JP2016152486A - Subscriber information storage device and subscriber information storage method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a subscriber information storage device capable of efficiently storing unique data of a subscriber using each service.SOLUTION: The subscriber information storage device includes: a processing amount measurement unit 60 for measuring a processing amount being the number of pieces of subscriber information to be processed per unit time; a processing amount distribution calculation unit 70 for calculating a distribution of processing amounts of each of a plurality of services; a subscriber information distribution calculation unit 40 for calculating a distribution of subscriber information of each of the plurality of services; a variation reduction function unit 50 for creating a small group for reducing variations of the distribution of processing amounts and the distribution of subscriber information; a two-dimensional variation correction control unit 80 for monitoring the distribution of processing amounts and the distribution of subscriber information, and controlling the variation reduction function unit 50 so as to create a small group of processing amounts and a small group of subscriber information in the case where one of variation amounts of the distribution of processing amounts and the distribution of subscriber information becomes larger than a threshold; and a storage correction control unit 90 for transferring a portion of the small group to another subscriber information storage device.SELECTED DRAWING: Figure 3

Description

  The present invention relates to a subscriber information accommodation apparatus and a subscriber information accommodation method for accommodating unique data of a subscriber who uses a service.

  The server responsible for subscriber profile management function (UPSF) in IMS (Ip Multimedia Subsystem), the server responsible for policy control function (PCRF) under standardization activities in 3GPP, and various cloud systems respond to requests from multiple service devices It contains the unique data of subscribers who use each service device for the purpose. An accommodation server for accommodating such subscriber specific data is generally added when the amount of information to be accommodated reaches its resource limit.

  Further, as a method for improving the accommodation efficiency of the accommodation server, the idea of improving the accommodation efficiency by improving the performance of processing that becomes a bottleneck is disclosed in Non-Patent Document 1, for example.

http://www.ieice.org/~nv/nv201211-01-tsuji.pdf

  When the performance of the accommodation server is not improved, the addition of the accommodation server is performed when either one of the accommodation processing amount or the accommodation subscriber information amount exceeds the resource limit. The accommodation processing amount is a processing amount that can be processed per unit time by one accommodation server. For example, in the case of a server that accommodates contractor information, a data operation requested from a service device per unit time ( Data registration, update deletion, reference) The number of processes. The accommodated subscriber information amount is a data amount determined by the product of the number of accommodated subscribers and the information amount of the unique information of each subscriber.

  The amount of processing and the amount of subscriber information for each of a plurality of services vary depending on the service characteristics, operation period, and sales promotion status for each service. Since the amount of processing and the amount of subscriber information vary independently from each other, the resources of the accommodation server may not be effectively used and the accommodation may be inefficient.

  FIG. 18 shows a state where the amount of processing and the amount of subscriber information are accommodated in three accommodation servers. FIG. 18 schematically shows a state of accommodation in the case where the accommodation subscriber information amount of the accommodation server # 1 or the accommodation processing amount of the accommodation server # 2 reaches the resource limit and the accommodation server # 3 is added. .

  In the accommodation server # 1, the accommodation subscriber information amount has reached the resource limit, but the accommodation processing amount is about half of the resource limit. On the contrary, in the accommodation server # 2, the accommodation processing amount has reached the resource limit, but the accommodation subscriber information amount is about half of the resource limit.

  Since the accommodation processing amount and the accommodation subscriber information amount of the additional accommodation server # 3 are both about half of the resource limit, the accommodation processing amount of the accommodation server # 3 is assigned to the accommodation server # 1 and the accommodation server # 3. In some cases, it is possible to eliminate the need for additional accommodation servers by accommodating the accommodation subscriber information amounts in the accommodation server # 2.

  As described above, when either the accommodation processing amount or the accommodation subscriber information amount exceeds the resource limit, the addition of the accommodation server may cause the accommodation of the accommodation server (hereinafter referred to as the subscriber information accommodation device) to be inefficient. is there.

  This invention is made in view of this subject, and it aims at providing the subscriber information accommodation apparatus and subscriber information accommodation method which were able to accommodate accommodation capacity and the number of accommodation subscribers efficiently. To do.

