JP2015121924A - Data management system, server device, control method of server device, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To arrange multiple pieces of data which are highly likely to be acquired at a time, on one database system.SOLUTION: A server device receives a request for processing a data group managed in a plurality of database systems from a client device, and executes data processing on the database systems. The server device analyzes a condition, which is to be executed to a data group to be processed and is set in a processing request on the database systems received from the client device, and acquires information for identifying a location of data corresponding to the analyzed condition, from any of data systems. The server device executes data processing operation on the data group in the data systems, on the basis of the acquired identifying information.

Description

  The present invention relates to a data management system, a server device, a server device control method, and a program.

  In a data management system that handles a large amount of data, the data is often distributed and managed in a plurality of database systems, rather than being handled by a single database system. Such an information processing system is called a distributed database system. By using such a distributed database system, it is possible to prevent a concentration of processing from being concentrated, and there is an effect of reducing the processing load of each database system.

Generally, in a distributed database system, how to distribute data is an important factor for improving performance. Patent Document 1 devised a system that distributes data so that the processing load of each database system is equal in consideration of the number of data accesses and the resources (for example, main storage amount) held by the database system. ing.
For example, a larger amount of data is allocated to a database system with more resources than a database system with fewer resources. By dispersing in this way, the processing load becomes uniform and local performance degradation can be prevented.

JP-A-6-259478

In general, when performing some processing (sorting, totaling, etc.) on data obtained from a database system, it is considered that the performance improvement is caused by causing the database system itself to perform those processing.
For example, it is assumed that the data to be processed is distributed in a plurality of database systems when performing the aggregation process. In this case, the application needs to connect to a plurality of databases to acquire data, and the application itself needs to aggregate these data.
As a result, in the information processing system, the performance is greatly degraded. For this reason, in the distributed database system, in order to realize the performance improvement, it is necessary to store the data processed simultaneously by one request in a single database system.

In the distribution method described in Patent Document 1, only the number of accesses and the resources of the database system are taken into consideration, so that data acquired simultaneously in one process is distributed and held in a plurality of database systems, and performance is improved. There is a possibility of deterioration. However, since the data acquired at the same time in one process varies greatly depending on the client issuing the request, it may be very difficult to combine the data acquired at one time into one database system.
The present invention has been made to solve the above-described problems, and an object of the present invention is to provide a mechanism that can arrange data that is highly likely to be acquired at one time in one database system.

The data management system of the present invention that achieves the above object has the following configuration.
A first server device that receives processing requests for data groups managed by a plurality of database systems from a client device and executes data processing for the plurality of database systems, and a second server that operates the data groups of each database system A first data analysis system including a first analysis unit configured to analyze a condition set in a process request for a plurality of database systems received from the client device; The acquisition means for acquiring the information specifying the location of the data corresponding to the condition analyzed by the analysis means of any one of the data systems, and the data of each database system based on the specified information acquired by the acquisition means Execution means for executing data processing operations on the group, This database system is analyzed by the first management means for acquiring the condition history analyzed by the first analysis means from the first server device and managing it using the storage device, and by the first analysis means. And a second management unit that manages information for specifying the location of data corresponding to the condition in the storage device, and the second server device acquires the condition history managed by any database system Second analysis means for analyzing whether or not a plurality of data to be processed simultaneously corresponding to each condition is distributed and arranged in each database system using information for specifying the location, and each condition When a plurality of data to be processed simultaneously corresponding to the above are arranged in each database system in a distributed manner, based on the analysis result by the second analysis means, A data group to be operated simultaneously corresponding to each condition is arranged in any database system using update means for updating information for specifying presence and information for specifying the location updated by the update means. And rearrangement means for performing rearrangement processing as described above.

  According to the present invention, data that is highly likely to be acquired at a time can be arranged in one database system.

