CN113961174B - Model development and deployment method based on cloud native microservice - Google Patents
Model development and deployment method based on cloud native microservice Download PDFInfo
- Publication number
- CN113961174B CN113961174B CN202111238955.7A CN202111238955A CN113961174B CN 113961174 B CN113961174 B CN 113961174B CN 202111238955 A CN202111238955 A CN 202111238955A CN 113961174 B CN113961174 B CN 113961174B
- Authority
- CN
- China
- Prior art keywords
- model
- development
- user
- mirror image
- mechanism model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000011161 development Methods 0.000 title claims abstract description 114
- 238000000034 method Methods 0.000 title claims abstract description 35
- 230000007246 mechanism Effects 0.000 claims abstract description 111
- 239000000178 monomer Substances 0.000 claims description 18
- 238000012360 testing method Methods 0.000 claims description 18
- 230000001419 dependent effect Effects 0.000 claims description 12
- 238000012795 verification Methods 0.000 claims description 9
- 238000004422 calculation algorithm Methods 0.000 claims description 7
- 238000003860 storage Methods 0.000 claims description 7
- 230000008569 process Effects 0.000 claims description 5
- 238000012549 training Methods 0.000 claims description 4
- 238000013507 mapping Methods 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 abstract description 2
- 238000007726 management method Methods 0.000 description 10
- 238000004364 calculation method Methods 0.000 description 3
- 238000010276 construction Methods 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 230000002452 interceptive effect Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013499 data model Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/20—Software design
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/60—Software deployment
- G06F8/61—Installation
- G06F8/63—Image based installation; Cloning; Build to order
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
- G06F2009/4557—Distribution of virtual machine instances; Migration and load balancing
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Stored Programmes (AREA)
Abstract
The invention relates to a model development and deployment method based on cloud native microservices, and provides a comprehensive method integrating modeling, compiling, debugging, instantiation and deployment. The invention embeds abundant model application templates and components, supports custom development, enables general developers to rapidly develop own mechanism models based on the embedded model templates and components, rapidly deploys and applies own mechanism models based on the containerization technology, realizes knowledge sharing, improves knowledge multiplexing rate and development efficiency, and can rapidly integrate other systems or platforms based on micro services and industrial Internet in a seamless manner.
Description
Technical Field
The invention relates to a model development and deployment method based on cloud native microservices, and belongs to the technical field of industrial digital transformation.
Background
The industrial mechanism model and the data driving model are the cores in industrial production, and the production process can be more reasonably and optimally arranged and controlled by means of the effective model, so that the production quality is improved, the production efficiency is enhanced, and the cost is reduced. Most models have strong specialization and need to be developed by related specialized personnel. However, small and medium-sized enterprises generally have no own professional development team, and the use of mechanism models and data models is severely limited.
The patent provides an interactive model development environment system and method based on cloud environment, which is based on an HDFS and Spark computing architecture and completes the construction of the development environment through Ipython interactive programming interfaces. By the invention, users do not need to care about complicated scheduling, optimizing, configuring and the like of the big data clusters, but the use threshold of the tool is relatively high, the tool can be used only by big data related programming due to the fact that the tool must be provided with a foundation for big data development.
The patent of intelligent model development method and intelligent control method carries out algorithm model development and intelligent control based on algorithm components, and the list of the algorithm components comprises rich algorithm components, but the model development and intelligent control in the patent has high coupling degree, the data source is locked during the model development, and the same model cannot be instantiated for multiple times and is applied to multiple objects.
The invention provides a self-service visual data analysis method based on a cloud native microservice architecture, which comprises a data layer, a processing layer, an analysis layer and an application display layer. However, the invention does not have model training environment related to artificial intelligence in a processing layer and an analysis layer and does not provide a custom development operator environment.