  The subscriber information accommodation apparatus according to the present invention provides a subscriber information storage unit that stores subscriber information of a plurality of subscribers for each service, and a service provision that provides a data operation function of the subscriber information to a service providing apparatus. A processing amount measuring unit that measures a processing amount that is the number of subscriber information processed per unit time, and a processing amount distribution calculating unit that calculates a measure of variation in the processing amount of the service for each service, A subscriber information distribution calculating unit for calculating a measure of variation in the subscriber information amount of the service, and monitoring the measure of variation in the processing amount and the measure of variation in the subscriber information amount, When the value of the scale is greater than or equal to a threshold value, the service small group that reduces the variation in the processing amount, and the variation in the subscriber information amount based on the small group is reduced. The two-dimensional variation correction control unit that causes the variation reduction function unit to create a small group of the service, and the order of the data size of each of the processing amount and the subscriber information amount are analyzed in ascending order of the analyzed order. A change amount of the data size is obtained, a value of the center of gravity of the obtained change amount is calculated, a first difference value that is a difference between the ranking of the minimum value of the data size and the value of the center of gravity, and a maximum of the data size A second difference value that is a difference between the ranking of the values and the value of the center of gravity is obtained, and a small number within the range of the first difference value is determined by the number of components of the ratio between the first difference value and the second difference value. The variation reducing function unit that creates the small group combining the data and the large data within the range of the second difference value with respect to the processing amount and the subscriber information amount; and the small amount of the subscriber information amount Group and subscriber information Wherein a portion of each data of the small group of the processing of the same configuration as the small group, and summarized in that comprises a housing correction control section for transferring to other subscriber information accommodation apparatus.

  The subscriber information accommodation method according to the present invention is a subscriber information accommodation method performed by a subscriber information accommodation apparatus, wherein a plurality of subscriptions stored in a subscriber information storage unit for each service with respect to a service providing apparatus. A service providing step for providing a data operation function of subscriber information of a subscriber, a processing amount measuring step for measuring a processing amount which is the number of the subscriber information processed per unit time, and the processing of the service for each service A processing amount distribution calculating step for calculating a measure of variation in amount, a subscriber information distribution calculating step for calculating a measure of variation in the subscriber information amount of the service, a measure of variation in the processing amount, and the subscriber information The amount of variation of the service is monitored, and when the value of any one of the measures is greater than or equal to a threshold value, A two-dimensional variation correction control step for causing a variation reducing function unit to create a group and a small group of the service that reduces variation in the subscriber information amount based on the small group, the processing amount, and the subscriber information Analyzing the rank of each data size of the amount, obtaining the change amount of the data size in ascending order of the analyzed rank, calculating the value of the center of gravity of the obtained change amount, ranking of the minimum value of the data size and the center of gravity A first difference value that is a difference between the first difference value and a second difference value that is a difference between the rank of the maximum value of the data size and the value of the center of gravity, and the first difference value and the second difference value The small group in which the small data within the range of the first difference value and the large data within the range of the second difference value are combined with the processing amount and the subscriber information amount. Before creating about A part of each data of the small group of the processing amount having the same configuration as the small group of the subscriber information amount and the small group of the subscriber information amount is accommodated in other subscriber information. The gist is to perform the accommodation correction control step to be transferred to the apparatus.

  According to the present invention, when either the accommodation processing amount or the accommodation subscriber information amount exceeds the resource limit, the subscriber information accommodation device and the subscriber information accommodation method capable of efficiently accommodating them. Can be provided.

It is a figure which shows typically a processing amount and subscriber information amount. It is a figure which shows the system configuration | structure of the subscriber information accommodation system 1000 which the subscriber information accommodation apparatus of 1st Embodiment comprises. It is an example of a functional structure of the subscriber information accommodation apparatus 100a of 1st Embodiment. It is a figure which shows the operation | movement flow of the variation reduction process which the subscriber information accommodation apparatus 100a performs. It is a figure which shows the example of the result of which the rank analysis part 52 analyzed the processing amount. 6 is a diagram illustrating an example in which a centroid calculating unit 53 determines a change amount in which the data size of the processing amount analyzed by the rank analyzing unit 52 changes in ascending order. It is a figure which shows typically about the case where the accommodation number of an average service is not divisible by a small group, and the case where it is divisible. It is a figure which shows the result of having simulated the change of the standard deviation at the time of changing the composition ratio of small data and large data. It is a figure which shows the variation in the processing amount which the storage part 55 recorded processing amount data by the structure of the small group 1. FIG. It is a figure which shows the variation in the data amount of the subscriber information amount which enabled subscriber information by the structure of the small group 1 for subscriber information. It is a figure which shows the example which arranged the subscriber information amount in ascending order. FIG. 10 is a diagram illustrating an example in which a centroid calculating unit 53 determines a change amount in which the data size of the subscriber information amount analyzed by the rank analyzing unit 52 changes in ascending order. It is a figure which shows the example of distribution of the processing amount recorded by the structure of the small group. It is a figure which shows the example of distribution of the subscriber information amount stored by the structure of the small group. It is a figure which shows the number of subscribers and the amount of processing of the conditions of simulation. It is a figure which shows a simulation result. It is a figure which shows the function structural example of the subscriber information accommodation apparatus 200 of 2nd Embodiment. It is a figure which shows a mode that the processing amount and the number of subscribers are accommodated in three accommodation servers.

