JP7189753B2 - Data distribution management system, data distribution management support device, and data distribution management method - Google Patents

Description

The present invention relates to a data distribution management system, a data distribution management support device, and a data distribution management method capable of improving the flexibility of data distribution.

In recent years, there has been a growing momentum for open data, in which companies and local governments disclose their own data to other organizations, and in the API economy, which promotes data distribution across organizational boundaries through REST APIs (Representational State Transfer Application Programming Interface). there is

Such data distribution between different industries and companies occupies an important position in the development of new systems that aim to solve complex social issues involving multiple stakeholders by utilizing data that transcends business fields. there is In addition, due to improvements in IoT (Internet of Things) and sensing technology, the quality and quantity of data obtained from data generation terminal devices managed by each company have increased significantly. In response to the above-mentioned trend, there is a demand for distribution while flexibly changing the specifications of data generated by in-house management terminal devices.

Systems that use such data often have a fixed system design that closely depends on the data before specification changes. For this reason, if the data specifications are changed, it will affect the existing and new systems in operation, so the hurdles for changing the data specifications are high.

As a technique related to such problems, there is a data distribution management system disclosed in Patent Document 1. In the data distribution management system disclosed in Patent Document 1, a matching means for matching data specifications provided by a data provider with data specifications desired by a user, and a data distribution contract for which matching has been established. It comprises means for managing and means for distributing data provided by the contract to the user's terminal device. This data distribution management system provides means for matching fixed data specifications presented by data providers and users.

JP 2017-84134 A

However, in Patent Document 1, when the specifications of data generated by a terminal device managed by a data provider (data provided to the data distribution management system) are changed after the distribution of matching data has started, data use There is no disclosure of a method for continuously distributing data requested by users.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and its object is to provide a data distribution management system, a data distribution management support device, and a data distribution management method capable of improving the flexibility of data distribution. be.

In order to achieve the above object, a data distribution management system according to a first aspect includes a management unit for managing data specifications of destinations of terminal data collected from terminal devices, and the destinations managed by the management unit. A processing node that is deployed based on the data specifications of the terminal data and executes processing related to the specification change of the terminal data.

According to the present invention, the flexibility of data distribution can be improved.

FIG. 1 is a block diagram showing the configuration of a data distribution management system according to the first embodiment. FIG. 2 is a diagram showing an example of input data at the time of use of FIG. FIG. 3 is a diagram showing an example of the user information management table of FIG. FIG. 4 is a diagram showing an example of the data schema management table of FIG. FIG. 5 is a diagram showing an example of the node setting management table of FIG. FIG. 6 is a diagram showing an example of the request transfer destination information management table of FIG. FIG. 7 is a flow chart showing the operation of the data distribution management system of FIG. 1 for the data provider. FIG. 8 is a flow chart showing the operation of the data distribution management system of FIG. 1 for data users. FIG. 9 is a flow chart showing the operation of the system management unit in FIG. 1 at the time of registration. FIG. 10 is a flow chart showing the operation of deploying a processing node by the system management unit in FIG. FIG. 11 is a diagram showing an example of a display screen of the console for users in FIG. FIG. 12 is a block diagram showing the configuration of a data distribution management system according to the second embodiment. FIG. 13 is a diagram showing an example of the business index-performance index mapping data of FIG. 14 is a diagram showing an example of the user information management table of FIG. 12. FIG. 15 is a diagram showing an example of the business index calculation logic management table of FIG. 12. FIG. FIG. 16 is a diagram showing an example of the business index-specific node optimum setting management table of FIG. FIG. 17 is a diagram showing an example of the request transfer destination information management table of FIG. FIG. 18 is a flow chart showing the operation at the time of registration of the data distribution system dynamic management unit of FIG. FIG. 19 is a flow chart showing the operation of the data distribution management system of FIG. 12 when changing the performance index. FIG. 20 is a block diagram showing the hardware configuration of the data distribution management system of FIG.

Embodiments will be described with reference to the drawings. In addition, the embodiments described below do not limit the invention according to the claims, and all of the elements described in the embodiments and their combinations are essential to the solution of the invention. Not necessarily.

FIG. 1 is a block diagram showing the configuration of a data distribution management system according to the first embodiment.
In FIG. 1 , the data distribution management system 100 includes a system management section 102 and a system execution section 103 . Data distribution management system 100 is connected to user console 101 , terminal data reception unit 106 , data provider system 107 , and data user system 108 .

When the data distribution management system 100 distributes IoT data collected from the terminal device 105 beyond the framework of the company 171 and the organization 172, the data specification (termed terminal data) generated by the terminal device 105 The data structure of IoT data, distribution methods such as streaming distribution, etc.) are flexibly changed for each data provider and data user and distributed.

The data specification refers to the terminal data transmitted from the terminal device 105, for example, addition or change of items included in the terminal data, data distribution method such as streaming method, and performance requirements such as latency during transmission.

A data provider refers to a company, organization, or individual who owns and manages the terminal device 105 that generates terminal data. In many cases, the data provider is a company that owns the terminal device 105 and collects terminal data for use in its existing business.

A data user refers to a company, an organization, or an individual who receives distribution of terminal data generated by the terminal device 105 managed by the data provider and uses the terminal data.

The user console 101 is an interface for system users to use the data distribution management function provided by the data distribution management system 100 . System users refer to both the above data providers and data users. A system user can use the user console 101 to input input data 109 at the time of use. The use input data 109 can designate data specifications of data provided to system users for each system user.

The terminal device 105 generates terminal data distributed via the data distribution management system 100 . The terminal device 105 is, for example, a mobile terminal such as a smart phone, a sensor installed in factory equipment, or an ambulance that transmits position information and sensing data of a patient's condition to a control center of an ambulance vehicle.

The terminal data reception unit 106 receives input of terminal data generated by the terminal device 105 and transfers it to the system execution unit 103 .

The data provider system 107 is provided with terminal data of the terminal device 105 owned and managed by the company via the data distribution management system 100 . Company A 171 as a data provider uses the data provider system 107 . Company A 171 manages a terminal group 174 that includes multiple terminal devices 105 .

The data user system 108 receives terminal data generated from a terminal device 105 owned and managed by another company via the data distribution management system 100 . A data user, organization B 172 , uses the data user system 108 .

The system management unit 102 manages data specifications of transmission destinations of terminal data collected from the terminal devices 105 . At this time, various information entered by system users, node packages that execute model conversion processing and sampling processing of provided data desired by system users, and parameters required when deploying node packages are managed. or deploy node packages.

The node package is a script group that automates the construction (deployment) of the processing node 131 for the purpose of achieving functional requirements and non-functional requirements related to terminal data specification changes. The node package may include middleware installation/construction processing that configures the processing node 131, a program for controlling execution of the middleware, clustering of computational resources into which the middleware is installed, network settings, and the like. Also, the node package may partially include a data structure conversion processing program executed in the constructed processing node 131 . By executing this script, the processing node 131 for executing model conversion of input request data (terminal data), sampling of streaming data, and the like is constructed and processing is executed. The entity of the processing node 131 may be a VirtualMachine (VM), a Container, or a cloud service type data processing function.

Specific products that implement node packages include Kubernetes developed by Google, which is a container-based IT system configuration management tool, and Ansible developed by Redhat.

