A resource orchestration and automated execution method and system based on hybrid cloud architecture

Abstract

The present invention relates to the field of cloud computing technology, and in particular to a resource orchestration and automated execution method and system based on a hybrid cloud architecture. The resource orchestration and automated execution method based on a hybrid cloud architecture pulls or receives monitoring data, performs streaming and batch cleaning, and generates the inputs required for training and reasoning in the feature processing pipeline; deploys the model, and delivers the prediction results to the decision module in real time, which triggers automated execution to scale resources; finally, compares the execution effect with the prediction results to achieve continuous optimization. The resource orchestration and automated execution method and system based on a hybrid cloud architecture realizes cross-cloud resource dependency parsing and task-driven orchestration execution, and also realizes policy-driven dynamic deployment and elastic scaling. It can be combined with artificial intelligence or rule engines to perform intelligent execution optimization and exception handling, thereby improving the resource utilization and operation and maintenance efficiency of the hybrid cloud environment and reducing the risk of human operation.

Description

Technical Field

[0001] The present invention relates to the field of cloud computing technology, and in particular to a resource orchestration and automated execution method and system based on a hybrid cloud architecture. Background Art

[0002] As enterprises deepen their digital transformation, more and more organizations are adopting hybrid cloud architectures, deploying their businesses across diverse infrastructures, including public clouds, private clouds, and edge clouds. However, inconsistent resource interfaces and heterogeneous service models across different cloud platforms create complex challenges in deploying and operating applications across multiple clouds.

[0003] Existing technologies mainly rely on manual operations or static templates for resource orchestration and execution, lack unified cross-cloud scheduling logic and adaptive capabilities, and are unable to meet dynamic business needs and efficient operation and maintenance requirements.

[0004] In order to solve the problems of resource heterogeneity, scheduling separation and low execution efficiency existing in the prior art, the present invention proposes a resource orchestration and automatic execution method and system based on a hybrid cloud architecture. Summary of the Invention

[0005] In order to overcome the deficiencies of the prior art, the present invention provides a simple and efficient resource orchestration and automated execution method and system based on a hybrid cloud architecture.

[0006] The present invention is achieved through the following technical solutions:

[0007] A resource orchestration and automated execution method based on a hybrid cloud architecture. It pulls or receives monitoring data at a custom fixed frequency through unified scheduling or message queues, cleans the data in streaming and batches, and then generates the input required for training and inference in the feature processing pipeline.

[0008] Deploy the model as an online service, deliver prediction results to the decision module in real time, and the decision module triggers automated execution to scale resources.

[0009] Finally, the monitoring module compares the execution results with the prediction results and provides feedback to the model and decision-making components to achieve continuous optimization;

[0010] The entire process described above uses encrypted transmission and centralized key management, and inter-service communication utilizes two-way authentication and fine-grained access control. All operations generate audit logs and are analyzed centrally. Security scans and compliance reviews are performed regularly according to custom cycles, and compliance reports are generated.

[0011] This resource orchestration and automated execution method and system based on hybrid cloud architecture encrypts all data and instructions end-to-end, centrally generates, distributes and rotates keys to achieve encrypted transmission;

[0012] When users access the system, their roles and policies are managed based on the principle of least privilege.

[0013] The following steps are involved:

[0014] Step S1: Perform resource discovery and unified modeling on multiple heterogeneous cloud platforms to form a standard resource abstract model and realize automatic identification of multi-cloud heterogeneous resources;

[0015] The standard resource abstraction model adopts a plug-in abstraction structure to support resource encapsulation and standardized processing of open source cloud platforms OpenStack, VMware, container orchestration system Kubernetes, Alibaba Cloud Aliyun, Amazon Web Services AWS, and Huawei Cloud HuaweiCloud platforms;

[0016] In a hybrid cloud environment, the standard resource abstraction model pulls or receives monitoring data in parallel using polling and event-driven methods. A streaming computing framework is used to clean monitoring data in real time, and historical verification is performed in batches according to custom cycles.

[0017] Based on the cleaned data, we combine streaming computing with batch processing to generate short-term and long-term statistical features, including the mean, maximum value, standard deviation, and growth rate within a moving window. We then normalize, reduce the dimensionality, and automatically filter these features to ensure efficient and redundant model input.