  Before describing the embodiment of the present invention, the basic idea of the present invention will be described with reference to FIG.

[Basic idea of the present invention]
The diagram on the left side of FIG. 1 shows the amount of processing, and the diagram on the right side shows the amount of subscriber information. The horizontal axis is service.

  The diagram of step (0) is a diagram schematically showing how the accommodation efficiency of the subscriber information accommodation device decreases when the processing amount of each service and the amount of subscriber information vary. The amount of processing for each service and the amount of data for the subscriber information amount vary as expressed in figures depending on the characteristics of the service, the operation period, the sales promotion situation, and the like. If the amount of processing and the amount of subscriber information are accommodated as they are in the subscriber information accommodation device in units of services, the number of services that can be accommodated per device is reduced.

  Therefore, the idea of the present invention is a method of efficiently accommodating the number of services on the horizontal axis in the subscriber information accommodation device by creating a small group in which the amount of processing and the amount of subscriber information are appropriately combined. is there.

  The figure of step (1) schematically shows that the data can be compressed in the service direction by forming a small group 1 that combines a processing amount Δ and □ with a small data amount and a trapezoid with a large data amount. ing. The figure of the subscriber information amount in the second stage is the same.

  The figure of step (2) shows a state in which the small group 2 is further combined and further compressed in the service direction. The number of services that can be accommodated in one subscriber information accommodation device can be increased by applying the processing constituting the small group to the amount of processing and the amount of subscriber information.

  The figure of step (3) shows a state in which the small group 2 created using the idea of the present invention is accommodated in the three subscriber information accommodating devices # 1, # 2, and # 3. Although the server number is not shown in the figure of step (0), more subscriber information accommodating devices than three are required.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. The same reference numerals are given to the same components in a plurality of drawings, and the description will not be repeated.

[First Embodiment]
FIG. 2 shows a system configuration of a subscriber information accommodation system 1000 configured by the subscriber information accommodation apparatus 100 according to the first embodiment of the present invention. The subscriber information accommodation system 1000 of this embodiment provides subscriber information to a service providing apparatus.

  The service providing apparatus is an apparatus that provides various services to a user, such as a connection server and a billing server. When providing a service to a user, the service providing apparatus acquires subscriber information, which is unique information such as user contract information, from the subscriber information accommodation system 1000.

  The subscriber information accommodation system 1000 is configured by connecting a plurality of subscriber information accommodation devices 100a, 100b, 100c,. The number of subscriber information accommodation devices is determined by the relationship between the resource limit of the subscriber information accommodation device, the amount of subscriber information, and the number of processes required from the service providing device.

  FIG. 3 shows a functional configuration example of the subscriber information accommodation apparatus 100a of the present embodiment. The subscriber information accommodation apparatus 100a is realized by reading a predetermined program into a computer configured with, for example, a ROM, a RAM, a CPU, and the like, and executing the program by the CPU. The functional configuration examples of the other subscriber information accommodation apparatuses 100b and 100c are the same as those in FIG. Therefore, it demonstrates as the subscriber information accommodation apparatus 100 below.

  The subscriber information accommodation apparatus 100 includes a communication I / F 10, a service providing unit 20, a subscriber information storage unit 30, a subscriber information distribution calculation unit 40, a variation reduction function unit 50, and a processing amount measurement unit 60. , A processing amount distribution calculation unit 70, a two-dimensional variation correction control unit 80, and an accommodation correction control unit 90. The communication I / F unit 10 is a functional unit that communicates with other subscriber information accommodation devices (100b, 100c,...) And a service providing device, and is a general one.

  The service providing unit 20 provides a data operation function (registration, update, deletion, reference) of subscriber information in response to a request from the service providing apparatus. The subscriber information storage unit 30 stores subscriber information of a plurality of service subscribers. Here, the subscriber information is determined by the product of the number of subscribers and unique data of each subscriber, and is hereinafter referred to as subscriber information amount. For example, there are three subscribers of service A, and the data size of unique information per person is represented by ■. The amount of subscriber information for service A is 3 × ■. The amount of subscriber information for service C is 5 × ◯.

  The processing amount measuring unit 60 measures a processing amount that is the number of subscriber information processed per unit time. The processing amount is the number of the above-mentioned ■, Δ, ○, etc. provided to the service providing apparatus per unit time. The processing amount measuring unit 60 records the measured processing amount of each service as processing amount data.