The mode of developing the model by the patent is closely bound with data in actual application, so that the model cannot be rapidly, conveniently and repeatedly reused, the model requirements cannot be effectively and rapidly met in industrial scenes, but when equipment data sources are different, a mechanism model cannot be rapidly instantiated for multiple times and is applied to multiple objects. In addition, the traditional WEB development model is single development, the code efficiency is low and the method is difficult to maintain in the same project, the micro-service architecture just makes up the defect of the single architecture, and quick development and deployment are realized through effective splitting application.
Disclosure of Invention
The invention aims to solve the technical problems that: the mechanism model has stronger specialty and needs to be developed by related professionals. However, small and medium-sized enterprises generally have no own professional development team, and the development environment, development method and use of the mechanism model are severely limited due to lack of professional support. .
In order to solve the technical problems, the technical scheme of the invention provides a model development and deployment method based on cloud native microservices, which is characterized by comprising the following steps:
Step 1, creating a model folder in a model information management module, and creating different model files in the model folder to set corresponding basic data information;
Step 2, providing an environment for developing a mechanism model for a user by using a model development module, wherein the model development module is embedded with a mechanism model template, a development component and a data source, and the model development module comprises the following components: the development component comprises a general algorithm component, a typical industry mechanism model component and a user-defined component; the data source is used for debugging, training and testing a mechanism model established by a user;
Step 3, a user drags the mechanism model template and/or the development component into a modeling area provided by the model development module in a dragging mode, and the mechanism model template and the development component selected by the user are subjected to series-parallel connection to complete the development of the mechanism model, or the development component selected by the user is subjected to series-parallel connection to complete the development of the mechanism model, or the mechanism model template is directly selected to complete the development of the mechanism model;
Step 4, after the user completes the development of the mechanism model, the model test verification module is utilized to realize the online test of the mechanism model; when performing online test, a user firstly completes the setting of the input parameters of the model, then verifies whether the input and output meet the set conditions of the mechanism model through a model test verification module, and if not, iteratively adjusts the setting;
after the mechanism model passes the test verification, the development is completed;
And 5, completing the operation, deployment and management of the mechanism model by using a model instance service management module, wherein the method specifically comprises the following steps of:
Step 501, storing a mechanism model and related configuration files thereof in a model library, wherein the configuration files comprise a model file, a project dependent file requirements. Txt, a construction mirror text file Dockerfile and the like, and the project dependent file requirements. Txt is a model running dependent environment configuration file; the mirror image text file Dockerfile is a text file for constructing a mirror image, and the content of the file comprises a command and an instruction required for constructing the mirror image, and when the mirror image text file Dockerfile is compiled, the command in the mirror image text file Dockerfile can reproduce the whole environment configuration process;
step 502, rewriting the model running dependency environment configuration in the mirror text file Dockerfile, and pointing to the project dependency file requirements.
Step 503, under the storage directory of the mirror image text file Dockerfile, constructing a model mirror image by executing a Docker Build command, after executing the Docker Build command, downloading and installing a dependency environment required by running the model according to the configuration in the project dependency file requirements.txt, and storing the model mirror image in a mirror image bin after the model mirror image is successfully constructed;
step 504, the user selects a target mechanism model by using a model instance service management module, and the model instance service management module pulls a corresponding model mirror image from a mirror image warehouse;
in the step, in order to improve the convenience of operation, after a target mechanism model is selected, an input/output port of the target mechanism model is presented on a user interface in the form of an interface list, input and output parameter points corresponding to the input/output port are bound with parameter points of an equipment model established according to actual equipment, and a mapping relation between an external input data source and input parameters of the target mechanism model and between output of the target mechanism model and the external data source is established; then, the parameter interface configuration information of the input/output port of the target mechanism model is added into a parameter information configuration file Arguementrequest;
Step 505, rewriting the model running dependent environment configuration in the mirror text file Dockerfile, and pointing to the parameter information configuration file armmentrequest elements.
Step 506, under the storage directory of the mirror image text file Dockerfile, constructing a new model mirror image by executing a Docker Build command, and after executing the Docker Build command, performing interface configuration according to the setting in the parameter information configuration file Arguertrequest.
