Operator edge processing method and device, equipment and storage medium

Abstract

Embodiments of the present application provide an operator edge processing method and device, equipment and a storage medium, applied to an operator edge hosting system, the operator edge hosting system has an operator center control service, the method comprises: obtaining a provided operator, constructing a container image for the operator; packaging the container image of the operator to obtain a container for the operator; deploying the container of the operator to an edge node, and managing the operator of the edge node through the operator center control service. While realizing the edge cloud deployment of the operator service, the management function of the operator can also be realized, achieving the purpose that the operator of the edge cloud node can meet the overall operation function when performing normal service operation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of cloud processing, in particular to an edge processing method of an operator, an edge processing device of an operator, a corresponding electronic device and a corresponding computer storage medium. BACKGROUND

[0002] With the rapid increase in the number of Internet intelligent terminal devices, and the advent of the 5G (5th Generation Mobile Communication Technology) and Internet of Things era, the traditional cloud computing center centralized storage and computing mode has been unable to meet the needs of terminal devices for timeliness, capacity and computing power, at which time edge cloud computing, referred to as edge cloud, can be used.

[0003] The edge cloud is based on the core and edge computing capabilities of cloud computing, and is a cloud computing platform built on edge infrastructure, while the edge hosting cluster service can be sunk to the edge side and terminal device, which mainly glues the cloud computing core capabilities and edge computing power to build a cloud computing platform on the edge infrastructure. Among them, the object of edge hosting can be an operator, but the related operator deployment tool usually focuses on operator deployment itself, and lacks the function implementation required for normal service operation. SUMMARY

[0004] In view of the above problems, the present application is proposed to provide an edge processing method of an operator, an edge processing device of an operator, a corresponding electronic device and a corresponding computer storage medium to overcome the above problems or at least partially solve the above problems.

[0005] The present application discloses an edge processing method of an operator, applied to an operator edge hosting system, the operator edge hosting system having an operator center control service, the method comprising:

[0006] Obtaining the provided operator, constructing a container image for the operator;

[0007] Packaging the container image of the operator to obtain a container for the operator;

[0008] Deploying the container of the operator to an edge node, and managing the operator of the edge node through the operator center control service.

[0009] Optionally, the constructing a container image for the operator comprises:

[0010] Obtaining the operator type of the provided operator; the operator type includes the form of the operator model source file and the form of the operator model container image;

[0011] Build a corresponding container image for the operator based on the operator type.

[0012] Optionally, the constructed container image includes a first container image constructed corresponding to the form of the operator model source file; the step of constructing a corresponding container image for the operator according to the operator type includes:

[0013] When the operator type is in the form of an operator model source file, obtain the model framework information of the operator model source file;

[0014] The model framework information of the operator is used to generate an image file adapted to the runtime environment;

[0015] Inject interface services and standardized services required for hosting the operator into the image file;

[0016] Using the image file, runtime environment deployment file, interface service, and standardized service, a first container image is constructed in the form corresponding to the operator model source file.

[0017] Optionally, the constructed container image includes a second container image constructed in a form corresponding to the operator model container image; the step of constructing a corresponding container image for the operator according to the operator type includes:

[0018] When the operator type is in the form of an operator model container image, if the operator model container image has a standardized service interface required to host the operator, then a management agent program is created for the operator model container image.

[0019] Using the operator model container image and the management agent, a second container image is constructed in a form corresponding to the operator model container image.

[0020] Optionally, it also includes:

[0021] When building a container image for the operator, the requirement parameter information is obtained: the requirement parameter information includes configuration information for the operator;

[0022] The parameters of the constructed container image are configured based on the configuration information of the operator.

[0023] Optionally, the operator edge hosting system includes an edge cloud platform and a container edge deployment platform, wherein deploying the operator's container to an edge node includes:

[0024] After building the container image for the operator, the requirement parameter information is obtained; the requirement parameter information includes the container deployment information for the operator on the edge node;

[0025] The container of the operator is deployed to the edge node of the edge cloud platform according to the container deployment information through the container edge deployment platform; the edge node includes a single edge node or multiple different edge nodes.

[0026] Optionally, the operator edge hosting system includes an operator center management device, and the management of operators on the edge nodes includes:

[0027] The operators hosted on the edge nodes are managed in a unified manner through the operator center management service provided by the operator center management device.

[0028] This application discloses an operator edge processing device applied to an operator edge hosting system, wherein the operator edge hosting system has an operator central management service, and the device includes:

[0029] The container image building module is used to obtain the provided operator and build a container image for the operator.

[0030] A container encapsulation module is used to encapsulate the container image of the operator to obtain a container for the operator;

[0031] The container deployment module is used to deploy the operator's container to the edge node and manage the operator on the edge node through the operator central management device.

[0032] Optionally, the container image building module includes:

[0033] The operator type acquisition submodule is used to acquire the operator type of the provided operator; the operator type includes the form of operator model source file and the form of operator model container image;

[0034] The container image building submodule is used to build the corresponding container image for the operator according to the operator type.