  From here, the operation of the subscriber information accommodation apparatus 100 will be described with reference to FIG. FIG. 4 is an operation flow of accommodation maintenance performed for the purpose of reducing the variation in the accommodated data performed after the processing amount and the subscriber information amount are stored in the subscriber information accommodation apparatus 100.

  The subscriber information distribution calculation unit 40 calculates a measure of variation in the amount of subscriber information for each of a plurality of services (step S1). The measure of variation in the subscriber information amount is a numerical value representing the size of the variation shown by the irregularities in the figure in FIG. 1. For example, the standard deviation of the distribution of the subscriber information amount is calculated.

  The processing amount distribution calculation unit 70 calculates a measure of variation in the processing amount measured by the processing amount measurement unit 60 (step S2). Similarly to the subscriber information distribution calculation unit 40, the standard deviation of the processing amount data is calculated. The scale may be any numerical value (statistic) as long as it is a numerical value corresponding to the size of variation.

  The variation reducing function unit 50 creates and stores a small group that combines processing amounts that reduce variation in processing amount and a small group that combines subscriber information amounts that reduce variation in subscriber information amount. Here, the small group in which the processing amounts are combined is a combination of a plurality of processing amounts for each service recorded by the processing amount measuring unit 60. The small group combining the subscriber information amounts is a combination of a plurality of service subscriber information amounts stored in the subscriber information storage unit 30. A small group is represented in the image of the variation image of step (1) in FIG. The variation reducing function unit 50 stores the created small group in the processing amount measuring unit 60 and the subscriber information storage unit 30.

  The two-dimensional variation correction control unit 80 monitors the processing amount scale and the subscriber information amount scale, and when any one of the scales is larger than the threshold value (YES in step S3). Then, the variation reducing function unit 50 is controlled so as to create a small group of services for reducing the variation in processing amount and a small group of services for reducing the variation in subscriber information based on the small group (step S4). ). Here, the threshold value is determined in advance by the system administrator and set in the two-dimensional variation correction control 80.

  The accommodation correction control unit 90 transfers a part of each data of the small group of the subscriber information amount and the small group of the processing amount having the same configuration as the small group of the subscriber information amount to the other subscriber information accommodation devices. (Step S5). For example, when the data size of the subscriber information amount of service C stored in the subscriber information storage unit 30 increases and exceeds the resource limit of the subscriber information accommodation apparatus 100a, the accommodation correction control unit 90 The subscriber information of the service C is transferred to, for example, the subscriber information accommodation apparatus 100b. At this time, the processing amount data of the service C recorded by the processing amount measuring unit 60 is also transferred together. Conversely, the accommodation correction control unit 90 receives, for example, a set of a small group of processing amount and a small group of subscriber information amount transferred from the subscriber information accommodation apparatus 100b.

  A group of small groups transferred by the accommodation correction control unit 90 to another subscriber information accommodation device can be selected by a plurality of selection criteria. For example, a small group with the maximum data amount may be transferred, or a small group with low access frequency may be transferred. Various criteria for selecting a small group to be transferred can be considered, and can be arbitrarily set based on the intention of the system administrator.

  The received small group set is stored in the subscriber information storage unit 30 via the service providing unit 20 in this embodiment. The path for the accommodation correction control unit 90 to read out the small group from the subscriber information storage unit 30 is in the opposite direction to the storage.

  The processing amount data may be stored in the subscriber information storage unit 30 in combination with the subscriber information amount. The processing amount data is a numerical value indicating the number of processings per unit time, and the data amount is relatively smaller than the subscriber information amount. Therefore, even if the processing amount data is stored in the subscriber information storage unit 30, the storage capacity is not greatly affected.

  According to the subscriber information accommodation apparatus 100 that operates as described above, for example, a sales promotion activity of a certain service is successful, and the subscriber information amount or the processing amount data of the service increases, and the subscriber information accommodation apparatus. When the resource limit of 100 is exceeded, the subscriber information and processing data can be automatically transferred to another subscriber information accommodating device. Therefore, it is not necessary to add a subscriber information accommodation device.

[Variation reduction function section]
Next, the operation of the variation reducing function unit 50 will be described in detail with a specific example. As shown in FIG. 3, the variation reduction function unit 50 includes an average accommodation number calculation unit 51, a rank analysis unit 52, a centroid calculation unit 53, a ratio calculation unit 54, and a storage unit 55.

  The average accommodation number calculation unit 51 obtains the average value of the number of subscribers stored in the subscriber information storage unit 30 and the average value of the processing amount recorded in the processing amount measurement unit 60, and determines the resource limit of the own device. The average number of services that can be accommodated by the device itself is calculated using the average value. For example, assuming that the average number of subscribers is 50755, the average throughput is 19534 sessions, the resource limit for subscribers is 3.4 million, and the resource limit for throughput is 1.300000 sessions, the average service capacity is 66 (3400000 ÷ 50755 ≒ 66, 1300000 ÷ 19534 ≒ 66). The session is, for example, the number of times per second. One second of the unit time is an example. For example, 100 milliseconds, 10 milliseconds, or any predetermined fixed time may be used.

