a required process.

Images