[0035] Optionally, the constructed container image includes a first container image constructed corresponding to the form of the operator model source file; the container image construction submodule includes:

[0036] The model framework information acquisition unit is used to acquire the model framework information of the operator model source file when the operator type is in the form of an operator model source file.

[0037] The image file generation unit is used to generate an image file adapted to the operating environment using the model framework information of the operator;

[0038] The service interface injection unit is used to inject interface services and standardized services required for hosting the operator into the image file.

[0039] The first container image building unit is used to obtain a first container image constructed in the form of the operator model source file by using the image file, the runtime environment deployment file, the interface service, and the standardized service.

[0040] Optionally, the constructed container image includes a second container image constructed in a form corresponding to the operator model container image; the container image construction submodule includes:

[0041] The management agent creation unit is used to create a management agent for the operator model container image when the operator type is in the form of an operator model container image, if the operator model container image has a standardized service interface required to host the operator;

[0042] The second container image building unit is used to obtain a second container image constructed in a form corresponding to the operator model container image by using the operator model container image and the management agent program.

[0043] Optionally, the container image building module further includes:

[0044] The configuration information acquisition submodule is used to acquire requirement parameter information when building a container image for the operator: the requirement parameter information includes configuration information for the operator;

[0045] The parameter configuration submodule is used to configure the parameters of the constructed container image based on the configuration information of the operator.

[0046] Optionally, the operator edge hosting system includes an edge cloud platform and a container edge deployment platform, wherein the container deployment module includes:

[0047] The container deployment information acquisition submodule is used to acquire requirement parameter information after building a container image for the operator; the requirement parameter information includes container deployment information for the operator on edge nodes;

[0048] The edge node deployment submodule is used to deploy the operator's container to the edge node of the edge cloud platform according to the container deployment information through the container edge deployment platform; the edge node includes a single edge node or multiple different edge nodes.

[0049] Optionally, the operator edge hosting system includes operator center management equipment, and the container deployment module includes:

[0050] The operator management submodule is used to uniformly manage the operators hosted on the edge nodes through the operator center management service provided by the operator center management device.

[0051] This application discloses an electronic device, including: a processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein when the computer program is executed by the processor, it implements the steps of any of the edge processing methods of the operator.

[0052] This application discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of any of the edge processing methods of the operator.

[0053] The embodiments of this application have the following advantages:

[0054] In this embodiment, a container image is built for the provided operator and the built container image is encapsulated to obtain a container for the operator. This container can encapsulate the operator provided by the user / third party in the form of a container. The encapsulated container is deployed to the edge node, and the operator center management service manages the operators on the edge node in a unified manner. While realizing the edge cloud deployment of operator services, it can also realize the management and control function of operators, so that the operators on the edge cloud node can have the necessary functions for overall operation when running normal services. Attached Figure Description

[0055] Figure 1 This is a flowchart illustrating the steps of an embodiment of an edge processing method for an operator according to this application;

[0056] Figure 2 This is a schematic diagram of the architecture of the operator edge hosting system provided in the embodiments of this application;

[0057] Figure 3 This is a schematic diagram of the operation of the operator edge hosting system provided in the embodiments of this application;

[0058] Figure 4 This is a flowchart illustrating the steps of another embodiment of the edge processing method using an operator in this application.

[0059] Figure 5 This is a schematic diagram illustrating an application scenario of the operator edge processing provided in the embodiments of this application;

[0060] Figure 6 This is a structural block diagram of an embodiment of an edge processing device for an operator according to this application. Detailed Implementation

[0061] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0062] To facilitate understanding of this application by those skilled in the art, the terms or nouns involved in the following embodiments of this application are explained below:

[0063] AI: Artificial Intelligence.

[0064] Operators can be understood as computational tools, typically belonging to a stage in the process of artificial intelligence / machine learning. Operators can be used to encapsulate a certain algorithm, or a step or link in an algorithm, and can serve as a specific model unit, that is, an operator model can be constructed based on several operators.

[0065] IaaS: Infrastructure as a Service, refers to providing IT (Information Technology) infrastructure as a service through a network, such as cloud-based services, pay-as-you-go services, and services for storage, networking, and virtualization.

[0066] PaaS: Platform as a Service, refers to a business model that provides services by offering server platforms as a service over a network, such as hardware and software tools available on the Internet.

[0067] SaaS: Software as a Service, which means providing software services over the network, such as software that can be obtained through a third party via the Internet.

[0068] API: Application Programming Interface.

[0069] The objects processed at the edge by the edge-hosted cluster service can be operators, which can be specific model units that encapsulate related algorithms. In this case, the computing power of the operators can be pushed down to the edge side or the device side, and the operators can be deployed on the edge cloud platform.