Step 507, after the new model image is generated, selecting the model to run on the edge side or the cloud:
If the model is selected to run in the cloud, starting a new container based on the new model image obtained in the step 506, generating a container instance, and thus deploying the model in the container and enabling the model to start to run in the container environment;
If the operation on the edge side is selected, the new model image obtained in step 506 and the configuration information of the parameter interface of the input/output port of the target mechanism model are issued to the edge side for operation, and the edge side is an edge side platform or an edge side gateway.
Preferably, in step 1, the basic data information includes metadata of the model, interface data of the model, and protocol data of the model.
Preferably, in step 2, the user can directly develop the operator online based on the Web terminal, then perform online compiling and debugging on the operator, and the operator after successful debugging is packaged into a component to be used as a user-defined component; the user can develop the operator locally and upload the operator to the model development module after compiling and debugging the operator is completed locally, and the model development module packages the received operator into components to serve as user-defined components.
Preferably, in step 3, a mechanism model consisting of only development components is defined as a monomer model, after the user completes the development of the monomer model, the monomer model can be secondarily packaged into a new mechanism model template which is embedded in a model development module, and a new model formed by the monomer model and the monomer model or the monomer model and the components in series-parallel connection is defined as a combined model.
Preferably, the user may click on the mechanism model template during development of the mechanism model to view the development component composition within the user.
Preferably, when the mechanism model is tested, any development component and/or mechanism model template in the mechanism model can be clicked to check the input and output states of the development component and/or mechanism model template, so that intermediate results can be checked, and problem points can be timely and reasonably positioned when problems occur in the intermediate results.
The invention provides a rapid development and deployment method for a mechanism model of an industrial Internet, which is a comprehensive method integrating modeling, compiling, debugging, instantiation and deployment. The invention embeds abundant model application templates and components, supports custom development, enables general developers to rapidly develop own mechanism models based on the embedded model templates and components, rapidly deploys and applies own mechanism models based on the containerization technology, realizes knowledge sharing, improves knowledge multiplexing rate and development efficiency, and can rapidly integrate other systems or platforms based on micro services and industrial Internet in a seamless manner.
Compared with the prior art, the invention has the following advantages:
1) According to the technical scheme provided by the invention, a model development and deployment system and method based on a cloud native micro-service architecture, which can be conveniently developed and deployed, are provided for small and medium-sized enterprises, mechanism model developers and mechanism model application engineers;
2) The method is embedded with a professional rich model template library and component library, and a general mechanism model can be directly built based on the component library and the template library, so that the difficulty of model development is reduced, and the development threshold and the working strength are reduced;
3) According to the model instantiation method provided by the invention, when the model principles or rules required by industrial scenes are the same, different data sources can be bound through the interface information by copying the model mirror image, and the container is constructed based on the model mirror image, so that repeated instantiation is performed, repeated multiplexing of the model is realized, and development workload is reduced;
4) The model is deployed and operated based on the container, and has the advantages that the container is used to eliminate the environmental difference between model development and model operation, and ensure the environmental consistency standardization of the application life cycle; based on container development and operation, the adaptability is stronger, cross-platform deployment can be realized, more and more cloud platforms support containers, and users do not need to worry about binding by the cloud platforms;
5) According to the instance method provided by the invention, when the instantiation deployment is carried out, the model instance can be deployed on the cloud end or the edge side according to actual requirements, the response capability is faster when the edge calculation is carried out on the edge side, and the pressure of the cloud end storage and calculation can be relieved after a part of data is processed on the edge side.
Drawings
FIG. 1 is a diagram of a model open system functional framework based on a cloud native microservice architecture;
FIG. 2 is a model development flow and an instantiated deployment application flow.
Detailed Description
The application will be further illustrated with reference to specific examples. It is to be understood that these examples are illustrative of the present application and are not intended to limit the scope of the present application. Furthermore, it should be understood that various changes and modifications can be made by one skilled in the art after reading the teachings of the present application, and such equivalents are intended to fall within the scope of the application as defined in the appended claims.