[0018] Step S2: Build a resource orchestration model based on business requirements, analyze resource dependencies and generate an execution graph;

[0019] Concurrently build time series models, ensemble learning models, and deep learning models, learning through a combination of offline large-scale training and online small-batch updates. An automated tuning mechanism is used to optimize hyperparameters, and versions are managed in a model repository. Grayscale verification and traffic segmentation are supported to ensure the stability and accuracy of the resource orchestration model once it is launched.

[0020] When building time series models, ensemble learning models, and deep learning models, it supports visual service design and provides import and export capabilities in YAML and JSON formats. It is compatible with mainstream template standards such as TOSCA and Terraform, which are cloud application topology and orchestration specifications.

[0021] The execution graph supports circular dependency detection and node priority scheduling, and has task failure isolation and local retry mechanisms;

[0022] Step S3: The scheduling engine drives task execution and calls the cross-cloud adapter to send the task to the corresponding cloud platform. The cloud platform delivers the prediction results to the decision module in real time.

[0023] The cross-cloud adapter module supports encapsulating API call parameters and authentication mechanisms through a unified interface, automatically adapting to resource creation, modification, and destruction operations on various cloud platforms;

[0024] Step S4: Optimize and adjust task execution based on a policy rule engine or an artificial intelligence (AI) model;

[0025] The decision-making module uses mathematical programming or self-learning algorithms to calculate the optimal scaling and cross-domain scheduling plan in real time based on prediction results, current resource status, service level requirements, and cost models. It also uses a built-in policy rule engine to handle emergencies and ensure continuous and stable business operations.

[0026] The policy rule engine builds a resource scheduling model based on machine learning algorithms, dynamically predicts resource bottlenecks and adjusts the task scheduling order or allocates cloud regions;

[0027] Using declarative configuration management and automated deployment practices, all resource changes are stored in the configuration repository as versioned configurations. Automated deployment tools synchronize resource changes to the runtime environment, enabling auditable and rollback-capable scaling and migration operations.

[0028] Step S5: Monitor the task status in real time and execute the automatic rollback strategy when an exception occurs;

[0029] After each scaling, key indicators are monitored in the short term, and actual results are compared with predicted results in multiple dimensions. The alarm system triggers model retraining or strategy adjustments, and core data is displayed in a visual dashboard to help the operations and algorithm teams respond promptly.

[0030] The automatic rollback strategy supports any combination of two or three strategies based on snapshot recovery, resource reconstruction, and state migration to ensure the eventual consistency of task execution and business continuity;

[0031] Step S6: Report the execution results to the management platform and generate an audit log.

[0032] In step S1, in a hybrid cloud environment, by deploying lightweight collection components and calling cloud platform monitoring interfaces, multi-dimensional indicators are aggregated into a message queue, written into a time series database in real time, and regularly archived to a big data storage layer, supporting online real-time query and offline batch analysis;

[0033] The acquisition component is deployed in a cluster and supports active / standby switching;

[0034] The time series database supports high compression and drill-down queries, and can meet real-time reading and writing requirements;

[0035] Schedule batch jobs based on custom cycles to write data into the big data storage layer and convert it into a columnar format, enabling data archiving into the lake.

[0036] Centrally manage data tables and field information for offline training.

[0037] In step S1, in feature engineering, statistical indicators within several time windows are calculated in real time and at regular intervals, and the timing period is user-defined;

[0038] Normalize or standardize the features to eliminate differences in different dimensions and improve the convergence speed of the model;

[0039] Dimensionality reduction technology is used to eliminate redundant features, and customized core features are automatically retained based on feature importance measurement.

[0040] In step S2, the resource orchestration model is updated in customized small batch increments using real-time collected data. The resource orchestration model parameters are automatically optimized based on the search strategy, and the parameters, indicators, and training records of each version of the resource orchestration model are tracked. Grayscale verification is supported to compare the effects of the new and old models within a customized small range before full deployment.