[0070] For deploying operators on edge cloud platforms, the relevant operator deployment tools usually focus on the operator deployment itself and lack the functions required for normal service operation, such as scheduling, billing, and high availability. In addition, existing platforms cannot deploy operators to edge cloud platforms and do not have the function of linking with other services of edge cloud.

[0071] To achieve both edge cloud deployment of operator services and operator management, ensuring that operators on edge cloud nodes can fulfill overall operational requirements during normal service operation, related technologies often involve users or third-party platforms purchasing cloud service IaaS layer resources (such as virtual machines and containers) and manually deploying the operator services to the edge cloud platform according to the specific nodes where the operators are deployed. A central cloud platform is then set up to manage the deployed operators. However, this manual deployment method is labor-intensive for both deployment and maintenance, and it cannot achieve batch replication, meaning it cannot perform batch deployment of operators or productize the operator deployment process.

[0072] One of the core ideas of this application's embodiments lies in building container images of the provided operators and encapsulating the built container images. This automatically encapsulates user / third-party provided operators in the form of containers, deploys the encapsulated containers to edge nodes, and allows the operator center management service to uniformly manage the operators deployed on the edge nodes. This ensures that the operators on the edge cloud nodes have the necessary functions for overall operation during normal service execution, including operator management, operator scheduling, metering and billing, and high availability. Furthermore, it not only enables edge cloud deployment of operator services but also achieves operator management and control functions and linkage with other edge cloud services through standardized interface capabilities injected into the containers. Moreover, based on the automatic construction of container images and the automatic encapsulation and deployment of containers, cumbersome manual deployment and maintenance operations are eliminated, enabling batch automated operations for operator edge hosting and realizing PaaS or SaaS productization capabilities for operator deployment.

[0073] Reference Figure 1 This diagram illustrates a flowchart of an embodiment of an operator edge processing method according to this application, applied to an operator edge hosting system. The operator edge hosting system has an operator center management service and may specifically include the following steps:

[0074] Step 101: Obtain the provided operator and build a container image for the operator;

[0075] Edge-hosted cluster services can extend cloud computing capabilities from the cloud to the edge and terminal devices. By combining core cloud computing capabilities with edge computing power, a cloud computing platform can be built on edge infrastructure. The objects processed at the edge by edge-hosted cluster services can be operators, which can be specific model units that encapsulate relevant algorithms. In this case, the computing power of the operators can be extended to the edge or device side.

[0076] Edge processing of operators, or edge hosting operations for operators, can be specifically manifested as deploying operators to edge nodes. To facilitate the deployment of operators on edge nodes, a corresponding container image can be built for the operator. The built container image can contain related applications and application dependencies, as well as process information of the application running at startup. During the construction of the operator's container image, the dependencies and process information contained in the container image can be represented as a whole, including the runtime environment, dependency files, etc., making it convenient to use the constructed whole for subsequent deployment on edge nodes.

[0077] The dependency files included in the constructed container image are mostly library files, including both dynamic and static libraries. If a dependency file for a certain operator is not deployed on the edge node, and only the operator itself is deployed, the deployed operator often cannot function properly. In this case, the dependency file can be used to ensure the normal operation of the operator. The runtime environment included in the container image can represent the specific runtime environment of the operator on the edge node, allowing for the configuration of runtime parameters within the container image to ensure that the deployed operator meets the required runtime environment. That is, the runtime environment can also be used to ensure the normal operation of the operator on the edge node. It should be noted that the construction of the container image for the operator can also be achieved by building a virtual machine image for the operator, etc., and this embodiment of the application does not impose any limitations on this.

[0078] In one embodiment of this application, the construction of the container image of the operator can be implemented by selecting a method that matches the operator type provided. The managed operator can be provided in the form of operator model source file or container image. That is, different image construction methods can be used to construct a first container image corresponding to the form of operator model source file, and / or to construct a second container image corresponding to the form of operator model container image.

[0079] It should be noted that the hosted operators can be provided by the operator edge hosting system, by the user, or by a third party. The type of operator provided is not affected by different sources. That is, when selecting the image building method that matches the operator type, it is not affected by different sources. Operators from any source can be image built by selecting the corresponding matching method based on their operator type. This application embodiment does not limit this.

[0080] Step 102: Encapsulate the container image of the operator to obtain a container for the operator;

[0081] After building the corresponding container image for the operator, the resulting container image can be encapsulated to obtain a container for the operator, which can then be deployed on edge nodes. Containerization of user / third-party provided operators can be achieved by creating a new container and loading the container image built for the operator into that container.

[0082] The constructed container image is injected with interface services and standardized service interfaces required for managing operators. Therefore, the encapsulated container not only has the computing power of the operator itself, but also interface service capabilities and standardized service capabilities.

[0083] Step 103: Deploy the operator container to the edge node and manage the operators on the edge node through the operator center management service.