  The rank analysis unit 52 analyzes the rank of the data size of the subscriber information amount and the processing amount. FIG. 5 shows an example of the result of analyzing the processing amount by the rank analysis unit 52. The horizontal axis in FIG. 5 is the rank of the data size of the processing amount, and the vertical axis is the processing amount.

  The rank analysis unit 52 first arranges and analyzes the processing amounts in ascending order. When the processing amounts are arranged in ascending order, the processing amount draws a rising curve.

  Subsequently, the center-of-gravity calculation unit 53 uses the rank of the data amount of the processing amount analyzed by the rank analysis unit 52 to obtain a change amount that changes the processing amount in ascending order as shown in FIG. The center of gravity is calculated.

  FIG. 6 is a diagram in which the horizontal axis represents each “order of size of processing amount” and the vertical axis represents “amount of change in processing amount” in ascending order of the order. The center of gravity of the change amount is a rank in which the cumulative value of the first half change amount (vertical axis) of the horizontal axis is balanced with the cumulative value of the second half change amount of the horizontal axis. The center of gravity of the example of FIG.

  The ratio calculation unit 54 includes a “first difference value” that is a difference between the ranking of the minimum value of the processing amount data size and the centroid calculated by the centroid calculation unit 53, and the ranking and centroid of the maximum value of the processing amount data size. The “second difference value”, which is the difference from Data within the range of the first difference value is classified as small data, and data within the range of the second difference value is classified as large data.

  And the ratio calculation part 54 calculates the ratio of a 1st difference value and a 2nd difference value, and calculates the number of the small groups comprised by the calculated | required ratio. In this example, since the first difference value = 351 and the second difference value = 148 (FIG. 6), the first difference value: the second difference value≈7: 3.

  From 7: 3, the number of services that make up a small group is calculated as a total of 10 small data and 7 large data. The average service capacity calculated by the average capacity calculation unit 51 is 66 in this example. The average service capacity of 66 is not divisible by the number of services that make up the small group (10).

  FIG. 7 shows the average number of services accommodated in a concave shape when it is not divisible by a small group and when it is divisible. The case where the left figure of FIG. 7 is not divisible is schematically shown.

  Two small groups with a relatively large data size are contained in the range of the average service capacity, but there is a surplus. The right diagram in FIG. 7 does not have a surplus by reducing the number of services that make up the small group. By not generating a surplus, subscriber information can be accommodated efficiently.

  In this way, when the average number of accommodated services cannot be divided by the number of services constituting the small group, the number of data constituting the small group is reduced. FIG. 8 shows the result of simulating the change in standard deviation when the composition ratio of small data and large data is changed. The horizontal axis is the size ratio (small: large), and the vertical axis is the standard deviation, which is a measure of variation.

  It can be seen that the variation can be reduced most at a large / small composition ratio of 2: 1. It can also be seen that the component ratio 1: 1 is improved to some extent. Therefore, since it was not divisible in the above example, the ratio calculation unit 54 determines 2: 1 which is a divisible number as the sub-optimal small group composition ratio.

  The storage unit 55 is a composition ratio of the small group determined by the ratio calculation unit 54, and configures a small group by combining the small data within the range of the first difference value and the large data within the range of the second difference value. Stored in the subscriber information storage unit 30.

  FIG. 9 shows an example of the small group processing amount stored in the storage unit 55. The horizontal axis in FIG. 9 represents the number of the small group 1 because it is the first variation reducing process with respect to the processing amount. The standard deviation value of the processing amount data recorded in the processing amount measuring unit 60 by combining the small data and the large data in the configuration of the small group 1 is 0.0803. Since the standard deviation value of the processing amount data in the state before the small group 1 is not formed is 0.8061, it can be seen that the variation is improved to about 1/10 of the standard deviation value.

  Subsequently, the variation reducing function unit 50 configures the small group 1 of the subscriber information amount using the composition ratio of the small group 1 obtained by the processing amount. FIG. 10 shows fluctuations in the data amount in which the variation in the subscriber information amount is reduced in the small group 1. In this state, the variation is large. The standard deviation value of the subscriber information amount in FIG. 10 is 0.2480.

  Note that the standard deviation value of the subscriber information amount before the small group 1 is not formed is 0.5699. The standard deviation is less than half of the standard deviation, but it is still insufficient.