[0041] In step S4, the decision module uses a mathematical programming algorithm to establish and solve a resource allocation optimization model, and uses an adaptive algorithm to optimize the cross-domain scheduling strategy;

[0042] The policy rule engine executes corresponding preset emergency strategies for sudden increases or abnormal situations, and updates performance and cost weights online according to business priorities to achieve dynamic weight adjustment.

[0043] In step S4, when managing the configuration repository, all resource definitions are stored in a versioned manner, and the automatic deployment tool continuously monitors configuration repository changes and synchronously executes resource orchestration; supports idempotent operations to ensure that the results of each automatic deployment are consistent without intermediate side effects; supports smooth scaling and version switching without business interruption, and achieves seamless updates.

[0044] In step S5, the actual effect is compared with the predicted result in multiple dimensions, and the prediction error, cost deviation and performance change are calculated. If the alarm system is triggered, the relevant personnel or components are notified, the model and decision parameters are dynamically updated through the open interface, and the operation and algorithm indicators are presented in real time through the dashboard.

[0045] A resource orchestration and automated execution system based on a hybrid cloud architecture is used to implement the above method. It adopts a microservice design and is divided into five parts: data layer, intelligent analysis layer, decision layer, execution layer, and monitoring feedback layer.

[0046] The data layer and the intelligent analysis layer are decoupled through message queues, and the remaining layers are decoupled through service meshes.

[0047] The data layer is responsible for data collection and preprocessing;

[0048] The intelligent analysis layer is responsible for implementing feature engineering services;

[0049] The decision-making layer is responsible for implementing model training and evaluation services;

[0050] The execution layer is responsible for implementing automatic deployment and execution services;

[0051] The monitoring feedback layer is responsible for real-time monitoring of task status and executing automatic rollback strategies when an exception occurs, triggering model retraining or strategy adjustment.

[0052] It also includes a configuration center and a service registration center;

[0053] The configuration center is responsible for centralized and unified storage and distribution of global operating parameters, and the service registration center is responsible for managing each module instance, including discovering new services and service health checks, to ensure high availability and scalability of the system.

[0054] A resource orchestration and automated execution device based on a hybrid cloud architecture includes a memory and a processor; the memory is used to store a computer program, and the processor is used to implement the above-mentioned method steps when executing the computer program.

[0055] A readable storage medium stores a computer program, and when the computer program is executed by a processor, the above method steps are implemented.

[0056] The beneficial effects of the present invention are: the resource orchestration and automated execution method and system based on hybrid cloud architecture supports automatic identification and standardized modeling of multi-cloud heterogeneous resources, realizes cross-cloud resource dependency analysis and task-driven orchestration execution, and also realizes policy-driven dynamic deployment and elastic scaling. It can be combined with artificial intelligence or rule engines for intelligent execution optimization and exception handling, thereby improving resource utilization and operation and maintenance efficiency of hybrid cloud environments and reducing the risks of human operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0058] Attachment Figure 1 Schematic diagram of the resource orchestration and automated execution system based on hybrid cloud architecture of the present invention.

[0059] Attachment Figure 2 Schematic diagram of the resource orchestration and automated execution method based on hybrid cloud architecture of the present invention. DETAILED DESCRIPTION

[0060] In order to enable those skilled in the art to better understand the technical solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work should fall within the scope of protection of the present invention.

[0061] As attached Figure 1 As shown in the figure, this resource orchestration and automated execution method based on a hybrid cloud architecture pulls or receives monitoring data at a custom fixed frequency through unified scheduling or message queues, then cleans the data in streaming and batches, and then generates the input required for training and inference in the feature processing pipeline;

[0062] Deploy the model as an online service, deliver prediction results to the decision module in real time, and the decision module triggers automated execution to scale resources.

[0063] Finally, the monitoring module compares the execution results with the prediction results and provides feedback to the model and decision-making components to achieve continuous optimization;

[0064] The entire process described above uses encrypted transmission and centralized key management, and inter-service communication utilizes two-way authentication and fine-grained access control. All operations generate audit logs and are analyzed centrally. Security scans and compliance reviews are performed regularly according to custom cycles, and compliance reports are generated.