The invention aims to reduce development threshold and development amount, provides a Web end development and deployment using method which is easy to develop and quick to deploy for mechanism models, data-driven model developers and users, and provides a model development and deployment method based on a cloud native micro-service architecture for small and medium-sized enterprises, thereby reducing the development threshold and improving development and application efficiency.
The method provided by the invention specifically comprises the following steps:
Step 1, creating a model folder in a model information management module, and creating different model files in the model folder to set corresponding basic data information, wherein the basic data information comprises metadata of a model, interface data of the model and protocol data of the model.
Step 2, providing an environment for developing a mechanism model for a user by using a model development module, wherein the model development module is embedded with a mechanism model template, a development component and a data source, and the model development module comprises the following components: the development component comprises a general algorithm component such as data processing, machine learning, deep learning and the like, a typical industry mechanism model component such as a motor, electricity, energy and the like and a user-defined component; the user can directly develop the operator on line based on the Web end, the invention supports python, java, matlab, R and other common model development languages, then carries out on-line compiling and debugging on the operator, and the successfully debugged operator is packaged into a component to be used as a user-defined component; the user can develop the operator locally, and upload the operator to the model development module after compiling and debugging the operator are finished locally, and the model development module packages the received operator into a component to be used as a user-defined component;
the data sources comprise simulation data, local uploading data, database data and the like, and are mainly used for debugging, training and testing a model created by a user.
Step 3, a user drags the mechanism model template and/or the development component into a modeling area provided by the model development module in a dragging mode, and the mechanism model template and the development component selected by the user are subjected to series-parallel connection to complete the development of the mechanism model, or the development component selected by the user is subjected to series-parallel connection to complete the development of the mechanism model, or the mechanism model template is directly selected to complete the development of the mechanism model;
Defining a mechanism model only consisting of a development component as a monomer model, defining the mechanism model consisting of the monomer model and the monomer model or the monomer model and the component as a combined model, and encapsulating the monomer model into a new mechanism model template for embedding in a model development module after a user of the combined model finishes the development of the monomer model;
a user may click on the mechanism model template during development of the mechanism model to view the development component composition within the user.
Step 4, after the user completes the development of the mechanism model, the model test verification module is utilized to realize the online test of the mechanism model; when performing online test, a user firstly completes the setting of the input parameters of the model, then verifies whether the input and output meet the set conditions of the mechanism model through a model test verification module, and if not, iteratively adjusts the setting;
when the mechanism model is tested and operated, any development component and/or mechanism model template in the mechanism model can be clicked to check the input and output states of the development component and/or mechanism model template, so that intermediate results can be checked, and problem points can be timely and reasonably positioned when problems occur in the intermediate results;
after the mechanism model passes the test verification, the development is completed.
And 5, completing the operation, deployment and management of the mechanism model by using a model instance service management module, wherein the method specifically comprises the following steps of:
In order to reduce the time for initializing the configuration environment during the instantiation of the mechanism model, after the development of the mechanism model is completed, a model base mirror image is constructed, a mirror image of the environment required by the calculation of the mechanism model is generated and stored in a container mirror image warehouse of cloud resources, and the process of constructing a mirror image file is as follows:
Step 501, storing a mechanism model and related configuration files thereof in a model library, wherein the configuration files comprise a model file, a project dependent file requirements. Txt, a construction mirror text file Dockerfile and the like, and the project dependent file requirements. Txt is a model running dependent environment configuration file; the mirror image text file Dockerfile is a text file for constructing a mirror image, and the content of the file comprises a command and an instruction required for constructing the mirror image, and when the mirror image text file Dockerfile is compiled, the command in the mirror image text file Dockerfile can reproduce the whole environment configuration process;
step 502, rewriting the model running dependency environment configuration in the mirror text file Dockerfile, and pointing to the project dependency file requirements.