  Therefore, the rank analysis unit 52, the centroid calculation unit 53, and the ratio calculation unit 54 perform processing (grouping) to reduce variation using the small group 1 for the subscriber information amount.

  First, the rank analysis unit 52 arranges the subscriber information amounts in the ascending order by combining the subscriber information amounts with the small group composition ratio obtained from the processing amount data. FIG. 11 shows the amount of subscriber information arranged in ascending order. The horizontal axis in FIG. 11 is the same small group 1 as in FIG. 9, and the vertical axis is the data amount.

  Next, the center-of-gravity calculation unit 53 uses the ranks of the data sizes of the subscriber information amounts arranged in ascending order by the rank analysis unit 52, and shows the amount of change in which the subscriber information amount changes in ascending order of rank as shown in FIG. And the center of gravity of the amount of change is calculated. The center of gravity of the example shown in FIG.

  Subsequently, the ratio calculation unit 54 calculates the difference between the ranking of the minimum value of the data size of the subscriber information amount and the centroid calculated by the centroid calculation unit 53, and the first difference value = 112, “Second difference value = 54”, which is the difference between the rank of the maximum value of the data size and the center of gravity, is obtained. Then, the data within the range of the first difference value is classified as small data, and the data within the range of the second difference value is classified as large data.

  Next, the ratio calculation unit 54 calculates the ratio between the first difference value and the second difference value, and calculates the number of constituents of the small group 2 configured with the obtained ratio. In this example, since the first difference value = 112 and the second difference value = 54 (FIG. 12), the first difference value: the second difference value≈2: 1.

  The optimal configuration of small group 2 is the ratio 2: 1 above, and the number of small groups 1 described above per small group 2 is 3 (the number of services per small group 1 is 3, so there are 9 services in terms of the number of services) The number of services that can be accommodated in the subscriber information accommodation apparatus in this example is 66. Divide 66 by 9 to get 3 more. Where 66 services can be stored, only 63 can be stored, and resources cannot be effectively used for 3 services. Therefore, the ratio calculation unit 54 determines sub-optimal first difference value: second difference value = 1: 1 that is divisible as the composition ratio of the small group 2.

  The storage unit 55 performs processing by combining small data within the range of the first difference value of processing amount and large data within the range of the second difference value one by one with the newly determined composition ratio of the small group 2. A small amount group 2 is configured and recorded in the processing amount measuring unit 60. Further, the subscriber information is also configured in the small group 2 with the same composition ratio and stored in the subscriber information storage unit 30.

  FIG. 13 shows the distribution of the processing amount data recorded with the configuration of the small group 2. The standard deviation value of the processing amount data in FIG. 13 is 0.0734, which is an improvement over the standard deviation value (0.0803) when the first small group 1 is grouped (FIG. 9).

  FIG. 14 shows the distribution of the subscriber information amount recorded in the small group 2 configuration. The variation between adjacent small groups 2 is about 10,000, and the standard deviation of the data amount is 0.0714. It has decreased to about 30% of the first variation (FIG. 10). In addition, it can be seen that the variation is improved compared to the standard deviation value when the first small group 1 is grouped.

  As described above, the variation reducing function unit 50 first creates a small group 1 of services from the viewpoint of the processing amount of each service using the composition ratio of the small groups obtained from the processing amount data (FIG. 9). The distribution of the subscriber information amount of the same small group 1 is as shown in FIG. Next, based on the small group 1, a small group 2 is created that reduces the variation in the amount of subscriber information in terms of the amount of subscriber information (FIG. 12). FIG. 13 shows the variation in the processing amount obtained by combining the services in the small group 2 configuration. Further, the variation in the amount of subscriber information is as shown in FIG.

  By operating the variation reducing function unit 50 in this way, the variation in the data amount between the processing amount and the subscriber information amount can be reduced. Note that the above description is an operation in which the subscriber information accommodation apparatus 100 reduces the variation when the accommodation of the subscriber information accommodation system 1000 occurs during system operation.

  The subscriber information accommodation apparatus 100 can also accept the transfer of subscriber information from the service providing apparatus. Even when the service providing apparatus accepts the transfer of subscriber information for which the service has been developed, the variation reducing function unit 50 operates in the same manner to reduce the variation in the data amount between the processing amount and the subscriber information amount. .

〔simulation〕
A simulation was performed on the server reduction effect of the subscriber information accommodation system 1000 constituted by a plurality of subscriber information accommodation devices 100a, 100b, 100c,. The simulation conditions are as follows.

  The number range of subscribers was 1,000 to 100,000, and the server usage rate range was a random number within the range of 10 to 60%. The number of services was 500. The subscriber capacity (resource limit) was 3.4 million, and the processing capacity (resource limit) was 1.3 million sessions.