[0065] This resource orchestration and automated execution method based on a hybrid cloud architecture encrypts all data and instructions end-to-end, centrally generates, distributes, and rotates keys to achieve encrypted transmission;

[0066] When users access the system, their roles and policies are managed based on the principle of least privilege.

[0067] The following steps are involved:

[0068] Step S1: Perform resource discovery and unified modeling on multiple heterogeneous cloud platforms to form a standard resource abstract model and realize automatic identification of multi-cloud heterogeneous resources;

[0069] The standard resource abstraction model adopts a plug-in abstraction structure to support resource encapsulation and standardized processing of open source cloud platforms OpenStack, VMware, container orchestration system Kubernetes, Alibaba Cloud Aliyun, Amazon Web Services AWS, and Huawei Cloud HuaweiCloud platforms;

[0070] In a hybrid cloud environment, the standard resource abstraction model pulls or receives monitoring data in parallel using polling and event-driven methods. A streaming computing framework is used to clean monitoring data in real time, and historical verification is performed in batches according to custom cycles.

[0071] Based on the cleaned data, we combine streaming computing with batch processing to generate short-term and long-term statistical features, including the mean, maximum value, standard deviation, and growth rate within a moving window. We then normalize, reduce the dimensionality, and automatically filter these features to ensure efficient and redundant model input.

[0072] Step S2: Build a resource orchestration model based on business requirements, analyze resource dependencies and generate an execution graph;

[0073] Concurrently build time series models, ensemble learning models, and deep learning models, learning through a combination of offline large-scale training and online small-batch updates. An automated tuning mechanism is used to optimize hyperparameters, and versions are managed in a model repository. Grayscale verification and traffic segmentation are supported to ensure the stability and accuracy of the resource orchestration model once it is launched.

[0074] When building time series models, ensemble learning models, and deep learning models, it supports visual service design and provides import and export capabilities in YAML and JSON formats. It is compatible with mainstream template standards such as TOSCA and Terraform, which are cloud application topology and orchestration specifications.

[0075] The execution graph supports circular dependency detection and node priority scheduling, and has task failure isolation and local retry mechanisms;

[0076] Step S3: The scheduling engine drives task execution and calls the cross-cloud adapter to send the task to the corresponding cloud platform. The cloud platform delivers the prediction results to the decision module in real time.

[0077] The cross-cloud adapter module supports encapsulating API call parameters and authentication mechanisms through a unified interface, automatically adapting to resource creation, modification, and destruction operations on various cloud platforms;

[0078] Step S4: Optimize and adjust task execution based on a policy rule engine or an artificial intelligence (AI) model;

[0079] The decision-making module uses mathematical programming or self-learning algorithms to calculate the optimal scaling and cross-domain scheduling plan in real time based on prediction results, current resource status, service level requirements, and cost models. It also uses a built-in policy rule engine to handle emergencies and ensure continuous and stable business operations.

[0080] The policy rule engine builds a resource scheduling model based on machine learning algorithms, dynamically predicts resource bottlenecks and adjusts the task scheduling order or allocates cloud regions;

[0081] Using declarative configuration management and automated deployment practices, all resource changes are stored in the configuration repository as versioned configurations. Automated deployment tools synchronize resource changes to the runtime environment, enabling auditable and rollback-capable scaling and migration operations.

[0082] Step S5: Monitor the task status in real time and execute the automatic rollback strategy when an exception occurs;

[0083] After each scaling, key indicators are monitored in the short term, and actual results are compared with predicted results in multiple dimensions. The alarm system triggers model retraining or strategy adjustments, and core data is displayed in a visual dashboard to help the operations and algorithm teams respond promptly.

[0084] The automatic rollback strategy supports any combination of two or three strategies based on snapshot recovery, resource reconstruction, and state migration to ensure the eventual consistency of task execution and business continuity;

[0085] Step S6: Report the execution results to the management platform and generate an audit log.

[0086] In step S1, in a hybrid cloud environment, by deploying lightweight collection components and calling cloud platform monitoring interfaces, multi-dimensional indicators are aggregated into a message queue, written into a time series database in real time, and regularly archived to a big data storage layer, supporting online real-time query and offline batch analysis;

[0087] The acquisition component is deployed in a cluster and supports active / standby switching;

[0088] The time series database supports high compression and drill-down queries, and can meet real-time reading and writing requirements;

[0089] Schedule batch jobs based on custom cycles to write data into the big data storage layer and convert it into a columnar format, enabling data archiving into the lake.

[0090] Centrally manage data tables and field information for offline training.

[0091] In step S1, in feature engineering, statistical indicators within several time windows are calculated in real time and at regular intervals, and the timing period is user-defined;

[0092] Normalize or standardize the features to eliminate differences in different dimensions and improve the convergence speed of the model;

[0093] Dimensionality reduction technology is used to eliminate redundant features, and customized core features are automatically retained based on feature importance measurement.

[0094] In step S2, the resource orchestration model is updated in customized small batch increments using real-time collected data. The resource orchestration model parameters are automatically optimized based on the search strategy, and the parameters, indicators, and training records of each version of the resource orchestration model are tracked. Grayscale verification is supported to compare the effects of the new and old models within a customized small range before full deployment.

[0095] In step S4, the decision module uses a mathematical programming algorithm to establish and solve a resource allocation optimization model, and uses an adaptive algorithm to optimize the cross-domain scheduling strategy;

[0096] The policy rule engine executes corresponding preset emergency strategies for sudden increases or abnormal situations, and updates performance and cost weights online according to business priorities to achieve dynamic weight adjustment.

[0097] In step S4, when managing the configuration repository, all resource definitions are stored in a versioned manner, and the automatic deployment tool continuously monitors configuration repository changes and synchronously executes resource orchestration; supports idempotent operations to ensure that the results of each automatic deployment are consistent without intermediate side effects; supports smooth scaling and version switching without business interruption, and achieves seamless updates.

[0098] In step S5, the actual effect is compared with the predicted result in multiple dimensions, and the prediction error, cost deviation and performance change are calculated. If the alarm system is triggered, the relevant personnel or components are notified, the model and decision parameters are dynamically updated through the open interface, and the operation and algorithm indicators are presented in real time through the dashboard.

[0099] As attached Figure 2 As shown in the figure, the resource orchestration and automated execution system based on hybrid cloud architecture is used to implement the above method. It adopts microservice design and is divided into five parts: data layer, intelligent analysis layer, decision layer, execution layer and monitoring feedback layer.

[0100] The data layer and the intelligent analysis layer are decoupled through message queues, and the remaining layers are decoupled through service meshes.

[0101] The data layer is responsible for data collection and preprocessing;

[0102] The intelligent analysis layer is responsible for implementing feature engineering services;

[0103] The decision-making layer is responsible for implementing model training and evaluation services;

[0104] The execution layer is responsible for implementing automatic deployment and execution services;

[0105] The monitoring feedback layer is responsible for real-time monitoring of task status and executing automatic rollback strategies when an exception occurs, triggering model retraining or strategy adjustment.

[0106] It also includes a configuration center and a service registration center;

[0107] The configuration center is responsible for centralized and unified storage and distribution of global operating parameters, and the service registration center is responsible for managing each module instance, including discovering new services and service health checks, to ensure high availability and scalability of the system.

[0108] The resource orchestration and automatic execution device based on hybrid cloud architecture includes a memory and a processor; the memory is used to store a computer program, and the processor is used to implement the above-mentioned method steps when executing the computer program.

[0109] The readable storage medium stores a computer program, which implements the above method steps when executed by a processor.

[0110] This resource orchestration and automated execution method and system based on a hybrid cloud architecture innovatively combines integrated stream and batch processing, model hybrid prediction, self-learning decision-making, and versioned orchestration. During online inference, it is called through a service interface to ensure millisecond-level response, taking into account both real-time and global perspectives. By comparing prediction errors and execution costs, it automatically adjusts subsequent process parameters, realizing a fully automatic closed loop of prediction-optimization-execution-feedback.

[0111] In addition, multi-model fusion and adaptive weighting ensure high-precision predictions of various loads. By combining self-learning algorithms with mathematical programming, a dynamic balance between cross-domain costs and performance is achieved. Versioned automated orchestration ensures traceability and high reliability of operations.