Step 503, under the storage directory of the mirror image text file Dockerfile, constructing a model mirror image by executing a Docker Build command, after executing the Docker Build command, downloading and installing a dependency environment required by running the model according to the configuration in the project dependency file requirements.txt, and storing the model mirror image in a mirror image bin after the model mirror image is successfully constructed;
step 504, the user selects a target mechanism model by using a model instance service management module, and the model instance service management module pulls a corresponding model mirror image from a mirror image warehouse;
In the step, in order to improve the convenience of operation, after a target mechanism model is selected, an input/output port of the target mechanism model is presented on a user interface in the form of an interface list, then input and output parameter points corresponding to the input/output port are bound with parameter points of an equipment model established according to actual equipment, and a mapping relation between an external input data source and input parameters of the target mechanism model and between output of the target mechanism model and the external data source is established, wherein the equipment model is stored in an IOT database in advance; then, the parameter interface configuration information of the input/output port of the target mechanism model is added into a parameter information configuration file Arguementrequest;
Step 505, rewriting the model running dependent environment configuration in the mirror text file Dockerfile, and pointing to the parameter information configuration file armmentrequest elements.
Step 506, under the storage directory of the mirror image text file Dockerfile, constructing a new model mirror image by executing a Docker Build command, and after executing the Docker Build command, performing interface configuration according to the setting in the parameter information configuration file Arguertrequest.
Step 507, after the new model image is generated, selecting the model to run on the edge side or the cloud:
If the model is selected to run in the cloud, starting a new container based on the new model image obtained in the step 506, generating a container instance, and thus deploying the model in the container and enabling the model to start to run in the container environment;
If the operation on the edge side is selected, the new model image obtained in step 506 and the configuration information of the parameter interface of the input/output port of the target mechanism model are issued to the edge side for operation, and the edge side is an edge side platform or an edge side gateway.
When the model principles or rules required by the industrial scene are the same, different data sources can be bound through interface configuration by copying the model images in the image warehouse, a new model image is constructed, then a container instance is built based on the new model image, repeated instantiation is performed, repeated multiplexing of the model is realized, and development workload is reduced. In order to ensure independence between model instances and prevent single instance operation errors from affecting other instances, in the invention, a new independent container is built based on a new model mirror image every time the model instance is built.
Claims (6)
1. The model development and deployment method based on the cloud native microservice is characterized by comprising the following steps of:
Step 1, creating a model folder in a model information management module, and creating different model files in the model folder to set corresponding basic data information;
Step 2, providing an environment for developing a mechanism model for a user by using a model development module, wherein the model development module is embedded with a mechanism model template, a development component and a data source, and the model development module comprises the following components: the development component comprises a general algorithm component, a typical industry mechanism model component and a user-defined component; the data source is used for debugging, training and testing a mechanism model established by a user;
Step 3, a user drags the mechanism model template and/or the development component into a modeling area provided by the model development module in a dragging mode, and the mechanism model template and the development component selected by the user are subjected to series-parallel connection to complete the development of the mechanism model, or the development component selected by the user is subjected to series-parallel connection to complete the development of the mechanism model, or the mechanism model template is directly selected to complete the development of the mechanism model;
Step 4, after the user completes the development of the mechanism model, the model test verification module is utilized to realize the online test of the mechanism model; when performing online test, a user firstly completes the setting of the input parameters of the model, then verifies whether the input and output meet the set conditions of the mechanism model through a model test verification module, and if not, iteratively adjusts the setting;
after the mechanism model passes the test verification, the development is completed;
And 5, completing the operation, deployment and management of the mechanism model by using a model instance service management module, wherein the method specifically comprises the following steps of:
Step 501, storing a mechanism model and related configuration files thereof in a model library, wherein the configuration files comprise a model file, a project dependent file requirements. Txt, and a build mirror text file Dockerfile, wherein the project dependent file requirements. Txt is a model running dependent environment configuration file; the mirror image text file Dockerfile is a text file for constructing a mirror image, and the content of the file comprises a command and an instruction required for constructing the mirror image, and when the mirror image text file Dockerfile is compiled, the command in the mirror image text file Dockerfile can reproduce the whole environment configuration process;
step 502, rewriting the model running dependency environment configuration in the mirror text file Dockerfile, and pointing to the project dependency file requirements.