It is a figure which shows an example of a data management system. It is a block diagram explaining each function of a data management system. It is a figure which shows the example of a processing result by a data usage frequency analysis part. It is a figure which shows the example of a correlation coefficient calculation result by a data grouping process part. It is a figure which shows an example of a grouping, integration, and a reintegration process result. It is a figure which shows the result example of a decision tree. It is a figure which shows the example of a schema of a query history table. It is a figure which shows the example of a schema of a connection information management table. It is a flowchart explaining the control method of a server apparatus. It is a flowchart explaining the control method of a server apparatus. It is a figure which shows the example of deviation of the number which came out in the grouping process result. It is a flowchart explaining the control method of a server apparatus.

Next, the best mode for carrying out the present invention will be described with reference to the drawings.
<Description of system configuration>
[First Embodiment]
FIG. 1 is a diagram illustrating an example of a data management system according to the present embodiment. In this example, a database system and a WEB server are connected to each other, and a client device (client PC (hereinafter simply referred to as a client)) processes information in a cloud computer system environment. The system according to the present embodiment receives a request for processing for a predetermined data group managed by a plurality of database systems from a client device, and executes data processing for the plurality of database systems, and each database. A data management system including a second server device that operates a data group of the system is taken as an example. Here, the database system can be constructed as an information processing apparatus or a server apparatus including hardware and software for storing and managing a data group as will be described later. In addition, in order to identify the plurality of server apparatuses shown in the present embodiment, the following description will be made of data management processing as a database system.

  In FIG. 1, a client 1, a WEB server 2, a Batch server 3, and database systems 4 and 5 are general computers including a CPU, a RAM, a ROM, an HDD, a network unit, and the like. The client 1 is connected to the web server 2 via the network 6 using a browser or a dedicated application. The WEB server 2 hosts a WEB service compliant with a protocol such as HTTP. In this WEB service, it is possible to receive a query related to data acquisition from the client 1 (for example, to acquire data whose value of the item A is larger than 10), acquire it from the database systems 4 and 5, and return it to the client 1. is there.

FIG. 2 is a block diagram for explaining each function of the data management system shown in FIG. Hereinafter, functions of the processing units of the WEB server 2 and the Batch server 3 will be described.
First, the function of each processing unit of the WEB server 2 will be described.
In FIG. 2, the request receiving unit 21 receives a request from the client 1. The request receiving unit 21 passes the query to the query recording unit 22 in order to parse the query included in the request issued from the client 1. The request here is a processing request such as deletion, update, or creation set for the data group. A query means a condition for limiting data to be processed, such as “a value of a certain item A is larger than 10”. For example, if the request is a deletion and the query is as described above, only the data whose item A value is larger than “10” is deleted from the database systems 4 and 5. The
The query recording unit 22 parses the query and records the result in a query history table 41 of the database system 4 described later. The connection information acquisition unit 23 receives the query analysis result from the query recording unit 22, and specifies information indicating where the data corresponding to the query is stored in the database systems 4 and 5. Obtained from the management table 42. The data operation unit 24 connects to the database systems 4 and 5 based on the connection information received from the connection information acquisition unit 23, executes the data processing operation according to the request and the query, and sends the processing result to the response transmission unit 25. hand over. The response transmission unit 25 returns the processing result to the client 1.

Next, functions of each processing unit of the Batch server 3 will be described.
The Batch server 3 is a server for performing various processes for realizing an optimal data arrangement. In the present embodiment, the WEB server 2 and the Batch server 3 are expressed as separate servers, but there is no problem even if they are operating in the same server.

First, the query history acquisition unit 31 acquires the query history information (condition history information) recorded by the query recording unit 22 in the query history table 41. Further, the data usage frequency analysis unit 32 obtains data existing in the current database corresponding to each query based on the contents of the history passed from the query history acquisition unit 31. As a result, as shown in FIG. 3, which data (d1 (data 301) to d6 (data 306)) corresponds to each query (q1 (query 311) to q6 (query 316)). A map is required.
For example, in FIG. 3, the data corresponding to the query q1 (query 311) means data d1 (data 301) and d3 (data 303).