[0084] After encapsulating the operators provided by users / third parties in the form of containers, the encapsulated containers can be deployed to edge nodes. The operator center management service can uniformly manage the operators on the edge nodes, so that the operators on the edge cloud nodes can have the functions required for the overall operation when running normal services. At the same time, based on the managed operation of operators, the service deployment on the edge nodes can be realized, achieving the effect of cloud-edge collaboration.

[0085] It should be noted that the services provided by the operator center management service can be related services provided when calling the capabilities of the deployed operators, such as operator management, operator scheduling, metering and billing, high availability, etc. This application embodiment does not limit these capabilities.

[0086] In the embodiments of this application, the above-mentioned construction of container images based on operators and the packaging and deployment of containers can be automated. That is, based on the automatic construction of container images and the automatic packaging and deployment of containers, the tedious manual deployment and maintenance operations are saved, and the batch automated operation of operator edge hosting is realized, which can realize the PaaS layer or SaaS productization capability of operator deployment.

[0087] In a preferred embodiment, the encapsulated container can possess the operator's own computing power, interface service capabilities, and standardized service capabilities. Furthermore, through the standardized interface capabilities injected into the container, it can achieve operator control and management functions, as well as the ability to interact with other edge cloud services. In practical applications, after the operator is successfully deployed to edge nodes and managed at the edge, the managed operator primarily provides services via network service APIs. Users can directly access the deployed operator through API calls or through a gateway service for load balancing. This application does not limit the specific usage method of the deployed operator.

[0088] In this embodiment, a container image is built for the provided operator and the built container image is encapsulated to obtain a container for the operator. This container can encapsulate the operator provided by the user / third party in the form of a container. The encapsulated container is deployed to the edge node, and the operator center management service manages the operators on the edge node in a unified manner. While realizing the edge cloud deployment of operator services, it can also realize the management and control function of operators, so that the operators on the edge cloud node can meet the overall operation function when running normal services.

[0089] Reference Figure 2 This diagram illustrates the architecture of the operator edge hosting system provided in this application embodiment. The operator edge hosting system can encapsulate user / third-party provided operators in the form of containers, then deploy the encapsulated containers to edge nodes, and the operator center management service uniformly manages the operators on the edge nodes. Specifically, the operator edge hosting system 210 includes a container image building platform 211, a container edge deployment platform 212, an edge cloud platform 213, and an operator center management device 214.

[0090] Among them, such as Figure 3 As shown, the container image building platform 211 is mainly responsible for building container images for operators and encapsulating the built container images. The container edge deployment platform 212 is mainly responsible for obtaining the container images built for operators and the containers encapsulated from the container images, and performing containerized operator deployment, that is, for deploying operators on the edge cloud platform. The operator deployment operation requires the deployment capabilities of the container edge deployment platform. The edge cloud platform 213 can have multiple edge nodes, which can be used for operator hosting, realizing the purpose of sinking cloud computing from the cloud to the edge side and terminal devices, that is, for deploying containerized operators on edge nodes. The operator center management device 214 has an operator center management service, which can provide related services for users or third-party platforms to call the capabilities of the deployed operators, such as operator management, operator scheduling, metering and billing, high availability and other capabilities.

[0091] Reference Figure 4 This illustrates a flowchart of another embodiment of the edge processing method for operators in this application, applied to, for example... Figure 2 The operator edge hosting system shown here has operator center management services, which may include the following steps:

[0092] Step 401: Construct container images of operators based on operator type using the operator edge hosting system;

[0093] In one embodiment of this application, in order to facilitate the deployment of the operator on the edge node, a corresponding container image can be built for the operator. This container image may contain related applications and application dependencies, as well as process information of the application running at startup. During the construction of the container image of the operator, the dependencies and process information contained in the container image can be represented as converting the operator into a whole including the runtime environment, dependency files, etc., so that the constructed whole can be used for deployment when it is deployed on the edge node later.

[0094] For building container images, a method matching the provided operator type can be selected. In the specific implementation, the operator type of the provided operator can be obtained, and a corresponding container image can be built based on the operator type. The managed operator can be provided in the form of an operator model source file or a container image. Different image building methods can be used to build a first container image corresponding to the operator model source file format, and / or a second container image corresponding to the operator model container image format.

[0095] Since most of the dependency files contained in the constructed container image are library files, including both dynamic and static libraries, if the dependency files of a certain operator are not deployed on the edge node, and only the operator itself is deployed, the deployed operator often cannot run normally. In this case, the dependency files can be used to ensure the normal operation of the operator. In this embodiment of the application, the dependency files can specifically be represented as image files. The runtime environment contained in the container image can be used to represent the specific runtime environment of the operator on the edge node, so that the runtime parameters in the container image can be configured so that the deployed operator meets the required runtime environment. Therefore, the runtime environment can also be used to ensure the normal operation of the operator on the edge node.

[0096] Specifically, regarding the construction of container images, such as Figure 3 As shown, in one scenario, when the operator type is in the form of an operator model source file, the model framework information of the operator model source file can be obtained. Based on the model framework information, such as the operator's framework type, an image file adapted to the runtime environment can be generated to automatically match the runtime environment. In order to enable subsequent calls to the operator and interaction with other services of the edge cloud after the operator is deployed on the edge node, interface services, such as API service capabilities, and standardized services required for hosting the operator, such as operator registration service, operator liveness detection service, operator start / stop service, and resource monitoring service, can be injected into the image file. Then, the generated image file, runtime environment deployment file, injected interface services, and standardized services can be containerized as a whole to obtain the first container image built in the form of the operator model source file.

[0097] The injection of interface services and standardized services can be achieved by configuring the corresponding interface capability information in the generated image file. The specific injection process will not be described in detail in this embodiment.

[0098] For building container images, such as Figure 3 As shown, in another case, when the operator type is in the form of an operator model container image, if the operator model container image has the standardized service interface required when hosting the operator, a management agent can be created for the operator model container image, and a second container image constructed in the form of the operator model container image can be obtained by using the operator model container image and the created management agent.

[0099] The standardized service interface capabilities required by the provided operator model container image can be manifested as standardized service capabilities such as operator registration service, operator liveness detection service, operator start / stop service, and resource monitoring service. The operator registration service is implemented based on the registration interface of the container image, which can return basic information about the operator, including but not limited to name, version, and type. The operator liveness detection service is implemented based on the liveness detection interface of the container image, which can return the status of the operator program, such as running, stopped, or in an abnormal state. The operator start / stop service is implemented based on the start / stop interface of the container image, which can be used to provide the ability to control the start and stop of operators.

[0100] For example, when an operator is provided as a container image, if the image has standardized interfaces required for operator hosting (registration, activity detection, start / stop, etc.), a management agent can be added based on the operator model container image, and then the whole thing can be built into a container image. In this case, the management agent may include calling the registration, activity detection, start / stop, and other interfaces of the AI ​​operator model container image, and providing service capabilities such as resource monitoring. The management agent can specifically manifest as a background program, such as a management agent program.

[0101] It should be noted that when an operator is provided in the form of a container image, if the operator model container image does not have the standardized interface required for operator hosting, since the standardized interface required for operator hosting is a standardized definition of the operator edge hosting system for its internal system, an operator model container image with the required standardized interface can be managed by the operator edge hosting system in a standardized way. In theory, an operator model container image without the required standardized interface cannot be managed by the edge hosting system. However, in practice, the lack of standardized interface in the container image can be injected through manual adaptation, so that the operator model container image in this case meets the conditions for being built.

[0102] In practical applications, when building a container image for an operator, requirement parameter information can also be obtained. This requirement parameter information may include configuration information for the operator. At this point, the parameters of the built container image can be configured based on the operator's configuration information. Specifically, the requirement parameter information can be user-provided requirements from the operator's edge hosting system. These requirements parameters can be used to inform the system about some parameter configurations during image building. Therefore, some parameters in the container image can be customized, including but not limited to service name, image name, service port number, node distribution, and resource requirements.

[0103] In a preferred embodiment, after building a container image for the operator, the built container image can be encapsulated to obtain a container for the operator, so that the operator container can be deployed on edge nodes. The built container image is injected with interface services and standardized service interfaces required for hosting the operator. In addition to the computing power of the operator itself, the encapsulated container can also have interface service capabilities and standardized service capabilities. The interface service capabilities can be used to ensure the function of linkage with other services of the edge cloud, while the standardized service capabilities can be manifested as integrated registration, activation detection, start-up and shutdown, monitoring, etc.

[0104] Step 402: Deploy the operator's container to the edge node of the edge cloud platform through the container edge deployment platform;

[0105] The container edge deployment platform is mainly responsible for obtaining the container image built for the operator and the container encapsulated from the container image, and performing containerized operator deployment. That is, the operation of deploying the operator on the edge cloud platform needs to rely on the deployment capabilities of the container edge deployment platform. At this time, the container of the operator can be deployed to the edge node of the edge cloud platform through the container edge deployment platform.

[0106] Specifically, after building the container image for the operator, requirement parameter information can be obtained. This information can include container deployment information for the operator on edge nodes. At this point, the operator's container can be deployed to the edge nodes of the edge cloud platform according to the container deployment information through the container edge deployment platform. The required parameters can be user-provided by the user into the operator edge hosting system. These parameters inform the system of configuration settings during image building, allowing for custom configuration of some parameters in the container image. These parameters include, but are not limited to, service name, image name, service port number, node distribution, and resource requirements. Furthermore, for the same containerized operator, the deployed edge nodes can be a single edge node or multiple different edge nodes. This enables automated batch operations for operator edge hosting by automatically building container images and automatically packaging and deploying containers, eliminating tedious manual deployment and maintenance.

[0107] Step 403: Through the operator center management service provided by the operator center management equipment, the operators hosted on the edge nodes are managed in a unified manner.