The system execution unit 103 accepts terminal data (request data), duplicates and transfers input data to individual node management tenants 123 deployed for each data supply destination system, and processes the transferred data to obtain data. Send it to the destination system. The data provider system is the data provider system 107 and the data user system 108 .

Here, the node management tenant 123 is deployed separately for each system user. Then, the node management tenant 123 can convert the data specifications of one type of terminal data so as to conform to the specifications of data that each system user wishes to receive. As a result, even if the data specifications of the terminal data are changed, the data distribution management system 100 can adapt to the specifications of the data that each system user wants to receive, and the data requested by the system user can be continued. can be distributed effectively.

The configurations and operations of the system management unit 102 and the system execution unit 103 will be described in detail below.

The system management unit 102 includes a system operation API (Application Programming Interface) providing unit 111, a user information management table 112, a data schema management table 113, a node setting management table 115, a node setting mapping unit 116, and a deploy execution 117 and a package management repository 118 .

The system operation API providing unit 111 exchanges necessary information between the company A 171 and the organization B 172 that use the data distribution management system 100 and the data distribution management system 100 through the user console 101, and provides the information to the user. The information management table 112, the data schema management table 113, and the node setting management table 115 are distributed and stored.

The user information management table 112 registers user information input to the system operation API providing unit 111, node setting information of the processing node 131 that performs data processing in the system execution unit 103, and the like.

The data schema management table 113 registers definitions of data schemas of data distributed in the data distribution management system 100 for each system user. For example, the data schema management table 113 can manage the presence or absence of provision and the range of provision partners for each data item included in the content of terminal data. The company A171, which is the data provider, has two data schemas, the data schema included in the terminal data and the data schema of the output data received by the company. Organization B 172, which is a data user, has only the data schema of the output data received from others.

The node setting management table 115 registers, for each node package accumulated in the package management repository 118, meta-information on parameter settings created in advance. For example, the node setting management table 115 stores data formats such as text, image, and video data for data contents, and data delivery formats (sequential delivery, batch data delivery, etc.) and data delivery protocols (HTTP or HTTP) for data provision methods. WebSocket, etc.), and templated node packages can be managed based on various conditions regarding data specifications such as latency, throughput and number of simultaneously connected terminals per performance requirements.

The node configuration mapping unit 116 maps functional and non-functional requirements required by system users to designs of deployable node packages. For example, the node setting mapping unit 116 implements a data model conversion between input and output data according to a data schema, and the deployment script of the processing node 131 for deploying the processing node 131 having the function of sampling the input data. It complements the parameters entered when selecting a node package and deploying a package.

For example, when converting streaming data into batch data, the node setting mapping unit 116 uses data specifications of data provided to system users or information such as batch intervals required for creating batch data based on inputs from data users. can be complemented as needed, or information such as an identifier for identifying data to be sampled and a sampling rate can be complemented as needed. In addition, the node setting mapping unit 116 determines whether structured text data (JSON (Javascript (registered trademark) Object Notation) or CSV (Comma-Separated Values)) is included in the data specification content of the data provided to the system user. In this case, by obtaining both the content definition of the data specification of the data provided to the system user and the content definition of the data specification of the terminal data obtained from the terminal device 105 as input, the data items included in the terminal data can be identified. You can delete or change the item name.

The deployment execution unit 117 receives the node package and parameters calculated by the node setting mapping unit 116 as inputs, and deploys the processing node 131 to the system execution unit 103 . Here, by deploying the processing node 131, a data processing function such as streaming data batch data conversion and sampling, and a data structure conversion function that converts the data structure of terminal data to the data structure of the destination are realized on the gateway. can do.

The package management repository 118 accumulates and manages node packages provided by the package manager 173 .

The system execution unit 103 includes a request transfer destination information management table 121 , a request replication/transfer execution unit 122 , and a node management tenant 123 .

The request transfer destination information management table 121 manages data path information and destination information for accepting terminal data as request data, and address information of the processing node 131 necessary when transferring request data to the processing node 131 .

The request duplicating/transfer executing unit 122 duplicates the terminal data input from the terminal data receiving unit 106 and transfers it to each processing node 131 managed by the node management tenant 123 prepared for each system user. For example, since the node management tenant 123 is separated for each system user, the request replication/transfer execution unit 122 replicates one request data sent from the terminal device 105, to the node management tenant 123 of The request data is transferred to the address of the processing node 131 indicated by the request transfer destination information management table 121 .

The node management tenant 123 operates a processing node 131 prepared for each system user. The node management tenant 123 may be constructed on a public cloud or a private cloud other than the server system on which the system execution unit 103 operates. The node management tenant 123 is a logically divided network environment created for each system user and in which the processing node 131 is stored.

Inside the system execution unit 103 , the domain name of the node management tenant 123 is generated when the node management tenant 123 is created and registered in the request transfer destination information management table 121 . Data communication to the node management tenant 123 may be possible using a tenant name in addition to the domain name. An example of the node management tenant 123 is Amazon Web Service's VPC (Virtual Private Cloud). Node managing tenant 123 provides load balancing functions to processing nodes 131 and routes data input for node managing tenant 123 to processing nodes 131 within node managing tenant 123 .

In addition, if the processing node 131 provides a terminal data sampling function, arbitrary terminal data must be input to a specific processing node 131, but the load balancing function performed by the node management tenant 123 is, for example, , the address information of the terminal device 105 and the data items included in the input data may be used as keys to distribute arbitrary terminal data to appropriate processing nodes 131 .

Node management tenant 123 comprises processing node 131 . Each processing node 131 has a data processing unit 132 and a request data transmission/reception unit 133 .

The processing node 131 is deployed based on the data specification of the transmission destination managed by the system management unit 102, and executes processing related to specification change of terminal data. At this time, the processing node 131 performs processing such as data model conversion, sampling, and micro batching on the request information input via the request replication/transfer execution unit 122 . For example, the processing node 131 converts the data structure of data included in the content of the terminal data received from the terminal device 105 into the data structure defined in the data specification of the data provided to the system user, or converts the data structure into the data structure defined in the data specification of the data provided to the system user. Processing such as batch data conversion of streaming data by data buffering and sampling is executed. The processing node 131 may be provided as a virtual machine (VM), a container, or a function-executing computing resource provided as a cloud service.

Data processing unit 132 provides the ability to execute processing logic such as data model transformation, sampling and micro-batching. Processing logic may be pre-included in a node package. In this case, the data processing unit 132 dynamically reads the input/output data schema included in the parameters at the time of deployment of the processing node 131, a program that converts the data structure, and requests input at the sampling rate specified in the parameters. Perform data sampling, etc.

The request data transmission/reception unit 133 transmits/receives request data processed by the processing node 131 to/from another processing node 131 or transmits the request data to a data supply destination system. In particular, the address information necessary for data transmission may be input at the timing when the parameters at the time of creating the processing node 131 or the information in the request transfer destination information management table 121 are updated.

FIG. 2 is a diagram showing an example of input data at the time of use of FIG.
In FIG. 2, input data 109 at the time of use is input by the system user when starting to use the system. The use input data 109 sets the processing type for terminal data processed in the data distribution management system 100 and the transmission destination information of the processed data.