Next, the data grouping processing unit 33 determines which data and which data are likely to be processed with a single request. Specifically, a correlation coefficient between data as shown in FIG. 4 is obtained using a technique such as collaborative filtering which is a general data analysis technique.
In FIG. 4, for example, the correlation coefficient between data d1 (data 401) and data d2 (data 412) is 0.1, and the correlation coefficient between data d1 (data 401) and data d3 (data 413). Represents 0.8. A higher correlation coefficient indicates that the data is more related.
Based on this correlation coefficient, grouping processing is performed so that data that can be processed simultaneously is arranged in the same database system. An example of grouping processing results is shown in FIG. In this example, those having a correlation coefficient of 0.7 or more are grouped. In the example of FIG. 4, data d3 (data 413) having a correlation coefficient of 0.7 or more with data d1 (data 401) is grouped into group 1 (group 501). By performing the grouping process, a grouped grouping result shown in FIG. 5A is obtained.

  The data integration processing unit 34 performs the group integration processing created by the data grouping processing unit 33. In this process, each group including the same data is integrated. In the example of FIG. 5A, data d3 (data 5012) and (data 5022) are grouped into both group 1 (group 501) and group 2 (502). Therefore, group 1 (group 501) and group 2 (group 502) are integrated. By performing this process, a group as shown in FIG. 5B is created.

  The data reintegration processing unit 35 further integrates the group integrated by the data integration processing unit 34 according to the number of data. Specifically, the data reintegration processing unit 35 integrates the groups created by the data integration processing unit 34 so that the number of groups is the number of divisions N input in advance to the system. At this time, the groups are integrated while considering the number of data so that the number of data included in the N groups is equal. For example, when N is “2”, a group as shown in FIG. 5C is finally created. Note that it is assumed that the value of N is obtained from, for example, the data amount or processing load.

The connection information analysis unit 36 uses the condition that each data is distributed to the group obtained by the reintegration process (for example, the item A having a value greater than “10” is group 1 (group 521), 10 The following obtains the condition of group 2 (group 522).
Specifically, the connection information analysis unit 36 uses an algorithm such as a decision tree, which is one of data analysis methods, to obtain a condition for distributing each data to the group after the reintegration processing. At this time, an example of the decision tree obtained by this processing is shown in FIG.
Items 601 to 605 shown in FIG. 6 correspond to the items of the data tables 43 and 51 shown in FIG. For example, when the value of TimeStamp of item 601 is 2012/7/1 or more (condition 611) and the value of Total (small) (condition 605) is “50” or more (condition 612), it is distributed to group 1 It means that it is done.
In this embodiment, a decision tree is used as an algorithm for obtaining a condition. However, there is no problem even if a condition is obtained using another algorithm. After obtaining the conditions for distributing data, the connection information analysis unit 36 registers the conditions in the connection information management table 42 described later.
In FIG. 2, the data rearrangement processing unit 37 rearranges the data of the database systems 4 and 5 in accordance with the conditions obtained by the connection information analysis unit 36.

Finally, the database systems 4 and 5 will be described.
In FIG. 2, the database system 4 includes a query history table 41, a connection information management table 42, and one or more data tables 43. The database system 5 is composed of one or more data tables 51. Each table will be described below.

  As shown in FIG. 7, the query history table 41 manages what queries are made for the items 711 to 718 of the data tables 43 and 51 for each query ID 701 uniquely assigned to each query. It is a table for. For example, in the query ID q6 (query 707), the item 711 is Tenant ID 711, which is equal to T1, and indicates that the value of total1 is larger than “200”. “Eq”, “Gt”, and “Lt” described in the table are operators that mean “equal”, “greater than”, and “less than”, respectively. The types of operators are not limited to these, and any operator may be used.

Next, the connection information management table 42 will be described with reference to FIG.
First, FIG. 8A shows the connection information management table 42 before data rearrangement is performed by this system.
The connection information management table 42 is a table that holds data for determining in which database system the data corresponding to the query is held based on the query included in the request issued from the client 1. .
The connection information management table 42 is composed of the same number of columns 801 to 808 as the items of the data table 43 and a column 809 indicating connection information to the database system. Before the present invention is implemented, as shown in FIG. 8A, only one line of data is registered in this table, and the values of all items are “*”. . “*” Means “regardless of conditions”. For example, even if conditions for TenantID are described in a query issued from the client 1, they are not considered and it means that a database system is selected using the connection information 811.