[0108] To ensure that the operators on edge cloud nodes have the necessary functions for normal operation, it is necessary to build container images of the provided operators and encapsulate the built container images. This allows for the automatic encapsulation of user / third-party provided operators into containers, which are then deployed to edge nodes. The operator center management service then manages the operators on the edge nodes in a unified manner.

[0109] In practical applications, the operators hosted on edge nodes can be uniformly managed through the operator center management service provided by the operator center management device. The services provided by the operator center management service can be related services provided when calling the capabilities of the deployed operators, such as operator management, operator scheduling, metering and billing, high availability and other capabilities. This application embodiment does not limit this.

[0110] The operator control function can be manifested in the unified management of the registration, startup, and shutdown of third-party hosted operators on different edge cloud nodes. Operator scheduling can be manifested in the unified task and request scheduling capabilities for operators on different edge cloud nodes. Metering and billing can be manifested in the collection of operation logs, request logs, and traffic logs of operators on each edge cloud node for billing and monitoring. High availability can be manifested in the activation of operator images on the same edge cloud node or on nearby edge cloud nodes that meet the requirements when an operator on a certain edge cloud node is unavailable or has high resource levels, thus achieving high service availability. In addition, the operator center control service can also provide functions such as operator version management and lifecycle management for different edge cloud nodes. The functions provided by the operator center control service can be any service for the operators deployed under control; this application embodiment does not impose any limitations on this.

[0111] In a preferred embodiment, the encapsulated container can possess the operator's own computing power, interface service capabilities, and standardized service capabilities. Furthermore, through the standardized interface capabilities injected into the container, it can achieve the function of controlling the operator and linking with other services of the edge cloud. Linking with other services includes, but is not limited to, storing AI-recognized videos / images to edge storage, which can fully utilize the advantages of edge cloud such as low bandwidth cost, wide coverage, and low latency.

[0112] In practical applications, after the operator is successfully deployed to the edge node, the hosted operator can provide services mainly in the form of network service APIs. That is, users can directly call the API or access the deployed operator through the gateway service in the form of load balancing. The specific usage method is not limited.

[0113] In this embodiment, a container image is built for the provided operator and the built container image is encapsulated to obtain a container for the operator. This container can encapsulate the operator provided by the user / third party in the form of a container. The encapsulated container is deployed to the edge node, and the operator center management service manages the operators on the edge node in a unified manner. While realizing the edge cloud deployment of operator services, it can also realize the management and control function of operators, so that the operators on the edge cloud node can meet the overall operation function when running normal services.

[0114] To enable those skilled in the art to understand the operator edge processing method proposed in this application, refer to... Figure 5 The diagram illustrates an application scenario of the operator edge processing provided in this application embodiment, and the following description is provided in conjunction with this application scenario:

[0115] This application scenario can be a view computing scenario. In this scenario, the edge cloud platform hosted by the operator in the operator edge hosting system can be represented as an edge cloud view computing platform. At this time, the characteristics of edge cloud such as bandwidth saving, low latency, and wide coverage can be fully utilized to provide users with functions such as video access to the cloud, cloud processing, and local storage.

[0116] AI recognition is a crucial component of cloud-based view computing. It enables intelligent analysis of incoming videos and takes actions such as alarms, screenshots, and recordings to detect anomalies. The requirements for AI recognition vary across different scenarios. In such cases, view computing can introduce operators, such as the ability to host AI operators. This allows users or third parties to host their AI operators on the edge cloud view computing platform, providing centralized control and the ability to interact with other aspects of view computing. This meets the view processing needs of different customers in different scenarios and greatly expands the coverage of view computing products.

[0117] For example, such as Figure 5 As shown, for cloud-based video surveillance products, suppose a user needs to perform traffic congestion detection on the video stream. Currently, the cloud-based video surveillance product, namely the edge cloud view computing platform, may have the same or different operators A, B, and C on multiple edge nodes, but it lacks an AI operator for traffic congestion recognition. Therefore, it needs to collaborate with an AI algorithm team, with a third-party partner providing an AI model for traffic congestion recognition. In this case, suppose the edge cloud view computing platform currently lacks an AI operator for traffic congestion recognition called operator D. Operator D needs to be hosted on the edge cloud view computing platform so that the platform can acquire traffic congestion recognition capabilities based on calls to operator D.

[0118] The traffic congestion recognition AI model can be hosted at the edge based on the operator edge hosting system proposed in this application embodiment, and this AI model can be built into an edge AI service with standard capabilities. After the operator is successfully deployed to the edge node, the hosted operator can mainly provide services in the form of network service APIs. That is, users can directly call the API or access the deployed operator through the gateway service in the form of load balancing. That is, the video monitoring service can use its own screenshot and callback functions to obtain the recognition result and make a judgment after each successful screenshot by calling the above-mentioned hosted AI service. If the current recognition result shows traffic congestion, the video of the five minutes before and after the traffic congestion can be stored to complete the call and use of the edge AI service.