  FIG. 15A shows the variation in the number of subscribers. FIG. 15B shows the variation in the processing amount. FIG. 16 shows the simulation result. The left column of FIG. 16 is a conventional method, and the right column is this embodiment.

  The number of devices when the subscriber information accommodation system is configured by the conventional method of adding subscriber information accommodation devices when either the accommodation processing amount or the number of accommodated subscribers exceeds the resource limit is nine. . On the other hand, the number of devices when the subscriber information accommodation system 1000 is configured by combining the subscriber information accommodation devices 100 of the present embodiment is eight.

  It can be seen that the subscriber information accommodation apparatuses (accommodation servers in FIG. 16) 1 to 7 of the present embodiment accommodate the number of accommodated subscribers and the accommodation throughput. Thus, according to the subscriber information apparatus of this embodiment, a subscriber information accommodation system can be constructed with a small number of units.

[Second Embodiment]
FIG. 17 shows a functional configuration example of the subscriber information accommodation apparatus 200 according to the second embodiment of the present invention. A subscriber information accommodation apparatus 200 according to the present embodiment is an apparatus that constitutes a subscriber information accommodation system by connecting to the subscriber information accommodation apparatus 100 according to the first embodiment. 100 accommodation states are controlled.

  The subscriber information accommodation apparatus 200 is different from the subscriber information accommodation apparatus 100 in that an accommodation correction control unit 290 is provided. The accommodation correction control unit includes a main pipe unit 291. The main management unit 291 has a system management table in which the processing amount of other subscriber information accommodation devices and the size information of the subscriber information are recorded, acquires a small group from the other subscriber information accommodation devices, and the small size of the own device. The group is transferred to another subscriber information accommodating device.

  The system management table is a table that can list the statuses of the amount of processing and the amount of subscriber information for each resource limit of all the subscriber information accommodation devices constituting the subscriber information accommodation system. By referring to the system management table, a small group of the own device can be appropriately transferred to another subscriber information accommodating device.

  According to the subscriber information accommodation apparatus 200, since the small group is transferred to and from other subscriber information accommodation apparatuses using the system management table, exchange of the small group between the subscriber information accommodation apparatuses is performed without waste. Make it possible.

  As described above, by using the subscriber information accommodation devices 100 and 200, efficient accommodation that can effectively use the resources is enabled.

  Note that the processing amount data is recorded in the processing amount measuring unit 60, but the processing amount data may be stored in the subscriber information storage unit 30 together with the subscriber information. As described above, the present invention is not limited to the above-described embodiment, and various modifications can be made within the scope of the gist thereof.

10: Communication I / F
20: Service providing unit 30: Subscriber information storage unit 40: Subscriber information distribution calculating unit 50: Variation reducing function unit 60: Processing amount measuring unit 70: Processing amount distribution calculating unit 80: Two-dimensional variation correcting control unit 90: Accommodating Correction control unit

Claims (8)