Next, an example of a connection information management table after implementing the present invention and rearranging data will be shown using FIG. 8B.
As described above, the conditions obtained from the decision tree obtained by the connection information analysis unit 36 are registered in this table. For example, the condition 812 represents the conditions 611 and 612 in FIG. For example, when a data request for data acquisition is received from the client 1 in which TimeStamp is greater than 2012/8/1 and the value of Total (small) is 100 or more, the connection information acquisition unit 23 selects (b) in FIG. ) Of data 812 is acquired.

Finally, the data table 43 will be described.
The data table 43 is not a table storing management information such as the query history table 41 or the connection information management table 42, but a table for handling actual data. In the case of a distributed database system, as shown in FIG. 2, only this table exists in a plurality of database systems. The user can specify a query as described above for each column (item) of this table.

  A processing flow of request processing from the client 1 in the WEB server 2 and data relocation processing in the Batch server 3 in the present embodiment will be described with reference to FIGS. 9 and 10.

[Process of WEB server 2]
FIG. 9 is a flowchart for explaining a control method of the server apparatus according to the present embodiment. This example is a processing example related to processing of a request from the client 1. Each step is realized by the CPU of the WEB server 2 executing a control program (module shown in FIG. 2) stored in the storage device. Hereinafter, the control entity will be described as the module shown in FIG.
In S <b> 901, the request reception unit 21 receives a request from the client 1, and passes the request content to the query recording unit 22. In S902, the query recording unit 22 determines whether a query is included in the passed request. If the query recording unit 22 determines that a query is included, the process proceeds to S903, and if not included, the process proceeds to S905 when the query recording unit 22 determines.
In S903, the query recording unit 22 parses the passed query and determines what conditions and what operators are used for each data item. In S904, the query recording unit 22 records the analysis result in the query history table 41.

  In S905, the connection information acquisition unit 23 acquires connection information from the connection information management table 42, and which database system holds data corresponding to the query based on the query condition parsed in S903. Is determined. Then, the obtained connection information to the database system is passed to the data operation unit 24.

  In S906, the data operation unit 24 connects to the database system based on the passed connection information, and executes processing according to the request and the query. The processing result is passed to the response transmission unit 25. In S907, the response transmission unit 25 returns the processing result to the client 1.

As described above, by performing the processing from S901 to S907, it is possible to record the contents of the query in the query history table 41 after executing the processing issued according to the request issued from the client and the query. .

[Process of Batch Server 3]
FIG. 10 is a flowchart for explaining a control method of the server apparatus according to the present embodiment. This example is a group integration process example associated with any group when the Batch server 3 executes data rearrangement. Each step is realized by the CPU of the Batch server 3 executing a control program (the module shown in FIG. 2) stored in the storage device. Hereinafter, the control entity will be described as the module shown in FIG. Further, it is assumed that the division number N and the correlation coefficient threshold value are input to the data management system before this flow is executed.
In S <b> 1001, the query history acquisition unit 31 acquires a query history group from the query history table 41.
In S1002, the data usage frequency analysis unit 32 acquires one query from the history group acquired in S1001, and obtains data corresponding to the query.

  In S1003, the data usage frequency analysis unit 32 confirms whether or not the processing of S1002 has been executed for all of the query history groups acquired in S1001, and if processing has been performed for all of them, S1004 If the process transitions to S1002, the process proceeds to S1002.

  In S1004, the data grouping processing unit obtains the correlation coefficient of each data using an algorithm such as collaborative filtering for the processing results of S1002 and S1003. By executing this processing, it is possible to obtain a map storing correlation coefficients between data as shown in FIG.

  In S1005, the data grouping processing unit selects one arbitrary data from the map obtained in S1004, detects data whose correlation coefficient with the data is equal to or greater than a predetermined threshold, and the data Group.