[0119] It should be noted that the operator edge deployment form and standardized interface capabilities defined in the embodiments of this application can be automatically managed by operators that conform to the form or standardized interface capabilities through the proposed solution. Furthermore, the provided operator edge deployment method can encapsulate and deploy user / third-party provided AI operators to the edge cloud platform, which is uniformly managed by the AI ​​operator center management service and can be linked with other edge cloud services. It can not only be applied to the view computing platform, but can also be regarded as a general AI operator hosting capability on the edge cloud platform, which can cover more scenarios.

[0120] In this embodiment, a container image is built for the provided operator and the built container image is packaged. This automatically packages the operator provided by the user / third party into a container and deploys the packaged container to the edge node. The operator center management service manages the operators on the edge node in a unified manner, so that the operators on the edge cloud node have the functions required for overall operation when the service is running normally, including operator management, operator scheduling, metering and billing, high availability and other capabilities. Moreover, it can not only realize the edge cloud deployment of operator services, but also realize the management function of operators and the function of linkage with other services of the edge cloud through the standardized interface capabilities injected in the container. Furthermore, based on the automatic construction of the container image and the automatic packaging and deployment of the container, the cumbersome manual deployment and maintenance operations are eliminated, realizing the batch automated operation of operator edge hosting, and realizing the PaaS layer or SaaS productization capability of operator deployment.

[0121] It should be noted that, for the sake of simplicity, the method embodiments are all described as a series of actions. However, those skilled in the art should understand that the embodiments of this application are not limited to the described order of actions, because according to the embodiments of this application, some steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also understand that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of this application.

[0122] Reference Figure 6 This diagram illustrates a structural block diagram of an embodiment of an operator edge processing device according to this application, applied to an operator edge hosting system. The operator edge hosting system has an operator center management service and may specifically include the following modules:

[0123] The container image building module 601 is used to obtain the provided operator and build a container image for the operator.

[0124] The container encapsulation module 602 is used to encapsulate the container image of the operator to obtain a container for the operator;

[0125] The container deployment module 603 is used to deploy the container of the operator to the edge node and manage the operator of the edge node through the operator central management device.

[0126] In one embodiment of this application, the container image building module 601 may include the following sub-modules:

[0127] The operator type acquisition submodule is used to acquire the operator type of the provided operator; the operator type includes the form of operator model source file and the form of operator model container image;

[0128] The container image building submodule is used to build the corresponding container image for the operator according to the operator type.

[0129] In one embodiment of this application, the constructed container image includes a first container image constructed corresponding to the form of the operator model source file; the container image construction submodule may include the following units:

[0130] The model framework information acquisition unit is used to acquire the model framework information of the operator model source file when the operator type is in the form of an operator model source file.

[0131] The image file generation unit is used to generate an image file adapted to the operating environment using the model framework information of the operator;

[0132] The service interface injection unit is used to inject interface services and standardized services required for hosting the operator into the image file.

[0133] The first container image building unit is used to obtain a first container image constructed in the form of the operator model source file by using the image file, the runtime environment deployment file, the interface service, and the standardized service.

[0134] In one embodiment of this application, the constructed container image includes a second container image constructed in a form corresponding to the operator model container image; the container image construction submodule may include the following units:

[0135] The management agent creation unit is used to create a management agent for the operator model container image when the operator type is in the form of an operator model container image, if the operator model container image has a standardized service interface required to host the operator;

[0136] The second container image building unit is used to obtain a second container image constructed in a form corresponding to the operator model container image by using the operator model container image and the management agent program.

[0137] In one embodiment of this application, the container image building module 601 may further include the following sub-modules:

[0138] The configuration information acquisition submodule is used to acquire requirement parameter information when building a container image for the operator: the requirement parameter information includes configuration information for the operator;

[0139] The parameter configuration submodule is used to configure the parameters of the constructed container image based on the configuration information of the operator.

[0140] In one embodiment of this application, the operator edge hosting system includes an edge cloud platform and a container edge deployment platform, and the container deployment module 603 may include the following sub-modules:

[0141] The container deployment information acquisition submodule is used to acquire requirement parameter information after building a container image for the operator; the requirement parameter information includes container deployment information for the operator on edge nodes;

[0142] The edge node deployment submodule is used to deploy the operator's container to the edge node of the edge cloud platform according to the container deployment information through the container edge deployment platform; the edge node includes a single edge node or multiple different edge nodes.

[0143] In one embodiment of this application, the operator edge hosting system includes an operator center management device, and the container deployment module 603 may include the following sub-modules:

[0144] The operator management submodule is used to uniformly manage the operators hosted on the edge nodes through the operator center management service provided by the operator center management device.