Step 503, under the storage directory of the mirror image text file Dockerfile, constructing a model mirror image by executing a Docker Build command, after executing the Docker Build command, downloading and installing a dependency environment required by running the model according to the configuration in the project dependency file requirements.txt, and storing the model mirror image in a mirror image bin after the model mirror image is successfully constructed;
step 504, the user selects a target mechanism model by using a model instance service management module, and the model instance service management module pulls a corresponding model mirror image from a mirror image warehouse;
in the step, in order to improve the convenience of operation, after a target mechanism model is selected, an input/output port of the target mechanism model is presented on a user interface in the form of an interface list, input and output parameter points corresponding to the input/output port are bound with parameter points of an equipment model established according to actual equipment, and a mapping relation between an external input data source and input parameters of the target mechanism model and between output of the target mechanism model and the external data source is established; then, the parameter interface configuration information of the input/output port of the target mechanism model is added into a parameter information configuration file Arguementrequest;
Step 505, rewriting the model running dependent environment configuration in the mirror text file Dockerfile, and pointing to the parameter information configuration file armmentrequest elements.
Step 506, under the storage directory of the mirror image text file Dockerfile, constructing a new model mirror image by executing a Docker Build command, and after executing the Docker Build command, performing interface configuration according to the setting in the parameter information configuration file Arguertrequest.
Step 507, after the new model image is generated, selecting the model to run on the edge side or the cloud:
If the model is selected to run in the cloud, starting a new container based on the new model image obtained in the step 506, generating a container instance, and thus deploying the model in the container and enabling the model to start to run in the container environment;
If the operation on the edge side is selected, the new model image obtained in step 506 and the configuration information of the parameter interface of the input/output port of the target mechanism model are issued to the edge side for operation, and the edge side is an edge side platform or an edge side gateway.
2. The method for developing and deploying a model based on a cloud native microservice according to claim 1, wherein in step 1, the basic data information includes metadata of the model, interface data of the model, and protocol data of the model.
3. The model development and deployment method based on the cloud native microservice according to claim 1, wherein in the step 2, a user can directly develop operators on line based on a Web end, then compile and debug the operators on line, and the operators after successful debugging are packaged into components to serve as user-defined components; the user can develop the operator locally and upload the operator to the model development module after compiling and debugging the operator is completed locally, and the model development module packages the received operator into components to serve as user-defined components.
4. The model development and deployment method based on the cloud native microservice according to claim 1, wherein in the step 3, a mechanism model consisting of only development components is defined as a monomer model, after the development of the monomer model is completed, a user can secondarily package the monomer model into a new mechanism model template which is embedded in a model development module, and a new model formed by serial-parallel connection of the monomer model and the monomer model or serial-parallel connection of the monomer model and the components is defined as a combined model.
5. The model development and deployment method based on cloud native microservices of claim 1 wherein a user can click on a mechanism model template to view the development component composition inside the mechanism model template in the process of developing the mechanism model.