[0112] The embodiment described above is only one specific implementation of the present invention. Common changes and substitutions made by those skilled in the art within the scope of the technical solution of the present invention should be included in the protection scope of the present invention.

Claims

1. A resource orchestration and automated execution method based on a hybrid cloud architecture, characterized by: Pull or receive monitoring data at a custom fixed frequency through unified scheduling or message queues, clean the data in streaming and batches, and then generate the input required for training and inference in the feature processing pipeline; Deploy the model as an online service, deliver prediction results to the decision module in real time, and the decision module triggers automated execution to scale resources. Finally, the monitoring module compares the execution results with the prediction results and provides feedback to the model and decision-making components to achieve continuous optimization; The entire process uses encrypted transmission and centralized key management, and inter-service communication utilizes two-way authentication and fine-grained access control. All operations generate audit logs and are centrally analyzed. Security scans and compliance reviews are conducted regularly according to customizable cycles, and compliance reports are generated. The following steps are involved: Step S1: Perform resource discovery and unified modeling on multiple heterogeneous cloud platforms to form a standard resource abstract model and realize automatic identification of multi-cloud heterogeneous resources; The standard resource abstraction model adopts a plug-in abstraction structure to support resource encapsulation and standardized processing of open source cloud platforms OpenStack, VMware, container orchestration system Kubernetes, Alibaba Cloud Aliyun, Amazon Web Services AWS, and Huawei Cloud HuaweiCloud platforms; In a hybrid cloud environment, the standard resource abstraction model pulls or receives monitoring data in parallel using polling and event-driven methods. A streaming computing framework is used to clean monitoring data in real time, and historical verification is performed in batches according to custom cycles. Based on the cleaned data, we combine streaming computing with batch processing to generate short-term and long-term statistical features, including the mean, maximum value, standard deviation, and growth rate within a moving window. We then normalize, reduce the dimensionality, and automatically filter these features to ensure efficient and redundant model input. Step S2: Build a resource orchestration model based on business requirements, analyze resource dependencies and generate an execution graph; Concurrently build time series models, ensemble learning models, and deep learning models, learning through a combination of offline large-scale training and online small-batch updates. An automated tuning mechanism is used to optimize hyperparameters, and versions are managed in a model repository. Grayscale verification and traffic segmentation are supported to ensure the stability and accuracy of the resource orchestration model once it is launched. When building time series models, ensemble learning models, and deep learning models, it supports visual service design, provides import and export capabilities in YAML and JSON formats, and is compatible with TOSCA templates and Terraform template standards. The execution graph supports circular dependency detection and node priority scheduling, and has task failure isolation and local retry mechanisms; Step S3: The scheduling engine drives task execution and calls the cross-cloud adapter to send the task to the corresponding cloud platform. The cloud platform delivers the prediction results to the decision module in real time. The cross-cloud adapter module supports encapsulating API call parameters and authentication mechanisms through a unified interface, automatically adapting to resource creation, modification, and destruction operations on various cloud platforms; Step S4: Optimize and adjust task execution based on a policy rule engine or an artificial intelligence (AI) model; The decision-making module uses mathematical programming or self-learning algorithms to calculate the optimal scaling and cross-domain scheduling plan in real time based on prediction results, current resource status, service level requirements, and cost models. It also uses a built-in policy rule engine to handle emergencies and ensure continuous and stable business operations. The policy rule engine builds a resource scheduling model based on machine learning algorithms, dynamically predicts resource bottlenecks and adjusts the task scheduling order or allocates cloud regions; Using declarative configuration management and automated deployment practices, all resource changes are stored in the configuration repository as versioned configurations. Automated deployment tools synchronize resource changes to the runtime environment, enabling auditable and rollback-capable scaling and migration operations. Step S5: Monitor the task status in real time and execute the automatic rollback strategy when an exception occurs; After each scaling operation, key indicators are monitored and actual results are compared with predicted results in multiple dimensions. The alarm system triggers model retraining or strategy adjustments, and core data is displayed in a visual dashboard to help the operations and algorithm teams respond promptly. The automatic rollback strategy supports any combination of two or three strategies based on snapshot recovery, resource reconstruction, and state migration to ensure the eventual consistency of task execution and business continuity; Step S6: Report the execution results to the management platform and generate an audit log.