The input data 109 at the time of use differs depending on whether the input data 201 is input by the data provider or the input data 202 is input by the data user. Data provider input data 201 may include a data schema for input data received from terminal device 105, but data consumer input data 202 does not require this. This is because the data user does not manage the terminal device 105 unlike the data provider.

Both input data 201 , 202 include user information 211 , data schema 212 , distribution information 213 and system requirements 214 .

The user information 211 is information for identifying system users. Identifiers for identifying system users may be organization names such as companies or organizations within companies, or new identifiers made up of alphanumeric characters may be used.

The data schema 212 is composed of a data schema (input data schema) of terminal data and a data schema (output data schema) of transmission (output) data that the system user expects when receiving data from the data distribution management system 100. be. The data schema of output data is information that constitutes part of the data specification of data provided to system users.

The distribution information 213 is data path information when the terminal device 105 inputs request data to the terminal data reception unit 106 . This data path information may be used when identifying terminal data within the data distribution management system 100 . The output data destination is destination information of data processed by the processing node 131 .

The system requirements 214 are information about specifications other than the data model among the data specifications of the request data collected from the terminal device 105 . The functional requirements may include output forms such as sampling, micro-batch conversion, or stream data conversion, and requirements regarding deployment destinations that define execution locations of the processing node 131 . Non-functional requirements may also include performance requirements such as information on the sampling rate for the output form, latency performance until the stream data is transmitted, and expected throughput.

FIG. 3 is a diagram showing an example of the user information management table of FIG.
In FIG. 3 , the user information management table 112 records user information of the data distribution management system 100 .

User information management table 112 includes user organization ID 311 , reference data path 312 , output data destination 313 , data manager flag 314 , schema ID 315 , node design pattern information 316 , and tenant name 317 .

The user organization ID 311 is identifier information assigned to each company, organization, department or individual using the data distribution management system 100 . A user organization ID 311 is assigned to each row 321 to 323 of the user information management table 112 . The user organization ID 311 may be created from the user information 211 of the input data 109 at the time of use, or may be assigned from the data distribution management system 100 .

The reference data path 312 is an endpoint identifier used when the terminal device 105 transmits data to the transmission data reception unit 106, and corresponds to a resource name in REST API. The reference data path 312 is determined by the use input data 109 input by the data provider, but the data distribution management system 100 may determine a unique one.

The output data destination 313 is the address information of the system that receives the data processed by the data distribution management system 100 in each of the data provider and the data user. This is assumed to be defined in the input data 109 at the time of use, but may be changeable from the user console 101 by the administrator of the system to which the data is sent.

The data manager flag 314 is information for identifying a data provider or a data user.

The schema ID 315 is information defined in the data schema management table 113 for each system user. For the data provider, the schema ID 315 defines the schema of data input from the managed terminal device 105 and the data schema of data (output data) transmitted from the data distribution management system 100 . For data users, only the data schema of data (output data) transmitted from the data distribution management system 100 is defined.

The node design pattern information 316 manages design information of the processing nodes 131 deployed in the node management tenant 123 prepared for each system user. The node design pattern information 316 is not directly input by the system user, but added by the node setting mapping unit 116 based on the input data 109 at the time of use.

The tenant name 317 is information for identifying the node management tenant 123 to which the processing node 131 is deployed. The tenant name 317 is not directly input by the system user, but is automatically input when the node management tenant 123 is created in the system execution unit 103 by the deployment execution unit 117 or the like.

FIG. 4 is a diagram showing an example of the data schema management table of FIG.
In FIG. 4, the data schema management table 113 manages input/output data schema information generated for each system user. The data schema management table 113 includes a schema ID 411 , an input data schema 412 and an output data schema 413 . The data schema management table 113 may contain schema names.

The input data schema 412 may be included in the input data 109 input by the data provider, or manually or automatically by the data provider providing a sample of terminal data to the data distribution management system 100. may be made. The output data schema 413 may be provided by the data provider together with the input data 109 at the time of use, or may be changed by the system user through the user console 101 .

The schema ID 411 is an identifier for each row 421, 422 of the data schema management table 113, and given by the data distribution management system 100. FIG. The schema ID 411 is the same as the schema ID 315 managed by the user information management table 112 .

The input data schema 412 is a schema of data generated from the terminal device 105 provided by the data provider and input to the data distribution management system 100 . In the input data schema 412, the presence or absence of provision and the range of provision partners can be set for each data item included in the content provided from the data provider to the data user. Since it is assumed that the input data schema 412 covers the items included in the terminal data to be input, input data samples other than those defined in the input data 109 at the time of use are received from the data provider. may be supplemented by However, the input data schema 412 also has a function to prevent each data item from being distributed to anyone other than the data provider. may be added. A GUI (Graphical User Interface)-based or CUI (Character User Interface)-based interface may be provided for the data provider. Further, the control of the data distribution range may be finely controlled by providing a role for each industry or each company group, for example.

The output data schema 413 is information that defines the data structure of data provided to system users. As with the input data schema 412, the output data schema 413 may be provided with a GUI-based or CUI-based interface for the purpose of defining the structure of the data that the system user receives.

FIG. 5 is a diagram showing an example of the node setting management table of FIG.
5, the node management table 115 includes pattern IDs 511, processing targets 512, and node design patterns 513. FIG. The processing target 512 includes a processing type 514 , a data type 515 and a data update interval 516 . A node design pattern 513 includes a node package 513 , a parameter 518 and a deployment destination 519 .

A pattern ID 511 is an identifier for each processing node pattern managed by the node management table 115 . The pattern ID 511 is assigned to each row 521-526 of the node management table 115. FIG. The pattern ID 511 may be assigned by the data distribution management system 100 .

The processing target 512 is combination information of a processing type 514, a data type 515, and a data update interval 516 for data input to the data distribution management system 100, which is targeted by individual patterns. The processing target 512 is information for distinguishing differences due to the format and processing content of data processed by the processing node 131 and linking this distinction with the design pattern of the processing node 131 .

The node design pattern 513 is information combining a node package 517 indicating a combination of middleware and services included in the processing node 131, parameter information 518 required for execution control of the node package, and a deployment destination 519 of the node package. The node design pattern 513 can specifically determine which node package to use and what parameters to give to control the behavior of the processing node 131 .

The processing type 514 is information indicating the content of processing implemented by the processing node 131 . For example, when a system user requests data model conversion and data sampling, the node management tenant 123 deploys two processing nodes 131 in the node management tenant 123 and sequentially processes the request data. processing. The ordering of the processing nodes 131 will be described later.

The data type 515 is information indicating a data format targeted by processing executed by the processing node 131 . The data type 515 is used to appropriately select packages containing different middleware when different middleware is used in the processing node 131 due to differences in data formats (eg, text data and image data). can be used.

The data update interval 516 is reference information as to how often the processing node 131 executes the data processing when processing the request data. The data update interval 516 affects the time until the data input to the data distribution management system 100 is transmitted to the system user's system. For example, the data update interval 516 is information that can be regarded as an indicator of non-functional requirements such as data processing latency.

The processing target 512 includes information on performance requirements such as data delivery format (sequential delivery, batch data delivery, etc.), data delivery protocol (HTTP, WebSocket, etc.), and performance requirements such as latency, throughput, and number of simultaneously connected terminals. may contain.

The node package 517 is information specifying a package used when deploying the processing node 131 and is a value for referring to the node package accumulated in the package management repository 118 .

A parameter 518 is meta information of a parameter input when deploying a node package. Parameters 518 control what processing is performed by the deployed node package (processing node 131). For example, row 521 of the node management table 115 is a processing node that performs model conversion of terminal data input in JSON format. must be set appropriately. In the example of row 521, a reference data path and an output data schema ID are shown as meta information of parameters indicating this.

The deployment destination 519 is information indicating the environment for deploying the node package. The deployment destination 519 may be an environment in which the data distribution management system 100 operates, an on-premises environment of the system user, or a public cloud. In addition, although the deployment destination 519 is shown fixedly in FIG. 5, the deployment destination 519 of the processing node 131 may be dynamically changed according to the request of the system user.

FIG. 6 is a diagram showing an example of the request transfer destination information management table of FIG.
6, the request transfer destination information management table 121 includes a reference data path 611, a tenant name 612, an output data transmission destination 613, and a processing node address 614. FIG.

The reference data path 611 is information for identifying terminal data that has arrived at the request copy/transfer execution unit 122 from the terminal device 105 via the request data reception unit 106 .

The tenant name 612 is the identification information of the node management tenant 123 created inside the system execution unit 103 or the transfer destination address information of the request data.

The output data transmission destination 613 is address information to which the terminal data processed by the processing node 131 inside the node management tenant 123 is finally transmitted. The output data transmission destination 613 is input by the system user with the use input data 109 or the like, and is set via the deploy execution unit 117 and the user information management table 112 .

The processing node address 614 is address information of the node package deployed within the node management tenant 123 . The processing node address 614 is automatically stored by the node deployment executing unit 117 or the like after the package is deployed. The processing node address 614 is assigned to each row 621-625 of the request transfer destination information management table 121. FIG.

Here, for one reference data path 611, the output data destination 613 of the data provider system 107=sys-a. com and the output data destination 613 of the data user system 108=sys-b. com in the request transfer destination information management table 121, the terminal data collected from the terminal device 105 can be input to the node management tenants 131 provided separately for the company A 171 and the organization B 172. Even if the data specifications of the data are changed, it is possible to adapt the specifications of the data that each system user wants to receive.

FIG. 7 is a flow chart showing the operation of the data distribution management system of FIG. 1 for the data provider.
In FIG. 7, in S710, the company A171 in FIG. 1 inputs the use input data 201 in FIG.

S720 is executed by the system operation API providing unit 111 in response to the input of S710. In S720, the system operation API providing unit 111 executes S721 to S723.

In S<b>721 , the system operation API providing unit 111 registers the user information 211 and the distribution information 213 in the user input data 109 in the user information management table 112 .

In S<b>722 , the system operation API providing unit 111 registers the data schema 212 of the use input data 109 in the data schema management table 113 .

In S<b>723 , the system operation API providing unit 111 inputs the data schema 212 and the system requirements 214 of the input data 109 at the time of use to the node setting management table 115 .

In S<b>730 , the node setting mapping unit 116 deploys the processing node 131 based on the system requirements 214 input from the system operation API providing unit 111 . In S730, the node setting mapping unit 116 executes S731 to S734.

In S731, the node setting mapping unit 116 selects from the node setting management table 115 a node package necessary for conversion of data specifications desired by the system user. At this time, the node setting mapping unit 116 focuses on the conversion (data model conversion) of the input/output data schema 212 in FIG. select.

In S732, the node setting mapping unit 116 complements the parameters 518 (reference data path and input/output data schema) of the package 517 of FIG. Register in parameters.

In S733, the node setting mapping unit 116 adds the package 517 that satisfies the conditions such as the processing type 514 in FIG. Choose one or more from

In selecting the package 517, in addition to the processing type 514, conditions such as the data type 515 of the processing target 512 and the data update interval 516 (such as the latency of the non-functional requirements of the system requirements 214 in FIG. 2) can be used.

Regarding the selection of packages 517, if more than one of "data model conversion", "sampling" and "micro-batching" is required, the package selection of "data model conversion" comes first as indicated at S730. It is assumed that it will be performed in the order of "sampling" and "micro-batch". The execution order of the processing nodes 131 may be defined as a separate rule. The processing type 512 may include types other than "data model conversion," "sampling," and "micro-batching."

In S734, the node setting mapping unit 116 supplements the parameters required by the node package selected in S733 with the system requirements 214 of FIG. 2, and adds them to the node design pattern information 316 of the user information management table 112 of FIG. Input to the deploy execution unit 117 . Since the parameter 518 of the node setting management table 115 is meta information of the parameter to be set when deploying the node package, it needs to be complemented. S734 is repeated for all node packages selected in S733.

In S950, the deployment execution unit 117 executes the node management tenant 123 and the processing node 131 according to the node design pattern information 316 of the user information management table 112 registered or updated by the series of processing S730 of the node setting mapping unit 116. Deploy to unit 103 . After that, the deploy execution unit 117 registers information necessary for transferring terminal data to the node management tenant 123 and the processing node 131 in the request transfer destination information management table 121 .

Examples of information necessary for transferring terminal data include the tenant name 612 and processing node address 614 in FIG. 6 managed by the request transfer destination information management table 121 . In the system execution unit 103, the transfer destination of these pieces of information may be uniquely specified by DNS (Domain Name System), which is a name resolution method of the IT system. In addition, as information for re-transferring the transfer data processed within the node management tenant 123 in FIG. 1 to the system provided by the system user, the output data destination 613 in FIG. 112 output data destinations 313 may be used to supplement.

In S741, the company A171 in FIG. 1 sets access information (reference data path) to the data reception unit 106 as a data transmission destination of the terminal device 105 included in the terminal group 174 managed by the company A171.

In S<b>751 , the terminal data reception unit 106 inputs the terminal data from the terminal device 105 to the request duplication/transfer execution unit 122 .

In S<b>752 , the request replication/transfer execution unit 122 transfers the terminal data (request data) to the node management tenant 123 according to the information in the request transfer destination information management table 121 .

In S753, the node management tenant 123 transfers the terminal data (request data) to the processing node 131, processes the terminal data so that it conforms to the data specifications of the data provider, and then sends the processed data to the data provider. Send and return to S751.

FIG. 8 is a flow chart showing the operation of the data distribution management system of FIG. 1 for data users.
In FIG. 8, S720, S730, S950, S752 and S753 of FIG. 7 are common in the case of use by the data user.

In S810, the organization B 172 in FIG. 1 inputs the use input data 202 in FIG.

Next, the system operation API providing unit 111 executes S720, the node setting mapping unit 116 executes S730, and the deploy executing unit 117 executes S950.

In S841, the terminal device 105 managed by the company A171 transmits terminal data to the terminal data reception unit 106 sequentially or periodically.

In S<b>842 , the request replication/transfer execution unit 122 transfers the terminal data (request data) to the node management tenant 123 according to the information in the request transfer destination information management table 121 .

In S843, the node management tenant 123 transfers the terminal data (request data) to the processing node 131, processes the terminal data so that it conforms to the data specifications of the data provider, and then sends the processed data to the data provider. Send and return to S841.

FIG. 9 is a flow chart showing the operation of the system management unit in FIG. 1 at the time of registration.
9, at S911, the package management repository 118 of FIG. 1 receives additions, deletions and updates of node packages by the package manager 173. In FIG.

In S912, the package manager 174 registers in the node setting management table 115 the functional requirements and non-functional requirements of the package and the parameters necessary for request data processing by the processing node 131 created by deploying the package.

FIG. 10 is a flow chart showing the operation of deploying a processing node by the system management unit in FIG.
In FIG. 10 , in S921, the system operation API providing unit 111 inputs the data schema 212 input by the system user and the system requirements 214 (functional/non-functional requirement information) to the node setting mapping unit 116 .

Next, the node setting mapping unit 116 executes S730 in FIG.

Next, in S<b>941 , the node setting mapping unit 116 inputs the target row data of the updated user information management table 112 to the deploy execution unit 117 .

Next, the deployment executing unit 117 executes S950. In S950, the deployment executing unit 117 executes S951 to S953.

In S<b>951 , the deploy execution unit 117 installs the node management tenant 123 that receives data input from the request replication/transfer execution unit 122 in the system execution unit 103 . A node management tenant 123 is a logical network environment in which processing nodes 131 such as VMs, containers, and public cloud services can be installed and executed.

In S952, the deployment executing unit 117 installs the processing node 131 in the node management tenant 123 based on the information described in the node design pattern information 316 of FIG.

In S<b>953 , the deploy execution unit 117 registers the destination information (such as addresses) of the processing node 131 and the node management tenant 123 in the request transfer destination information management table 121 .

In S<b>961 , the request replication/transfer execution unit 122 reads the information in the request data transfer destination information management table 121 and caches information for transferring terminal data to the node management tenant 123 and processing node 131 .

FIG. 11 is a diagram showing an example of a display screen of the console for users in FIG.
In FIG. 11, the user console 101 of FIG. 1 displays a display screen 1011 for managing terminal data by a data provider and a display screen 1021 for managing specification of data provided by a data user.

The display screen 1011 includes a user information display 1012, a collected data list 1013, a used data list 1014, address information 1015 of the terminal data reception unit 106 including the data path of the collected data, and an input data scheme management screen 1016. .

The user information display 1012 shows the information of the user currently using the management system, and for example, the user organization ID 311 of FIG. 3 may be displayed.

The collected data list 1013 shows a list of collected data from the terminal device 105 managed by itself, and the registered reference data path 312 may be shown. By clicking and selecting the reference data path 312, the address information 1015 of the terminal data reception unit 106 and the management screen 1016 of the input data scheme may be appropriately switched to those related to the corresponding data.

The used data list 1014 shows a list of data provided by the company or others, and may show the reference data path 312 of each data. Similar to the collected data list 1013 , by selecting the reference data path 312 , the display screen 1021 may be changed to a screen for managing data specifications of data to be used, as will be described later.

The address information 1015 of the terminal data reception unit 106 indicates the address, URL information, etc. of the terminal data reception unit 106 that receives the selected data (data generated by the terminal device 105 managed by the company).

The input data scheme management screen 1016 shows the data structure of the terminal data. Data column names and type names may be automatically completed by entering a sample of terminal data. The management screen 1016 may be able to change whether or not to disclose individual data items when providing data to other data users.

The display screen 1021 includes a user information display 1012 , a collection data list 1013 , a utilization data list 1022 , a data utilization method setting screen 1023 and a data structure setting screen 1024 .

Usage data list 1022 provides similar information as usage data list 1014 . However, the output data schema can be flexibly changed by the system user, and it may be possible to create a plurality of output data schemas for each data and select and use them as appropriate. The available data list 1022 may display a list of available (output) data schemas, or may indicate which (output) data schemas are applied, such as tick marks.

On the data usage method setting screen 1023, it is possible to set the sampling rate among the data specifications of the usage data. The data usage method setting screen 1023 may indicate an original data collection method (data generation frequency, etc.) of terminal data provided by others. Based on this information, data users can set their own desired data utilization method (sampling rate and collection interval of microbatch data).

The data structure setting screen 1024 may display data items made public by the data provider if the data provider is other than the company itself, or display all data items to be collected if the data provider is the data provider itself. On the data structure setting screen 1024, when the management system 100 provides data to the data provider, it is possible to set the necessity of data items to be included in the output data.

Although FIG. 1 shows an example in which the system management unit 102 and the system execution unit 103 are provided in the data distribution management system 100, the system management unit 102 and the system execution unit 103 are arranged in a distributed manner. 102 and system execution unit 103 may be connected via a network. For example, the system management unit 102 may be commonly provided for a plurality of system users, and the system execution unit 103 may be provided separately for each system user.

FIG. 12 is a block diagram showing the configuration of a data distribution management system according to the second embodiment.
In FIG. 12, the data distribution management system 100B includes a system management section 102B and a system execution section 103B. The data distribution management system 100B is connected to the user console 101, the terminal data reception unit 106, the data provider system 107, and the data user system 108. FIG.

When the data distribution management system 100B distributes IoT data collected from the terminal device 105 beyond the framework of the company 171 or the organization 172, the data distribution management system 100B sets the data specification of the terminal data generated by the terminal device 105 to the data provider and the data provider. It is flexibly changed for each user and distributed. At this time, the data distribution management system 100B can change the data specification of the terminal data received by the data user via any reference data path, according to the content of the terminal data.

Here, the system user can use the user console 101 to input the usage input data 109 and the business index-performance index mapping data 1101 . The business index-performance index mapping data 1101 is used to associate the terminal data received from the terminal device 105 with the business index unique to the system user, and to associate the business index with the system requirements 214 of FIG. This is reference information.

The system management unit 102B is necessary when deploying node packages that execute various information input by system users, model conversion processing and sampling processing of provided data desired by system users, and node packages. Manage parameters and deploy node packages. At this time, the system management unit 102B sequentially monitors the status of each terminal device 105, associates changes in the terminal data received from the terminal devices 105 with specific business indicators, and selects specific terminal devices according to changes in the business indicators. Data specification changes can be dynamically given to the terminal data from 105 .

The system execution unit 103B accepts terminal data, duplicates and transfers the input data to individual node management tenants 123B deployed according to the content of the terminal data for each data supply destination system, and processes the transferred data. and send it to the data provider system.

Here, the node management tenant 123B is deployed as a separate entity according to the content of terminal data for each system user. Then, the node management tenant 123B changes the data specifications of the terminal data according to the state of the terminal device 105 that outputs the terminal data so as to conform to the specifications of the data that each system user wants to receive. can be done. As a result, even if the data specifications of the terminal data are changed, the data distribution management system 100B can adapt to the specifications of the data that each system user wants to receive, and can continue to provide the data requested by the data user. can be distributed effectively.

The system management unit 102B includes a data distribution system dynamic management unit 1120 in addition to the configuration of the system management unit 102 in FIG. The system management unit 102B has a user information management table 112B instead of the user information management table 112 of FIG. The data distribution system dynamic management unit 1120 includes a business index calculation logic management table 1121 , a business index-specific node optimum setting management table 1122 , a business index extraction unit 1123 , and a target performance index calculation unit 1124 .

The user information management table 112B manages the use input data 109 and the business index-performance index mapping data 1101 for each user.

The data distribution system dynamic management unit 1120 calculates the correspondence between the received data defined in the business index-performance index mapping data 1101 and the business index for the terminal data received from the terminal device 105, and detects changes in the business index. In this case, the deployment instruction of the processing node 131B and the transfer rule of the terminal data are changed.

The data distribution system dynamic management unit 1120 includes a business index calculation logic management table 1121 , a business index-specific node optimum setting management table 1122 , a business index extraction unit 1123 , and a target performance index calculation unit 1124 .

The business index calculation logic management table 1121 manages logic for converting terminal data received from the terminal device 105 into business indices specific to the data user. As for the information managed by the business index calculation logic management table 1121 , part of the information defined in the business index-performance index mapping data 1101 may be stored by the system operation API providing unit 111 .

The business index-specific node optimum setting management table 1122 manages the package and parameters of the processing node 131B deployed in the node management tenant 123B for each business index unique to the data user. Information calculated by the node setting mapping unit 116 may be stored as the information related to the deployment of the processing node 131B in the node management tenant 123B among the information managed by the business index-specific node optimum setting management table 1122 .

The business index extraction unit 1123 extracts terminal data received from the terminal device 105 based on the data transferred from the request replication/transfer execution unit 122 and the logic information managed in the business index calculation logic management table 1121. Calculates business metrics specific to a person.

The target performance index calculation unit 1124 calculates changes in the data user's business index based on the business index unique to the data user calculated by the business index extraction unit 1123 and the information in the node optimum setting management table for each business index 1122. , determination of the node management tenant 123B to which the terminal data should be transferred, change of transfer information of the data received from the terminal device 105, and the like are performed.

The system execution unit 103B includes a request transfer destination information management table 121B, a request replication/transfer execution unit 122, and a node management tenant 123B.

The node management tenant 123B operates a processing node 131B prepared for each system user according to the contents of terminal data. The node management tenant 123B is a logically divided network environment that is created for each system user and for each content of terminal data and stores the processing node 131B.

Node management tenant 123B comprises processing node 131B. Each processing node 131B includes a data processing section 132B and a request data transmission/reception section 133B.

The processing node 131B performs processing such as data model conversion, sampling, and micro batching on request information input via the request replication/transfer execution unit 122 .

Data processing unit 132B provides the ability to execute processing logic such as data model transformation, sampling and micro-batching.

The request data transmitting/receiving unit 133B transmits/receives the request data processed by the processing node 131B to/from another processing node 131B or transmits the request data to the data providing destination system.

In the following, regarding the operation of the data distribution management system 100B in FIG. 12, a use case is assumed in which the heart rate and electrocardiogram information regarding the state of an emergency patient collected by an ambulance is notified to the emergency patient management system of the hospital to which the emergency patient is scheduled to be transported. explain. In this use case, an ambulance that transmits emergency status information can be regarded as one terminal device 105, and the emergency management system can be regarded as a data supply destination system.

In the data distribution management system 100 of FIG. 1, the data specifications to be received by the emergency patient management system can be set by the administrator of the emergency patient management system from the user console 101. FIG. However, in the data distribution management system 100 of FIG. 1, one data specification is applied to each reference data path, so the emergency management system can respond Accordingly, the data specification cannot be dynamically changed only for the terminal device 105 whose terminal data has changed.

On the other hand, the data distribution management system 100B in FIG. 12 sequentially monitors the status of each terminal device 105, and detects changes in the terminal data received from the terminal devices 105 as specific business indicators such as "emergency patient status changes". , and the data specification can be dynamically changed for terminal data from a specific terminal device 105 according to changes in the business index.

FIG. 13 is a diagram showing an example of the business index-performance index mapping data of FIG.
In FIG. 13, data user business index-performance index mapping data 1101 includes user information 211 , business index 1211 , calculation formula 1212 , and system requirements by business index 1213 . The user information 211 is the same as the contents of FIG.

The business index 1211 is information unique to the data user. The work index 1211 is, for example, an index parameter such as "urgency of emergency state" in the case of an emergency state notification system during ambulance transportation. The data distribution management system 100B of FIG. 12 changes the specification of the data provided to the data user according to the change in the business index 1211. FIG. The business index 1211 may include identification information of data managing the business index 1211 (terminal data identifier 1411 described later).

A change in the business index 1211 may differ for each ambulance vehicle (for each terminal device 105) like the emergency patient information notification system. In other words, most ambulance vehicles have a low "emergency level of urgency", and there are cases where the urgency level is high only for a specific emergency case of an ambulance vehicle. In such a case, the data distribution management system 100B identifies an ambulance (terminal device 105) containing information on a specific emergency, and identifies the node management tenant 123B to which the terminal data of the majority of the terminal devices 105 is transferred. Forward specific emergency information to another node management tenant 123 . In the example of FIG. 13, the terminal data identifier 1411 is patientInfo. In the above-described emergency patient information notification system, changes in the work index 1211 are managed for each emergency patient name.

The calculation formula 1212 is information relating to a formula for calculating the business index 1211 . For example, the business index 1211 is calculated from the data of the terminal device 105 used by the data user. The calculation formula 1212 includes, in addition to the calculation formula for the business index 1211, a reference data path for identifying terminal data and a calculation use data item as information specifying data used for calculation. The information of the calculation formula 1212 of the business index-performance index mapping data 1101 is set in the calculation formula 1413, the reference data path 312, and the calculation use data item 1412 of the business index calculation logic management table 1121 of FIG.

The system requirements by business index 1213 are information in which system requirements required by data users are defined for each index parameter of the business index 1211 . In the system requirements by business index 1213, information similar to the system requirements 214 of the input data 109 at the time of use may be individually defined for each index parameter.

Further, the business index-performance index mapping data 1101 may be created by the data user himself or may be created by the service providing the data distribution management system 100B through SI (System Integration) for the data user 172. good.

14 is a diagram showing an example of the user information management table of FIG. 12. FIG.
14, the user information management table 112B has a user organization ID 311, a reference data path 312, an output data transmission destination 313, a data administrator flag 314, a business index ID 1311, and a tenant name 1312. The user organization ID 311, reference data path 312, output data destination 313, and data manager flag 314 are the same as those shown in FIG. The user organization ID 311 of the user information management table 112B is assigned to each row 1321-1323.

The business index ID 1311 is identifier information for referring to the business index-performance index mapping data 1101 in FIG. The business index ID 1311 is given in the business index calculation logic management table 1121 of FIG. 12 for each business index-performance index mapping data 1101 .

The tenant name 1312 is identification information of the node management tenant 123B that processes the data of the reference data path 312 provided to the data user. Since multiple node management tenants 123 may be deployed to generate provided data with different data specifications for the business index 1211 unique to the data user, one or more tenant names 1312 are managed. .