A subscriber information storage unit for storing subscriber information of a plurality of subscribers for each service;
A service providing unit for providing a data operation function of the subscriber information to a service providing device;
A throughput measuring unit that measures a throughput that is the number of subscriber information to be processed per unit time; and
A processing amount distribution calculating unit that calculates a measure of variation in the processing amount of the service for each service;
A subscriber information distribution calculating unit that calculates a measure of variation in the subscriber information amount of the service;
The scale of variation of the processing amount and the scale of variation of the subscriber information amount are monitored, and when the value of any one of the scales exceeds a threshold value, the variation of the processing amount is reduced. A two-dimensional variation correction control unit that causes a variation reducing function unit to create a small group of the service to be reduced and a small group of the service that reduces variation in the subscriber information amount based on the small group;
Analyzing the ranking of each data size of the processing amount and the subscriber information amount, obtaining the change amount of the data size in ascending order of the analyzed order, calculating the value of the center of gravity of the obtained change amount, and calculating the data size A first difference value that is a difference between the minimum value rank and the centroid value, and a second difference value that is a difference between the maximum value rank and the centroid value of the data thirds. The small group combining the small data in the range of the first difference value and the large data in the range of the second difference value with the number of components of the ratio between the difference value and the second difference value, The variation reducing function unit to create the amount and the subscriber information amount;
Accommodation that transfers a part of each data of the small group of the subscriber information amount and the small group of the processing amount of the same configuration as the small group of the subscriber information amount to another subscriber information accommodation device A subscriber information accommodation device comprising: a correction control unit. In the subscriber information accommodation device according to claim 1,
The variation reducing function unit is
The average number of subscribers is calculated and the average value of the processing amount is obtained, and the average number of services that can be accommodated by the own device using the resource limit of the own device and the average value is calculated. And
Obtaining the first difference value and the second difference value, calculating the number of constituents of the small group constituted by the ratio of the first difference value and the second difference value, and calculating the average of the service A ratio that determines whether the accommodation number is divisible by the number of the components, and if it is divisible, determines the number of the components as the composition ratio of the small group, and if not divisible, determines the sub-optimal configuration ratio that is the divisible number of components A calculation unit;
A storage unit configured to store the small group by configuring the small group by combining the data within the range of the first difference value and the data within the range of the second difference value in the configuration ratio. A subscriber information accommodation device as a feature. In the subscriber information accommodation device according to claim 1 or 2,
The variation reducing function unit is
The composition ratio of the small group 2 that is the second small group is determined from the value of the center of gravity obtained from the change amount of the small group of the subscriber information, the first difference value, and the second difference value. ,
Using the composition ratio of the small group 2 to reduce the variation in the processing amount,
The subscriber information accommodation apparatus, wherein variation of the subscriber information is reduced by using the composition ratio of the small group 2. In the subscriber information accommodation device according to any one of claims 1 to 3,
The accommodation correction control unit
Having a system management table that records the processing amount of the other subscriber information accommodation device and the data size of the subscriber information, and refers to the system management table from the other subscriber information accommodation device to A subscriber information accommodation apparatus comprising: a main unit that acquires a part of a group and transfers a part of the small group of the own apparatus to another subscriber information accommodation apparatus. A subscriber information accommodation method performed by a subscriber information accommodation device,
A service providing step of providing a data operation function of subscriber information of a plurality of subscribers stored in the subscriber information storage unit for each service to the service providing device;
A throughput measuring step for measuring a throughput that is the number of subscriber information to be processed per unit time; and
A processing amount distribution calculating step for calculating a measure of variation in the processing amount of the service for each service;
A subscriber information distribution calculating step for calculating a variation measure of the subscriber information amount of the service;
The scale of variation of the processing amount and the scale of variation of the subscriber information amount are monitored, and when the value of any one of the scales exceeds a threshold value, the variation of the processing amount is reduced. A two-dimensional variation correction control step for causing a variation reducing function unit to create a small group of the service to be performed and a small group of the service that reduces variation in the subscriber information amount based on the small group;
Analyzing the ranking of each data size of the processing amount and the subscriber information amount, obtaining the change amount of the data size in ascending order of the analyzed order, calculating the value of the center of gravity of the obtained change amount, and calculating the data size A first difference value that is a difference between the minimum value rank and the centroid value, and a second difference value that is a difference between the maximum value rank and the centroid value of the data thirds. The small group combining the small data in the range of the first difference value and the large data in the range of the second difference value with the number of components of the ratio between the difference value and the second difference value, The variation reducing function step for creating the amount and the amount of subscriber information;
Accommodation correction for transferring a part of each data of the small group of the subscriber information amount and the small group of the processing amount of the same configuration as the small group of the subscriber information amount to another subscriber information accommodation device A subscriber information accommodation method comprising: performing a control step. In the subscriber information accommodation method according to claim 5,
The variation reducing function step includes:
The average number of subscribers is calculated and the average value of the processing amount is obtained, and the average number of services that can be accommodated by the own device using the resource limit of the own device and the average value is calculated. Steps,
Obtaining the first difference value and the second difference value, calculating the number of constituents of the small group constituted by the ratio of the first difference value and the second difference value, and calculating the average of the service A ratio that determines whether the accommodation number is divisible by the number of the components, and if it is divisible, determines the number of the components as the composition ratio of the small group, and if not divisible, determines the sub-optimal configuration ratio that is the divisible number of components A calculation step;
And storing the small group by storing the small group by combining the data within the range of the first difference value and the data within the range of the second difference value by the composition ratio. Subscriber information accommodation method. In the subscriber information accommodation method according to claim 6,
The variation reducing function step includes:
The composition ratio of the small group 2 that is the second small group is determined from the value of the center of gravity obtained from the change amount of the small group of the subscriber information, the first difference value, and the second difference value. ,
Using the composition ratio of the small group 2 to reduce the variation in the processing amount,
The subscriber information accommodation method, wherein variation in the subscriber information is reduced by using the composition ratio of the small group 2. In the subscriber information accommodation method according to any one of claims 5 to 7,
The accommodation correction control step includes:
Having a system management table that records the processing amount of the other subscriber information accommodation device and the data size of the subscriber information, and refers to the system management table from the other subscriber information accommodation device to A subscriber information accommodation method comprising: a main step of acquiring a part of a group and transferring a part of the small group of the own apparatus to another subscriber information accommodation apparatus.

Images