In S1006, the grouping processing unit determines whether or not the processing in S1005 has been performed on all data included in the map obtained in S1004. If it has been performed on all data, the process proceeds to S1007, and if not, the process proceeds to S1005. By performing the processing from S1005 to S1006, a grouping result as shown in FIG. 5A is completed.
In S1007, the data integration processing unit 34 selects an arbitrary group from the groups obtained by the data grouping processing unit 33. In S1008, the data integration processing unit 34 selects one arbitrary data from the group selected in S1007.

In S1009, the data integration processing unit 34 searches for a group including the data selected in S1008 other than the group selected in S1007. If the data integration processing unit 34 determines that it exists, the process proceeds to S1010. If the data integration processing unit 34 determines that it does not exist, the process proceeds to S1011.
In S1010, the data integration processing unit 34 integrates the group detected in S1009 and the group selected in S1007.

  In S1011, the data integration processing unit 34 confirms whether or not the processing of S1008 to S1010 has been performed on all data existing in the group selected in S1007. If so, the process proceeds to S1012. If not, the process proceeds to S1008.

In S1012, the data integration processing unit 34 checks whether or not the processing of S1007 to S1011 has been performed on all the groups obtained by the data grouping processing unit 33. If it is implemented, the process proceeds to S1013. If it is not implemented, the process proceeds to S1007. By performing the processing from S1007 to S1012, a grouping result as shown in FIG. 5B is completed.
In step S <b> 1013, the data reintegration processing unit 35 selects N groups in order from the group with the largest number of data included from all the groups obtained by the data integration processing unit 34.

In S1014, the data reintegration processing unit 35 selects the group having the largest number of data from the groups that are included in all the groups obtained by the data integration processing unit 34 and are not the groups selected in S1013.
In S1015, the data reintegration processing unit 35 compares the number of data included in the N groups selected in S1013. In S1016, the data reintegration processing unit 35 selects a group having the smallest number of data from the comparison result in S1015, and integrates the group obtained in S1014.

  In S1017, the data reintegration processing unit 35 determines whether or not the processing of S1014 to S1016 has been performed on all the groups obtained by the data integration processing unit 34. If the data reintegration processing unit 35 determines that the process is being performed, the process proceeds to S1018. If the data reintegration processing unit 35 determines that the process is not being performed, the process proceeds to S1014. By performing the processing of S1013 to S1017, the grouping result as shown in FIG. 5C is completed.

  In S1018, the connection information analysis unit 36 obtains a decision tree in which data included in the group after reintegration obtained in the processes of S1013 to S1017 is classified into the group after reintegration. By performing this process, a decision tree as shown in FIG. 6 is completed.

In S1019, the connection information analysis unit 36 updates the connection information management table 42 based on the decision tree obtained in S1018. Thereby, a connection information management table as shown in FIG. 8B is completed. In S1020, the data rearrangement processing unit 37 rearranges the data in the database systems 4 and 5 based on the connection information management table obtained in S1019.
As described above, by performing the processing from S1001 to S1020, it is possible to realize data arrangement such that data that are likely to be used at the same time are collected in one database system.
[Second Embodiment]
A second embodiment according to the present invention will be described. Only the differences from the first embodiment will be described below with respect to the configuration.

  In the first embodiment, when the data grouping processing unit 33 performs the grouping process, for example, as shown in FIG. 11A, a large deviation may occur in the number of data included in each group. . At this time, when the integration process and the reintegration process (the number of divisions is set to 2) are performed in order, as shown in FIG. 11B, there is a large deviation in the number of data included in the finally created group. Will occur.

  Specific examples of the above description will be described below. For example, a case where the first embodiment of the present invention is applied to a system having a restriction that data that can be operated by the client 1 is limited to data associated with an organization (tenant) to which the client 1 belongs. Suppose. Here, in order to realize the restriction that “only the data of the tenant to which the user belongs can be manipulated”, the request reception unit 21 may automatically assign a query character string such as “tenantID eq 'XXX'”. Conceivable. This XXX is an ID (tenant ID) indicating the tenant to which the client 1 belongs, and this ID is given to the query by the request receiving unit 21 based on the authentication information of the client 1. In this way, by automatically and forcibly assigning the tenant ID on the system side, it becomes possible to limit the tenant data operated by the client 1 to the tenant to which the client belongs.