15 is a diagram showing an example of the business index calculation logic management table of FIG. 12. FIG.
15, the business index calculation logic management table 1121 includes a business index ID 1311, a reference data path 312, a user organization ID 311, a terminal data identifier 1411, a calculation use data item 1412, and a calculation formula 1413. The business index ID 1311 is the same as the business index ID 1311 of the user information management table 112B of FIG. The business index ID 1311 of the business index calculation logic management table 1121 is assigned to each row 1412 of the business index calculation logic management table 1121 . The user organization ID 311 and reference data path 312 are the same as the user organization ID 311 and reference data path 312 of the user information management table 112 in FIG. 3, respectively.

The terminal data identifier 1411 is identification information of data managing changes in the business index 1211 in FIG.

The calculation use data item 1412 is information included in the calculation formula 1212 in FIG. 13, and is information specifying an arbitrary data item in the data of the terminal device 105 for calculating the business index. Although one data item is specified in the example of FIG. 15, a plurality of data items may be specified.

The calculation formula 1413 is information included in the calculation formula 1212 in FIG. 13, and is formula information for calculating the business index using the data item specified in the calculation use data item 1412 . The calculation formula 1413 in FIG. 15 can be represented by, for example, a graph 1431B.

FIG. 16 is a diagram showing an example of the business index-specific node optimum setting management table of FIG.
In FIG. 16 , the business index-specific node optimum setting management table 1122 includes a business index ID 1311 , a business index 1511 , a performance index 1512 , a schema ID 411 , and node design pattern information 316 . The node design pattern information 316 is the same as the node design pattern information 316 of the user information management table 112 of FIG. The schema ID 411 is the same as the schema ID 411 of the data schema management table 113 of FIG. The business index ID 1311 is the same as the business index ID 1311 of the user information management table 112B of FIG.

The business index 1511 indicates possible values of the business index unique to the data user. In FIG. 16, the business index 1511 is a discrete value such as "large", "medium", and "small", but may be a continuous value such as "urgency=a (0≦a)".

The performance index 1512 is the value of the performance index corresponding to the value of the business index 1511 . A performance index can indicate the performance requirements of the data provision. As shown in FIG. 16, the values of the business index 1511 and the performance index 1512 may take discrete values that seem to have little relationship, or may take continuous values such as "data update interval δ=ax+b".

FIG. 17 is a diagram showing an example of the request transfer destination information management table of FIG.
17, the request transfer destination information management table 121B includes a reference data path 611, an output data destination 613, a tenant name 612, a rule application order 1611, a distribution rule 1612, a business index 1613, and a processing node address. 614. The reference data path 611, output data destination 613, tenant name 612, and processing node address 614 are the same as the contents of the request transfer destination information management table 121 in FIG.

The rule application order 1611 indicates the order in which the request data sorting rules 1612 are applied to each tenant name 612 indicating the transfer destination of the request data. In FIG. 17, the work index 1613 specified in advance by the data user (administrator of the output data destination 613) changes to "urgency = high", and some request data (PatientInfo.name shown in the sorting rule 1612) = (a, b, c) data) is required to be transferred to the tenant name 612 "SysB2".

The distribution rule 1612 defines a rule for identifying request data when distributing the request data to the tenant name 612 . The business index 1613 is a value that indicates the business index corresponding to the process performed by the node management tenant 123B specified by the tenant name 612. FIG.

Here, for example, the output data destination 613 of one data user system 108=sys-b. com, processing node address corresponding to urgency=low=b1. local and processing node address corresponding to urgency=high=b2. By setting local in the request transfer destination information management table 121B, the terminal data collected from the terminal device 105 can be input to the node management tenant 131B separately provided according to the business index 1613. Performance indicators 1512 can be varied accordingly 1613 .

FIG. 18 is a flow chart showing the operation at the time of registration of the data distribution system dynamic management unit of FIG.
In FIG. 18, in S1711, the system operation API providing unit 111 receives input of the business index-performance index mapping data 1101 from the data user.

In S 1712 , the system operation API providing unit 111 inputs the business index 1211 of the business index-performance index mapping data 1101 to the data distribution system dynamic management unit 1120 .

In S<b>1713 , the data distribution system dynamic management unit 1120 registers the business index 1211 and the calculation formula 1212 in the business index calculation logic management table 1121 .

FIG. 19 is a flow chart showing the operation of the data distribution management system of FIG. 12 when changing the performance index.
19, in S751, the terminal data reception unit 106 inputs the terminal data from the terminal device 105 to the request duplication/transfer execution unit 122. FIG.

S1810 is executed by the request replication/transfer execution unit 122 upon receiving the input of S751. In S1810, the request replication/transfer execution unit 122 executes S1811 to S1814.

In S1811, the request replication/transfer execution unit 122 refers to the request transfer destination information management table 121B. Information in the request transfer destination information management table 121B may be cached in the request replication/transfer execution unit 122 in S961 of FIG.

In S1812, the request replication/transfer execution unit 122 receives the input of the terminal data, and determines the distribution rule of FIG. 17 in which the reference data path 611 of FIG. Using 1612 as a key, a row of the request transfer destination information management table 121B is searched. In this process, based on changes in the business index of the system user, it is necessary to transfer the terminal data to several node management nodes 123B used by individual system users for each terminal data identifier 1411 of the terminal data. Get information.

In S1813, the request duplication/transfer execution unit 122 duplicates the terminal data and transfers the terminal data to the tenant name 612 specified by the search line.

In S<b>1814 , the request duplication/transfer execution unit 122 transfers the business index 1613 of the search line and the terminal data to the business index extraction unit 1123 .

S1820 is executed by the business index extraction unit 1123 in response to the input of S1814. In S1820, the business index extraction unit 1123 executes S1821 to S1822.

In S1821, the business index extraction unit 1123 searches for a line from the business index calculation logic management table 1121 using the business index 1613 as a key.

In S1822, the business index extraction unit 1123 calculates a business index using the terminal data information corresponding to the calculation use data item 1412 in FIG. In addition, the business index extraction unit 1123 inputs the business index 1613 searched in S1814 (different from the business index calculated in S1822) to the target performance index calculation unit 1124 .

S1830 is executed by the target performance index calculation unit 1124 in response to the input of S1822. In S1830, the target performance index calculation unit 1124 executes S1831 to S1834.

In S1831, the target performance index calculation unit 1124 compares the business index calculated in S1822 with the business index 1613 searched in S1814. If both match, it is determined that there has been no change in the business index for the terminal data identified by the terminal data identifier 1411, and the process ends. If the two do not match, it is determined that the business index has changed, and the process proceeds to S1832.

In S1832, the target performance index calculation unit 1124 performs a line search on the request forwarding destination information management table 121B using the business index calculated in S1822 as a key. If the line search hits, it means that the node management tenant 123B that processes the terminal data indicating the business index calculated in S1822 already exists. In this case, there is no need to deploy a new node management tenant 123B, so the target performance index calculation unit 1124 adds the terminal data identifier 1411 to register.

In S1833, the target performance index calculation unit 1124 searches the business index-specific node optimum setting management table 1122 for a row that includes the business index calculated in S1822.

In S<b>1834 , the node design pattern information 316 of the search line is input to the deploy execution unit 117 .

In S1841, the deploy executing unit 117 and the node setting mapping unit 116 deploy the new node management tenant 123B and processing node 131B by the processes shown in S730-S950 of FIG.