[0145] In this embodiment, a container image is built for the provided operator and the built container image is encapsulated to obtain a container for the operator. This container can encapsulate the operator provided by the user / third party in the form of a container. The encapsulated container is deployed to the edge node, and the operator center management service manages the operators on the edge node in a unified manner. While realizing the edge cloud deployment of operator services, it can also realize the management and control function of operators, so that the operators on the edge cloud node can have the necessary functions for overall operation when running normal services.

[0146] As the device embodiment is basically similar to the method embodiment, the description is relatively simple, and relevant parts can be found in the description of the method embodiment.

[0147] This application also provides an electronic device, including:

[0148] It includes a processor, a memory, and a computer program stored in the memory and capable of running on the processor. When the computer program is executed by the processor, it implements the various processes of the edge processing method embodiment of the above-described operator and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0149] This application also provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the various processes of the edge processing method embodiment of the above-described operator and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0150] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0151] Those skilled in the art will understand that embodiments of this application can be provided as methods, apparatus, or computer program products. Therefore, embodiments of this application can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of this application can take the form of computer program products implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0152] This application describes embodiments with reference to flowchart illustrations and / or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0153] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0154] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal equipment, causing a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0155] Although preferred embodiments of the present application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present application.

[0156] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0157] The foregoing has provided a detailed description of an edge processing method, an edge processing device, a corresponding electronic device, and a corresponding computer storage medium provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. An edge processing method for an operator, characterized in that, Applied to an operator edge hosting system, wherein the operator edge hosting system has an operator center management service, the method includes: Obtain the provided operator, and construct a container image for the operator using a method that matches the operator type; the operator is provided by the operator edge hosting system, user, or third party; The container image of the operator is encapsulated to obtain a container for the operator; the container has the operator's computing power, interface service capabilities, and standardized service capabilities, and the container is injected with standardized interface capabilities, which are used to realize the control capabilities of the operator and the function of linkage with other services of the edge cloud. The operator container is deployed to the edge node, and the operator on the edge node is managed through the operator center management service.

2. The method according to claim 1, characterized in that, The method of constructing a container image for the operator by matching the operator type includes: Obtain the operator type of the provided operator; the operator type includes the form of operator model source file and the form of operator model container image; Build a corresponding container image for the operator based on the operator type.

3. The method according to claim 2, characterized in that, The constructed container image includes a first container image constructed corresponding to the form of the operator model source file; the step of constructing a corresponding container image for the operator according to the operator type includes: When the operator type is in the form of an operator model source file, obtain the model framework information of the operator model source file; The model framework information of the operator is used to generate an image file adapted to the runtime environment; Inject interface services and standardized services required for hosting the operator into the image file; Using the image file, runtime environment deployment file, interface service, and standardized service, a first container image is constructed in the form corresponding to the operator model source file.

4. The method according to claim 2 or 3, characterized in that, The constructed container image includes a second container image constructed in a form corresponding to the operator model container image; the step of constructing a corresponding container image for the operator according to the operator type includes: When the operator type is in the form of an operator model container image, if the operator model container image has a standardized service interface required to host the operator, then a management agent program is created for the operator model container image. Using the operator model container image and the management agent, a second container image is constructed in a form corresponding to the operator model container image.

5. The method according to claim 1 or 2, characterized in that, Also includes: When building a container image for the operator, the requirement parameter information is obtained: the requirement parameter information includes configuration information for the operator; The parameters of the constructed container image are configured based on the configuration information of the operator.

6. The method according to claim 1 or 2, characterized in that, The operator edge hosting system includes an edge cloud platform and a container edge deployment platform. Deploying the operator's container to edge nodes includes: After building the container image for the operator, the requirement parameter information is obtained; the requirement parameter information includes the container deployment information for the operator on the edge node; The container of the operator is deployed to the edge node of the edge cloud platform according to the container deployment information through the container edge deployment platform; the edge node includes a single edge node or multiple different edge nodes.

7. The method according to claim 1 or 2, characterized in that, The operator edge hosting system includes operator center management equipment, and the management of operators at the edge nodes includes: The operators hosted on the edge nodes are managed in a unified manner through the operator center management service provided by the operator center management device.

8. An edge processing device for an operator, characterized in that, An apparatus for use in an operator edge hosting system, wherein the operator edge hosting system has operator center management and control services, the apparatus comprising: The container image building module is used to obtain the provided operator and build a container image for the operator using a method that matches the operator type; the operator is provided by the operator edge hosting system, user, or third party. The container encapsulation module is used to encapsulate the container image of the operator to obtain a container for the operator; the container has the computing power, interface service capabilities and standardized service capabilities of the operator, and the standardized interface capabilities are injected into the container to realize the control capabilities of the operator and the function of linkage with other services of the edge cloud. The container deployment module is used to deploy the operator's container to the edge node and manage the operator on the edge node through the operator central management device.

9. An electronic device, characterized in that, include: A processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein the computer program, when executed by the processor, implements the steps of the edge processing method of the operator as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, which, when executed by a processor, implements the steps of the edge processing method of the operator as described in any one of claims 1 to 7.

Images