2. The resource orchestration and automated execution method based on hybrid cloud architecture according to claim 1, characterized in that: All data and instructions are encrypted end-to-end, with centralized key generation, distribution, and rotation to achieve encrypted transmission. When users access the system, their roles and policies are managed based on the principle of least privilege.

3. The resource orchestration and automated execution method based on hybrid cloud architecture according to claim 1, characterized in that: In step S1, in a hybrid cloud environment, by deploying lightweight collection components and calling cloud platform monitoring interfaces, multi-dimensional indicators are aggregated into a message queue, written into a time series database in real time, and regularly archived to a big data storage layer, supporting online real-time query and offline batch analysis; The acquisition component is deployed in a cluster and supports active / standby switching; The time series database supports high compression and drill-down queries, and can meet real-time reading and writing requirements; Schedule batch jobs based on custom cycles to write data into the big data storage layer and convert it into a columnar format, enabling data archiving into the lake. Centrally manage data tables and field information for offline training.

4. The resource orchestration and automated execution method based on hybrid cloud architecture according to claim 1, characterized in that: In step S1, in feature engineering, statistical indicators within several time windows are calculated in real time and at regular intervals, and the timing period is user-defined; Normalize or standardize the features to eliminate differences in different dimensions and improve the convergence speed of the model; Dimensionality reduction technology is used to eliminate redundant features, and customized core features are automatically retained based on feature importance measurement.

5. The resource orchestration and automated execution method based on hybrid cloud architecture according to claim 1, characterized in that: In step S2, the resource orchestration model is updated in a customized small batch incremental manner using real-time collected data, the resource orchestration model parameters are automatically optimized based on the search strategy, and the parameters, indicators, and training records of each version of the resource orchestration model are tracked; Supports grayscale verification, comparing the effects of new and old models in a small, customized range before full deployment.

6. The resource orchestration and automated execution method based on hybrid cloud architecture according to claim 1, characterized in that: In step S4, the decision module uses a mathematical programming algorithm to establish and solve a resource allocation optimization model, and uses an adaptive algorithm to optimize the cross-domain scheduling strategy; The policy rule engine executes corresponding preset emergency strategies for sudden increases or abnormal situations, and updates performance and cost weights online according to business priorities to achieve dynamic weight adjustment.

7. The resource orchestration and automated execution method based on hybrid cloud architecture according to claim 1, characterized in that: In step S4, when managing the configuration repository, all resource definitions are stored in a versioned manner, and the automatic deployment tool continuously monitors configuration repository changes and synchronously executes resource orchestration; It supports idempotent operations to ensure consistent results for each automatic deployment without intermediate side effects. It also supports smooth scaling and version switching without business interruption, enabling seamless updates.

8. The resource orchestration and automated execution method based on hybrid cloud architecture according to claim 1, characterized in that: In step S5, the actual effect is compared with the predicted result in multiple dimensions, and the prediction error, cost deviation and performance change are calculated. If the alarm system is triggered, the relevant personnel or components are notified, the model and decision parameters are dynamically updated through the open interface, and the operation and algorithm indicators are presented in real time through the dashboard.

9. A resource orchestration and automated execution system based on a hybrid cloud architecture, characterized by: Used to implement the method according to any one of claims 1 to 8, the method adopts a microservice design and is divided into five parts: data layer, intelligent analysis layer, decision layer, execution layer and monitoring feedback layer; The data layer and the intelligent analysis layer are decoupled through message queues, and the remaining layers are decoupled through service meshes. The data layer is responsible for data collection and preprocessing; The intelligent analysis layer is responsible for implementing feature engineering services; The decision-making layer is responsible for implementing model training and evaluation services; The execution layer is responsible for implementing automatic deployment and execution services; The monitoring feedback layer is responsible for real-time monitoring of task status and executing automatic rollback strategies when an exception occurs, triggering model retraining or strategy adjustment. It also includes a configuration center and a service registration center; The configuration center is responsible for centralized and unified storage and distribution of global operating parameters, and the service registration center is responsible for managing each module instance, including discovering new services and service health checks, to ensure high availability and scalability of the system.

Images