In S1842, the set of tenant name 612, processing node access information, and business index calculated in S1822 in FIG. 17 is stored in the request transfer destination information management table 121B in FIG.

In S1843, the request transfer destination information management table 121B is searched for a row containing the business index calculated in S1822, and the terminal data identifier is registered in the distribution rule 1612. FIG.

FIG. 20 is a block diagram showing the hardware configuration of the data distribution management system of FIG.
20, data distribution management system 100 includes processor 11, communication control device 12, communication interface 13, main storage device 14 and external storage device 15. FIG. Processor 11 , communication control device 12 , communication interface 13 , main memory device 14 and external memory device 15 are interconnected via internal bus 16 . Main memory device 14 and external memory device 15 are accessible from processor 11 .

An input device 20 and an output device 21 are provided outside the data distribution management system 100 . Input device 20 and output device 21 are connected to internal bus 16 via input/output interface 17 . The input device 20 is, for example, a keyboard, mouse, touch panel, card reader, voice input device, or the like. The output device 21 is, for example, a screen display device (liquid crystal monitor, organic EL (Electro Luminescence) display, graphic card, etc.), an audio output device (speaker, etc.), a printing device, or the like.

The processor 11 is hardware that controls the operation of the entire data distribution management system 100 . The main memory device 14 can be composed of semiconductor memory such as SRAM or DRAM, for example. The main memory device 14 can store a program being executed by the processor 11 and can provide a work area for the processor 11 to execute the program.

The external storage device 15 is a storage device with a large storage capacity, such as a hard disk device or an SSD (Solid State Drive). The external storage device 15 can hold executable files of various programs and data used for executing the programs. The external storage device 15 can store a data distribution management program 15A and management information 15B. The data distribution management program 15A may be software that can be installed in the data distribution management system 100, or may be incorporated in the data distribution management system 100 as firmware.

The communication control device 12 is hardware having a function of controlling communication with the outside. Communication control device 12 is connected to network 19 via communication interface 13 . The network 19 may be a WAN (Wide Area Network) such as the Internet, a LAN (Local Area Network) such as WiFi or Ethernet (registered trademark), or a mixture of WAN and LAN. good too.

The input/output interface 17 converts data input from the input device 20 into a data format processable by the processor 11, and converts data output from the processor 11 into a data format processable by the output device 21. .

The processor 11 reads the data distribution management program 15A into the main storage device 14 and executes the data distribution management program 15A to collect IoT data from the terminal device 105 via the network 19, and use the data specifications of the IoT data in the system. It can be flexibly changed and distributed for each person. The processor 11 can refer to the management information 15B when executing the data distribution management program 15A. At this time, the data distribution management program 15A performs the functions of the system operation API provider 111, the node setting mapping unit 116, the deploy execution unit 117, the request duplication/transfer execution unit 122, and the node management tenant 123 shown in FIG. can be realized. The management information 15B can include information of the user information management table 112, the data schema management table 113, the node setting management table 115, the package management repository 118, and the request transfer destination information management table 121 of FIG. .

Execution of the data distribution management program 15A may be shared among a plurality of processors or computers. Alternatively, the processor 11 may instruct a cloud computer or the like via the network 19 to execute all or part of the data distribution management program 15A and receive the execution result.

In addition, the present invention is not limited to the above-described embodiments, and includes various modifications. For example, the above-described embodiments have been described in detail in order to explain the present invention in an easy-to-understand manner, and are not necessarily limited to those having all the described configurations. In addition, it is possible to replace part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. Moreover, it is possible to add, delete, or replace a part of the configuration of each embodiment with another configuration. Further, each of the above configurations, functions, processing units, processing means, and the like may be realized by hardware, for example, by designing a part or all of them using an integrated circuit.

DESCRIPTION OF SYMBOLS 100... Data distribution management system 101... User console 102... System management part 103... System execution part 105... Terminal device 106... Terminal data reception part 107... Data provider system 108... Data user system , 109 ... input data at the time of use, 111 ... system operation API providing unit, 112 ... user information management table, 113 ... data schema management table 115 ... node setting management table, 116 ... node setting mapping unit, 117 ... deploy execution unit, 118... package management repository, 121... request transfer destination information management table, 121... request replication/transfer execution unit, 123... node management tenant, 131... processing node, 132... data processing unit, 133... request data transmission/reception unit, 171... Company A 172 Organization B 173 Package manager 174 Terminal group managed by Company A 1101 Business index-performance index mapping data 1120 Data distribution system dynamic management unit 1121 Business index calculation logic Management table 1122: Business index-specific node optimum setting management table 1123: Business index extraction unit 1124: Target performance index calculation unit 1131: Request transfer destination information management table

Claims (12)

a management unit that manages data specifications of destinations of terminal data collected from terminal devices;
a processing node that is deployed based on data specifications of the destination managed by the management unit and executes processing related to specification changes of the terminal data;
A node management tenant that manages the processing node ;
A data distribution management system in which the node management tenant is provided for each provider who provides the terminal data and each user who uses the terminal data . 2. The data distribution management system according to claim 1, wherein said node management tenant is provided for each said destination. 2. A data distribution management system according to claim 1 , wherein the presence or absence of provision and the range of provision partners are defined for each data item included in the content provided from the provider to the user. 2. The data distribution management according to claim 1 , wherein an address of said processing node is set for each said node management tenant for one reference data path which is an identifier of an endpoint used when said terminal data is transmitted. system. a repository that manages a package in which data conversion functions related to specification changes of the terminal data are templated;
a mapping unit that sets parameters related to the data specification change to the package;
2. The data distribution management system according to claim 1 , further comprising a deploy execution unit that deploys the package in which the parameter is set as the node management tenant. The deploy execution unit
Regarding the content contained in the terminal data, the data format of text, image and video, the transmission form of the terminal data and the data provision method including the data provision protocol, the latency, throughput and number of simultaneously connected terminals of the data provision 6. The data distribution management system according to claim 5 , wherein the templated package is selected based on performance requirements included. The processing node is
converting the data structure of the terminal data into a data structure defined by the data specification of the destination;
2. The data distribution management system according to claim 1, wherein batch data conversion of streaming data is performed by buffering data whose data structure has been converted, or sampling of data whose data structure has been converted is performed. Supplementing information including the batch interval required for creating the batch data based on the data specifications of the destination or external input,
8. The data distribution management system according to claim 7 , wherein the information including the identifier and sampling rate of the data to be sampled is complemented based on the data specification of the destination or an external input. If the content of the destination data specification contains structured text data,
acquiring a content definition of the data specification of the terminal data and a content definition of the data specification of the destination;
8. The data distribution management system according to claim 7 , wherein data items included in said terminal data are deleted or item names are changed. A node management tenant that manages the processing node;
2. The data distribution management system according to claim 1, wherein said node management tenants are divided according to the contents of said terminal data. 11. The data distribution management system according to claim 10 , wherein the node management tenants are divided based on a business index relating to the business of users who use the terminal data. managing the business index for each terminal device or for each identifier included in the terminal data;
calculating the business index based on the terminal data acquired from the terminal device;
12. The data distribution management system according to claim 11 , wherein a node management tenant corresponding to the calculated business index is deployed based on a comparison result between the managed business index and the calculated business index.

Images