  When such a query is given by the request receiving unit 21, the query history table 41 always records a condition related to “tenantID”. As described above, when the grouping process is performed based on the query history table 41 in which “tenantID” is always recorded, data having the same tenant ID is easily grouped. At this time, if the number of data held for each tenant is greatly different, the number of data included in each created group is greatly different as a result of the previous grouping.

In general, one factor that greatly affects the performance of processing such as search and deletion is the total number of data held in the database system. The larger the total number, the longer it takes to search and delete. Therefore, when there is a bias in the number of data held by each database system among a plurality of database systems, the performance is significantly degraded due to the difference in the database system to be operated. This leads to a poor user experience. In general, in any situation, it is important to provide a system to users with equivalent performance. For these reasons, in the distributed database system, it is important that the number of data held in each database system is evenly arranged without being biased.

Therefore, in the second embodiment, an object is to correct the difference in the number of data held by each database system while collecting the data that are likely to be used at the same time into one database system.
In the case of the above specific example, the number of data included in each group of the grouping processing result is compared, and when there is a significant difference in the number of data, the conditions common to the data included in the group are the case of the above example. And the tenant ID are deleted, and the grouping process is performed again. By doing this, there is a condition that must be given to the request, and even if there is a difference in the number of data due to that condition, relocation processing that corrects the difference in the number of data is performed. It becomes possible.

[Process of Batch Server 3]
FIG. 12 is a flowchart for explaining a control method of the server apparatus according to the present embodiment. This example is a processing example when the Batch server 3 executes data rearrangement. Each step is realized by the CPU of the Batch server 3 executing a control program (the module shown in FIG. 2) stored in the storage device. Hereinafter, the control entity will be described as the module shown in FIG. Further, it is assumed that the threshold α is input to the data management system by some means before the flow is executed.

In the present embodiment, the correction processing for the deviation in the number of data is realized by the processing from S1201 to S1205 in FIG. These processes are performed by the data grouping processing unit 33 between S1006 and S1007 shown in FIG. Since the processes of S1001 to S1006 and S1007 to S1020 are not different from those described above, the description thereof is omitted.
In S1201, the data grouping processing unit 33 calculates the number of data existing in all database systems. This total number of data is defined by a variable A. In S1202, the data grouping processing unit 33 obtains the number of data A / N when the data is evenly distributed to N database systems.

  In S1203, the data grouping processing unit 33 determines whether there is a group having the number of data equal to or greater than the value obtained by multiplying the A / N determined in S1202 by the threshold value α in the groups obtained in S1005 to S1006. The purpose of this processing is to determine whether or not there is a group having a number of data that deviates significantly from a state where data is evenly distributed (A / N data exists in each group). If the data grouping processing unit 33 determines that there is a group having a significantly different number of data, the process proceeds to S1204. If the data grouping processing unit 33 determines that there is no group, the process proceeds to S1007.

In S1204, the data grouping processing unit 33 determines whether there is a query issued in common to all data included in the group determined in S1203 from the query history group acquired in S1001. If the data grouping processing unit 33 determines that there is a common condition, the data grouping processing unit 33 transitions to S1205. If the data grouping processing unit 33 determines that there is no common condition, the data grouping processing unit 33 transitions to S1007.
In S1205, the data grouping processing unit 33 deletes the condition obtained in S1204 from the query history group acquired in S1001, and the process proceeds to S1002.
As described above, by performing the processing of S1201 to S1205, it is possible to correct the deviation in the number of data.

  Each process of the present invention can also be realized by executing software (program) acquired via a network or various storage media by a processing device (CPU, processor) such as a personal computer (computer).

  The present invention is not limited to the above embodiment, and various modifications (including organic combinations of the embodiments) are possible based on the spirit of the present invention, and these are excluded from the scope of the present invention. is not.

101 Computer 105 Print Device 302 Print Management Device 304 Device Information Management Device

Claims (8)

A first server device that receives processing requests for data groups managed by a plurality of database systems from a client device and executes data processing for the plurality of database systems, and a second server that operates the data groups of each database system A data management system including a device,
The first server device is
First analysis means for analyzing conditions set in processing requests for a plurality of database systems received from the client device;
Obtaining means for obtaining information identifying the location of data corresponding to the condition analyzed by the first analyzing means from any database system;
An execution means for executing a data processing operation on a data group of each database system based on the information to be specified acquired by the acquisition means;
One of the database systems
First management means for acquiring a condition history analyzed by the first analysis means from the first server device and managing the condition history using a storage device;
And second management means for managing information specifying the location of data corresponding to the condition analyzed by the first analysis means in the storage device,
The second server device is
Information for identifying the location as to whether or not a plurality of data to be processed simultaneously corresponding to each condition is distributed and arranged in each database system by acquiring the condition history managed by any database system Second analysis means for analyzing using,
Update to update the information specifying the location based on the analysis result by the second analysis means when a plurality of data to be processed simultaneously corresponding to each condition is distributed in each database system Means,
Relocation means for performing relocation processing so that a data group to be operated simultaneously corresponding to each condition is arranged in any database system using the information specifying the location updated by the update means; A data management system comprising: The second analyzing means includes
Calculation means for obtaining a correlation coefficient of each data operated simultaneously under each condition using information for specifying the location acquired from any database system;
Grouping means for grouping each data from the value of the correlation coefficient calculated by the calculation means;
The data management system according to claim 1, further comprising an integration unit that integrates groups associated with data selected from the specific group into one group. The second server device is
First determination means for determining whether the data group rearranged by the rearrangement means is biased in any database system;
A second determination means for determining whether or not a condition common to the grouped data group exists when it is determined that the data group is biased in any database system;
When it is determined that a condition common to the grouped data group exists, a deletion unit for deleting from the condition history,
The data management system according to claim 1, further comprising: A server device that receives processing requests for data groups managed by a plurality of database systems from a client device, and executes data processing for a plurality of database systems,
First analysis means for analyzing conditions set in processing requests for a plurality of database systems received from the client device;
Obtaining means for obtaining information identifying the location of data corresponding to the condition analyzed by the first analyzing means from any database system;
An execution unit that executes a data processing operation on a data group of each database system based on the specified information acquired by the acquisition unit. A server device that operates a data group of a plurality of database systems for storing and managing a predetermined data group,
Uses information that identifies the location of whether or not a plurality of data that should be processed simultaneously corresponding to each condition is distributed and arranged in each database system by acquiring condition history information managed by any database system Second analysis means for analyzing
Update to update the information specifying the location based on the analysis result by the second analysis means when a plurality of data to be processed simultaneously corresponding to each condition is distributed in each database system Means,
Relocation means for performing relocation processing so that a data group to be operated simultaneously corresponding to each condition is arranged in any database using the information specifying the location updated by the update means; A server device comprising: A method of controlling a server device that receives a request for processing for a data group managed by a plurality of database systems from a client device and executes data processing on the plurality of database systems,
A first analysis step of analyzing conditions set in processing requests for a plurality of database systems received from the client device;
An acquisition step of acquiring information specifying the location of data corresponding to the condition analyzed by the first analysis step from any database system;
An execution step of executing a data processing operation on a data group of each database system based on the specified information acquired by the acquisition step. A control method for a server device that operates a data group of a plurality of database systems that stores and manages a predetermined data group,
Uses information that identifies the location of whether or not a plurality of data that should be processed simultaneously corresponding to each condition is distributed and arranged in each database system by acquiring condition history information managed by any database system A second analysis step to analyze,
An update for updating the information specifying the location based on the analysis result of the second analysis step when a plurality of data to be processed simultaneously corresponding to each condition is distributed in each database system Process,
A relocation step for performing a relocation process so that a data group to be operated simultaneously corresponding to each condition is arranged in any database system using the information specifying the location updated in the update step; A control method for a server device, comprising: A program for causing a computer to function as each means according to claim 4 or 5.

Images