6. The cloud native microservice-based model development and deployment method of claim 1, wherein when a mechanism model is tested, any development component and/or mechanism model template in the mechanism model can be clicked to check the input and output states of the development component and/or mechanism model template, so that intermediate results can be checked, and problem points can be timely and reasonably positioned when problems occur in the intermediate results.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111238955.7A CN113961174B (en) | 2021-10-25 | 2021-10-25 | Model development and deployment method based on cloud native microservice |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111238955.7A CN113961174B (en) | 2021-10-25 | 2021-10-25 | Model development and deployment method based on cloud native microservice |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113961174A CN113961174A (en) | 2022-01-21 |
| CN113961174B true CN113961174B (en) | 2024-04-30 |
Family
ID=79466611
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111238955.7A Active CN113961174B (en) | 2021-10-25 | 2021-10-25 | Model development and deployment method based on cloud native microservice |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113961174B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111857694A (en) * | 2020-07-25 | 2020-10-30 | 苏州浪潮智能科技有限公司 | Code editing method and device based on cloud native program development and deployment |
| CN112329945A (en) * | 2020-11-24 | 2021-02-05 | 广州市网星信息技术有限公司 | Model deployment and reasoning method and device |
| KR20210040884A (en) * | 2020-04-17 | 2021-04-14 | 베이징 바이두 넷컴 사이언스 앤 테크놀로지 코., 엘티디. | Edge computing test methods, devices, electronic devices and computer-readable media |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108415795B (en) * | 2018-02-12 | 2019-04-05 | 人和未来生物科技（长沙）有限公司 | A kind of container Dockerfile, container mirror image rapid generation and system |
-
2021
- 2021-10-25 CN CN202111238955.7A patent/CN113961174B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20210040884A (en) * | 2020-04-17 | 2021-04-14 | 베이징 바이두 넷컴 사이언스 앤 테크놀로지 코., 엘티디. | Edge computing test methods, devices, electronic devices and computer-readable media |
| CN111857694A (en) * | 2020-07-25 | 2020-10-30 | 苏州浪潮智能科技有限公司 | Code editing method and device based on cloud native program development and deployment |
| CN112329945A (en) * | 2020-11-24 | 2021-02-05 | 广州市网星信息技术有限公司 | Model deployment and reasoning method and device |
Non-Patent Citations (1)
| Title |
|---|
| Docker容器技术在MES系统部署中的应用;相铮;石春鹏;韩立新;;制造业自动化;20200425(04);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113961174A (en) | 2022-01-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112965785B (en) | Container-based micro-service application development method and development platform | |
| CN107831757B (en) | dSPACE-based motor controller hardware-in-loop automatic test system and method | |
| JP4950454B2 (en) | Stack hierarchy for test automation | |
| US7865350B1 (en) | Partitioning a model in modeling environments | |
| CN105512304B (en) | It is online to generate internet application method and system integration method and support platform | |
| CN105608258B (en) | A kind of Model-based diagnosis and information flow visual simulation system and method | |
| CN101739334A (en) | Automatic testing method of embedded software | |
| CN104615541A (en) | Simulation testing modeling system and method based on flow chart | |
| CN111813420B (en) | Method for carrying out automatic performance test on OpenStack cluster | |
| CN112328229A (en) | Method for quickly creating Jenkins assembly line | |
| CN107450897A (en) | The cross-platform implantation method and device of graphics engine | |
| CN103309800A (en) | Automatic webpage testing method and system | |
| EP4459525A1 (en) | Workflow generation method and system, and medium and program product | |
| Bartusevics et al. | Models for implementation of software configuration management | |
| CN113254054B (en) | Intelligent contract one-stop development system and method | |
| CN115185539A (en) | Method, device and storage medium for generating executable dynamic link library file | |
| Fursin | The collective knowledge project: Making ML models more portable and reproducible with open APIs, reusable best practices and MLOps | |
| CN113961174B (en) | Model development and deployment method based on cloud native microservice | |
| KR102293274B1 (en) | Method and apparatus for constructing test environment | |
| TWI695284B (en) | Agile software development system and method thereof | |
| CN114265595A (en) | Cloud native application development and deployment system and method based on intelligent contracts | |
| CN112148271B (en) | Method for automatically generating and injecting assembly process codes | |
| CN117908847B (en) | Programmable AI development platform | |
| CN112685051A (en) | Method, device, platform and storage medium for automatically executing shell script | |
| CN117762397B (en) | System rapid development platform and method based on control engine |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |