A method and system for intelligent capacity planning in a hybrid cloud environment

Abstract

The present invention relates to the field of cloud computing technology, and in particular to a method and system for intelligent capacity planning in a hybrid cloud environment. The method and system for intelligent capacity planning in a hybrid cloud environment utilize a declarative description language to model heterogeneous resources in a hybrid cloud, identify dependencies between resources, and construct a directed acyclic graph (DAG); divide resource operation tasks into batches, and execute tasks without dependencies in parallel; synchronize the operating status of each resource in real time, perform status consistency verification and anomaly marking; trigger a retry mechanism when an anomaly occurs; and after the resource orchestration process is completed, summarize the entire process data, and perform performance evaluation and strategy optimization on resource operations. The method and system for intelligent capacity planning in a hybrid cloud environment can monitor and analyze resource usage in real time, accurately predict resource demand, generate reasonable resource scheduling strategies, reduce manual intervention, and achieve elastic expansion and optimized configuration of cloud resources.

Description

Technical Field

[0001] The present invention relates to the field of cloud computing technology, and in particular to a method and system for intelligent capacity planning in a hybrid cloud environment. Background Art

[0002] With the rapid development of cloud computing technology, more and more enterprises and organizations are choosing to adopt hybrid cloud architecture, combining the advantages of private cloud and public cloud to achieve a balance between flexibility and cost control.

[0003] In hybrid cloud environments, traditional capacity planning methods are no longer able to meet the resource scheduling needs of dynamic environments due to load volatility, uncertainty in resource demand, and differences in resource management across cloud platforms. Traditional capacity planning relies on static rules or manual intervention, making it impossible to adjust cloud resources in real time, resulting in wasted resources or insufficient performance.

[0004] With the development of machine learning, deep learning, and optimization algorithms, intelligent capacity planning has gradually become an effective solution to this problem. To achieve flexible response to load fluctuations, reduce resource waste, and optimize cost-effectiveness, this paper proposes a method and system for intelligent capacity planning in a hybrid cloud environment. Summary of the Invention

[0005] In order to overcome the deficiencies of the prior art, the present invention provides a simple and efficient method and system for intelligent capacity planning in a hybrid cloud environment.

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

[0007] A method for intelligent capacity planning in a hybrid cloud environment includes the following steps:

[0008] Step S1: Resource modeling

[0009] Use a unified declarative description language to model heterogeneous resources in the hybrid cloud, including computing, storage, networking, and middleware. Define the target state, configuration parameters, dependencies, and lifecycle hooks of heterogeneous resources, and accurately describe the target state and configuration intent of resources through structured documents.

[0010] Step S2: Dependency parsing and directed acyclic graph (DAG) generation

[0011] To ensure the order and correctness of complex resources during deployment and change, after parsing the modeling data, the dependencies between resources are automatically identified. Based on these dependencies, a directed acyclic graph (DAG) is constructed using a topological sorting algorithm to clarify the order of resource creation and change.

[0012] Step S3: Task scheduling and execution control

[0013] Based on the directed acyclic graph (DAG), resource operation tasks are divided into batches. Non-dependent tasks are executed in parallel through a distributed scheduler, and the concurrency and execution rhythm of tasks are dynamically adjusted according to system load and cloud platform limitations.

[0014] Step S4: State synchronization and consistency check

[0015] To ensure that the system view is consistent with the actual cloud resource status, the operating status of each resource is synchronized in real time during resource operations. A mechanism combining event-driven and polling is used to obtain status change information, and a state machine is used to perform status consistency verification and anomaly marking.

[0016] Step S5: Anomaly detection and automatic rollback

[0017] When an exception occurs in a resource operation, the exception type is automatically identified, a retry mechanism is triggered for transient exceptions, a rollback process is triggered for serious exceptions, and related resources are destroyed in reverse topological sorting order to ensure the integrity of dependency relationships;

[0018] Step S6: Execution feedback and continuous optimization

[0019] After the resource orchestration process is completed, task execution data is automatically collected, including execution time, failure rate, and retry count indicators. After summarizing the data of the entire process, the built-in analysis engine is used to perform performance evaluation and strategy optimization on resource operations to achieve intelligent evolution.

[0020] In step S1, the data serialization format YAML or JSON format standard language is used to encapsulate the interface parameters of mainstream cloud platforms, support unified description of various types of resources, including computing, network, storage, security group, load balancing, database and middleware, and support model parameterization, templating, public module reference and resource combination modeling;

[0021] Model parameterization and templating refer to achieving high reuse, dynamic adjustment, and flexible expansion of resource models through variable definition, reference to common modules, and nested combination.

[0022] Introducing a resource abstraction adapter to connect to the resource application programming interface (API) of different cloud vendors (including Alibaba Cloud, Huawei Cloud, and Amazon Web Services, etc.), automatically convert resource parameters, and achieve cross-platform modeling and unified delivery;

[0023] Build a version management and audit mechanism to uniformly store all resource description files in a centralized version management system, supporting configuration comparison, change approval, version backtracking, and audit tracking to meet security compliance and team collaboration requirements;

[0024] Supports embedding lifecycle hook functions (pre_create, post_delete), compliance check items, resource quota control, and permission tags in models to enhance the integrity, security, and governance capabilities of resource delivery.

[0025] In step S2, the multi-level dependency identification technology automatically analyzes explicit dependencies and implicit dependencies, constructs a complete dependency graph, uses a topological sorting algorithm to convert the directed acyclic graph (DAG) into a strictly ordered execution plan, and customizes the division of concurrent batches to ensure the correctness of dependency relationships while improving scheduling efficiency.

[0026] During the DAG generation phase, the system automatically identifies critical path nodes that affect the overall execution time, prioritizes their scheduling, optimizes the critical path, and dynamically evaluates task execution costs using a prediction model to optimize the order of resource usage.

[0027] Supports pruning or expanding the directed acyclic graph (DAG) based on the actual operating status of resources to improve processing performance and scenario adaptability;

[0028] The generated directed acyclic graph (DAG) is visualized through a graphical interface for administrators to review, adjust, and simulate the execution path, thereby reducing the risk of execution failure due to modeling errors.

[0029] In step S3, the distributed scheduler executes tasks in batches according to the directed acyclic graph (DAG) topology order, ensuring the order of resource dependencies while executing them in parallel between non-dependent nodes to maximize execution efficiency.

[0030] Based on the cloud platform application programming interface (API) rate limit, resource type processing characteristics, and current system load, the task concurrency and execution rate are adaptively and dynamically adjusted to achieve refined scheduling of resources and interfaces.

[0031] Supports remote deployment of execution nodes. The distributed scheduler dispatches tasks to the nearest cloud platform or region based on the resource, avoiding delays and failures caused by cross-region calls.

[0032] Each resource operation task has an idempotent mechanism, with customizable timeout thresholds and exponential backoff retry strategies to ensure that the task can self-recover from brief exceptions.

[0033] The distributed scheduler has a state feedback capability and can determine in real time whether the task is successfully advanced. If a state blockage or error is found, it will actively trigger a rescheduling and compensation mechanism.

[0034] In step S4, a dual-channel mechanism based on the cloud platform's native event bus and timed polling is adopted to achieve low-latency status updates by subscribing to event notifications from cloud vendors; for resource types that do not support event push, a differentiated polling mechanism is enabled for status compensation;

[0035] Build a standard resource state transition model, clearly define the state paths from "creating" to "running" and from "failed" to "rolling back", and support state-triggered automatic processing;

[0036] Combined with a distributed consistency verification strategy, it compares the resource description model with the actual status fields to identify status drift and configuration deviations, and automatically triggers synchronization, repair, or alarm processes to ensure synchronization timeliness and integrity.

[0037] Use a distributed storage system to store state information to ensure multiple copies of data, transaction consistency, and high concurrent reading and writing;

[0038] Record a snapshot of resource status before a critical operation. If the operation fails, perform reverse operations or compensation processing by comparing the rollback point status to enhance the state consistency closed loop.

[0039] In step S5, the anomalies are divided into transient, structural and strategic categories, and a custom response strategy is formulated for each category;

[0040] Transient anomalies include network jitter and interface congestion, structural anomalies include missing dependencies and parameter errors, and policy anomalies include quota exceeding and insufficient permissions.

[0041] Retry and skip mechanism: For transient exceptions, retries are automatically performed using an exponential backoff strategy. For custom-selected non-critical resources that tolerate failure, a user-configured skip strategy is used to ensure the overall process progress.

[0042] When a task fails and cannot be recovered, the rollback task is executed in the reverse topological order of the directed acyclic graph (DAG) to avoid resource residue and configuration drift caused by uncleared factor resources.

[0043] Supports users to declare rollback hook functions in the model and inject custom recovery scripts or notification mechanisms to implement custom compensation logic to meet differentiated scenarios;

[0044] All exception information, execution stack information, context variables, retry history, and rollback results are written into the audit log to achieve comprehensive audit records for post-problem review and intelligent optimization analysis.

[0045] In step S6, the task execution data collected includes the following dimensions: task completion rate, average / maximum time consumption, concurrent utilization, application programming interface (API) failure rate, retry distribution, failure type distribution and critical path length;

[0046] Based on data analysis, performance bottleneck nodes and high-risk resources are identified, and the directed acyclic graph (DAG) structure and scheduling strategy are automatically adjusted to improve execution stability.

[0047] Use machine learning algorithms to model historical task execution data, build resource behavior models and deployment time estimation models, and support optimal path prediction and concurrency setting before task scheduling;

[0048] Based on historical successful experiences, it automatically recommends the optimal resource combination, application programming interface (API) call parameters, and concurrency control strategies to achieve automatic process tuning. It also automatically writes the optimization results into the model template and scheduling strategy library, forming the system's self-learning and strategy self-evolution capabilities to support business scale expansion.

[0049] A system for intelligent capacity planning in a hybrid cloud environment, used to implement the above method, includes:

[0050] The declarative resource modeling module is responsible for modeling heterogeneous resources in the hybrid cloud, including computing, storage, networking, and middleware, using a unified declarative description language. It defines the target state, configuration parameters, dependencies, and lifecycle hooks of heterogeneous resources, and accurately describes the target state and configuration intent of resources through structured documents.

[0051] The dependency parsing module is responsible for automatically identifying the dependencies between resources after parsing the modeling data;

[0052] The DAG generation module is responsible for constructing a DAG based on the dependencies between resources using a topological sorting algorithm to clarify the order of resource creation and modification.

[0053] The distributed task scheduling module is responsible for batching resource operation tasks based on the directed acyclic graph (DAG) and executing independent tasks in parallel through the distributed scheduler.

[0054] The execution control and state synchronization module is responsible for dynamically adjusting the concurrency and execution rhythm of tasks based on system load and cloud platform limitations. It also synchronizes the operating status of each resource in real time during resource operations. It uses a combination of event-driven and polling mechanisms to obtain state change information and performs state consistency verification and anomaly marking through a state machine.

[0055] The exception handling and automated rollback module is responsible for automatically identifying the exception type when an exception occurs in a resource operation, triggering a retry mechanism for transient exceptions, triggering a rollback process for severe exceptions, and destroying related resources in reverse topological sorting order to ensure the integrity of dependency relationships;

[0056] The execution feedback and optimization module is responsible for automatically collecting task execution data after the resource orchestration process is completed, including execution time, failure rate, and retry count indicators. After summarizing the data of the entire process, it combines the built-in analysis engine to perform performance evaluation and strategy optimization of resource operations to achieve intelligent evolution.

[0057] A device for intelligent capacity planning in a hybrid cloud environment includes a memory and a processor; the memory is used to store a computer program, and the processor is used to implement the above method steps when executing the computer program.

[0058] A readable storage medium stores a computer program, and when the computer program is executed by a processor, the above method steps are implemented.

[0059] The beneficial effects of the present invention are: the method and system for intelligent capacity planning in a hybrid cloud environment can not only monitor and analyze resource usage in real time and accurately predict resource demand, but also generate reasonable resource scheduling strategies based on performance and cost constraints, and reduce manual intervention through automated execution, thereby achieving elastic expansion and optimized configuration of cloud resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] 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.

[0061] Attachment Figure 1 Schematic diagram of the method for intelligent capacity planning in a hybrid cloud environment of the present invention.

[0062] Attachment Figure 2 Schematic diagram of the system architecture for intelligent capacity planning in a hybrid cloud environment according to the present invention. DETAILED DESCRIPTION

[0063] 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.

[0064] By learning from historical data, intelligent capacity planning methods can predict future resource requirements based on load changes and automatically make scaling decisions, thereby improving resource utilization efficiency, reducing operating costs, and ensuring high availability and performance requirements of services.

[0065] The method for intelligent capacity planning in a hybrid cloud environment includes the following steps:

[0066] Step S1: Resource modeling

[0067] Use a unified declarative description language to model heterogeneous resources in the hybrid cloud, including computing, storage, networking, and middleware. Define the target state, configuration parameters, dependencies, and lifecycle hooks of heterogeneous resources, and accurately describe the target state and configuration intent of resources through structured documents.

[0068] In heterogeneous hybrid cloud environments, resource types are numerous, structures are complex, and configuration standards vary. Traditional script-based configuration methods are difficult to adapt to large-scale, rapidly changing resource orchestration needs. The declarative resource modeling approach adopted in step S1 fundamentally improves the portability and consistency of resource management.

[0069] Step S2: Dependency parsing and directed acyclic graph (DAG) generation

[0070] To ensure the order and correctness of complex resources during deployment and change, after parsing the modeling data, the dependencies between resources are automatically identified. Based on these dependencies, a directed acyclic graph (DAG) is constructed using a topological sorting algorithm to clarify the order of resource creation and change.

[0071] Step S3: Task scheduling and execution control

[0072] Based on the directed acyclic graph (DAG), resource operation tasks are divided into batches. Non-dependent tasks are executed in parallel through a distributed scheduler, and the concurrency and execution rhythm of tasks are dynamically adjusted according to system load and cloud platform limitations.

[0073] Step S4: State synchronization and consistency check

[0074] To ensure that the system view is consistent with the actual cloud resource status, the operating status of each resource is synchronized in real time during resource operations. A mechanism combining event-driven and polling is used to obtain status change information, and a state machine is used to perform status consistency verification and anomaly marking.

[0075] Step S5: Anomaly detection and automatic rollback

[0076] When an exception occurs in a resource operation, the exception type is automatically identified, a retry mechanism is triggered for transient exceptions, a rollback process is triggered for serious exceptions, and related resources are destroyed in reverse topological sorting order to ensure the integrity of dependency relationships;

[0077] Step S6: Execution feedback and continuous optimization

[0078] After the resource orchestration process is completed, task execution data is automatically collected, including execution time, failure rate, and retry count indicators. After summarizing the data of the entire process, the built-in analysis engine is used to perform performance evaluation and strategy optimization on resource operations to achieve intelligent evolution.

[0079] In step S1, the data serialization format YAML or JSON format standard language is used to encapsulate the interface parameters of mainstream cloud platforms, support unified description of various types of resources, including computing, network, storage, security group, load balancing, database and middleware, and support model parameterization, templating, public module reference and resource combination modeling;

[0080] Model parameterization and templating refer to achieving high reuse, dynamic adjustment, and flexible expansion of resource models through variable definition, reference to common modules, and nested combination.

[0081] Introducing a resource abstraction adapter to connect to the resource application programming interface (API) of different cloud vendors (including Alibaba Cloud, Huawei Cloud, and Amazon Web Services, etc.), automatically convert resource parameters, and achieve cross-platform modeling and unified delivery;

[0082] Build a version management and audit mechanism to uniformly store all resource description files in a centralized version management system, supporting configuration comparison, change approval, version backtracking, and audit tracking to meet security compliance and team collaboration requirements;

[0083] Supports embedding lifecycle hook functions (pre_create, post_delete), compliance check items, resource quota control, and permission tags in models to enhance the integrity, security, and governance capabilities of resource delivery.

[0084] In step S2, the multi-level dependency identification technology automatically analyzes explicit dependencies and implicit dependencies, constructs a complete dependency graph, uses a topological sorting algorithm to convert the directed acyclic graph (DAG) into a strictly ordered execution plan, and customizes the division of concurrent batches to ensure the correctness of dependency relationships while improving scheduling efficiency.

[0085] During the DAG generation phase, the system automatically identifies critical path nodes that affect the overall execution time, prioritizes their scheduling, optimizes the critical path, and dynamically evaluates task execution costs using a prediction model to optimize the order of resource usage.

[0086] Supports pruning or expanding the directed acyclic graph (DAG) based on the actual operating status of resources to improve processing performance and scenario adaptability;

[0087] The generated directed acyclic graph (DAG) is visualized through a graphical interface for administrators to review, adjust, and simulate the execution path, thereby reducing the risk of execution failure due to modeling errors.

[0088] In step S3, the distributed scheduler executes tasks in batches according to the directed acyclic graph (DAG) topology order, ensuring the order of resource dependencies while executing them in parallel between non-dependent nodes to maximize execution efficiency.

[0089] Based on the cloud platform application programming interface (API) rate limit, resource type processing characteristics, and current system load, the task concurrency and execution rate are adaptively and dynamically adjusted to achieve refined scheduling of resources and interfaces.

[0090] Supports remote deployment of execution nodes. The distributed scheduler dispatches tasks to the nearest cloud platform or region based on the resource, avoiding delays and failures caused by cross-region calls.

[0091] Each resource operation task has an idempotent mechanism, with customizable timeout thresholds and exponential backoff retry strategies to ensure that the task can self-recover from brief exceptions.

[0092] The distributed scheduler has a state feedback capability and can determine in real time whether the task is successfully advanced. If a state blockage or error is found, it will actively trigger a rescheduling and compensation mechanism.

[0093] In step S4, a dual-channel mechanism based on the cloud platform's native event bus and timed polling is adopted to achieve low-latency status updates by subscribing to event notifications from cloud vendors; for resource types that do not support event push, a differentiated polling mechanism is enabled for status compensation;

[0094] Build a standard resource state transition model, clearly define the state paths from "creating" to "running" and from "failed" to "rolling back", and support state-triggered automatic processing;

[0095] Combined with a distributed consistency verification strategy, it compares the resource description model with the actual status fields to identify status drift and configuration deviations, and automatically triggers synchronization, repair, or alarm processes to ensure synchronization timeliness and integrity.

[0096] Use a distributed storage system to store state information to ensure multiple copies of data, transaction consistency, and high concurrent reading and writing;

[0097] Record a snapshot of resource status before a critical operation. If the operation fails, perform reverse operations or compensation processing by comparing the rollback point status to enhance the state consistency closed loop.

[0098] In step S5, the anomalies are divided into transient, structural and strategic categories, and a custom response strategy is formulated for each category;

[0099] Transient anomalies include network jitter and interface congestion, structural anomalies include missing dependencies and parameter errors, and policy anomalies include quota exceeding and insufficient permissions.

[0100] Retry and skip mechanism: For transient exceptions, retries are automatically performed using an exponential backoff strategy. For custom-selected non-critical resources that tolerate failure, a user-configured skip strategy is used to ensure the overall process progress.

[0101] When a task fails and cannot be recovered, the rollback task is executed in the reverse topological order of the directed acyclic graph (DAG) to avoid resource residue and configuration drift caused by uncleared factor resources.

[0102] Supports users to declare rollback hook functions in the model and inject custom recovery scripts or notification mechanisms to implement custom compensation logic to meet differentiated scenarios;

[0103] All exception information, execution stack information, context variables, retry history, and rollback results are written into the audit log to achieve comprehensive audit records for post-problem review and intelligent optimization analysis.

[0104] In step S6, the task execution data collected includes the following dimensions: task completion rate, average / maximum time consumption, concurrent utilization, application programming interface (API) failure rate, retry distribution, failure type distribution and critical path length;

[0105] Based on data analysis, performance bottleneck nodes and high-risk resources are identified, and the directed acyclic graph (DAG) structure and scheduling strategy are automatically adjusted to improve execution stability.

[0106] Use machine learning algorithms to model historical task execution data, build resource behavior models and deployment time estimation models, and support optimal path prediction and concurrency setting before task scheduling;

[0107] Based on historical successful experiences, it automatically recommends the optimal resource combination, application programming interface (API) call parameters, and concurrency control strategies to achieve automatic process tuning. It also automatically writes the optimization results into the model template and scheduling strategy library, forming the system's self-learning and strategy self-evolution capabilities to support business scale expansion.

[0108] The system for intelligent capacity planning in a hybrid cloud environment, used to implement the above method, includes:

[0109] The declarative resource modeling module is responsible for modeling heterogeneous resources in the hybrid cloud, including computing, storage, networking, and middleware, using a unified declarative description language. It defines the target state, configuration parameters, dependencies, and lifecycle hooks of heterogeneous resources, and accurately describes the target state and configuration intent of resources through structured documents.

[0110] The dependency parsing module is responsible for automatically identifying the dependencies between resources after parsing the modeling data;

[0111] The DAG generation module is responsible for constructing a DAG based on the dependencies between resources using a topological sorting algorithm to clarify the order of resource creation and modification.

[0112] The distributed task scheduling module is responsible for batching resource operation tasks based on the directed acyclic graph (DAG) and executing independent tasks in parallel through the distributed scheduler.

[0113] The execution control and state synchronization module is responsible for dynamically adjusting the concurrency and execution rhythm of tasks based on system load and cloud platform limitations. It also synchronizes the operating status of each resource in real time during resource operations. It uses a combination of event-driven and polling mechanisms to obtain state change information and performs state consistency verification and anomaly marking through a state machine.

[0114] The exception handling and automated rollback module is responsible for automatically identifying the exception type when an exception occurs in a resource operation, triggering a retry mechanism for transient exceptions, triggering a rollback process for severe exceptions, and destroying related resources in reverse topological sorting order to ensure the integrity of dependency relationships;

[0115] The execution feedback and optimization module is responsible for automatically collecting task execution data after the resource orchestration process is completed, including execution time, failure rate, and retry count indicators. After summarizing the data of the entire process, it combines the built-in analysis engine to perform performance evaluation and strategy optimization of resource operations to achieve intelligent evolution.

[0116] The device for intelligent capacity planning in a hybrid cloud environment includes a memory and a processor; the memory is used to store a computer program, and the processor is used to implement the above method steps when executing the computer program.

[0117] The readable storage medium stores a computer program, which implements the above method steps when executed by a processor.

[0118] Compared with existing technologies, the method and system for intelligent capacity planning in a hybrid cloud environment have the following characteristics:

[0119] 1) Improve the efficiency and accuracy of flow resource delivery: Through declarative modeling and automatic dependency resolution, resources are automatically processed from definition to deployment, avoiding manual configuration errors and duplication of work, significantly improving the efficiency and accuracy of resource delivery.

[0120] 2) Enhanced compatibility and uniformity across cloud platforms: A unified modeling and scheduling framework was built to shield API differences between different cloud vendors, enabling centralized modeling, unified orchestration, and uniform execution of multi-cloud resources, improving the system's heterogeneous adaptability.

[0121] 3) Achieved high-availability and elastic resource scheduling control: Adopting a distributed task scheduling architecture, it supports highly concurrent and scalable resource creation and change processes, and ensures high availability and stability of resource operations through dynamic concurrency control and exception handling mechanisms.

[0122] 4) Ensures consistent resource status and automatic self-healing: A real-time status synchronization mechanism and consistency verification process are introduced to ensure that the system view is always synchronized with the actual resource status. This supports automatic discovery and repair of status drift issues, significantly improving the reliability of resource management.

[0123] 5) Improved exception handling capabilities and fault recovery efficiency: Equipped with a comprehensive automatic retry, classification processing, and rollback mechanism, it can automatically adopt corresponding strategies based on the type of exception, effectively reducing deployment failure rates, shortening fault recovery time, and improving system robustness.

[0124] 6) Intelligent optimization and continuous evolution of resource management: By collecting and analyzing various types of data during the execution process, dynamically optimizing task scheduling strategies, and forming intelligent recommendations and automatic learning capabilities based on historical data, it can promote the continuous evolution and self-optimization of the resource orchestration system.

[0125] 7) Support closed-loop management of the entire resource lifecycle: This solution covers the entire lifecycle of resources, including definition, deployment, monitoring, exception handling, and optimization feedback, truly achieving end-to-end automated and intelligent closed-loop management from resource intent to final state.

[0126] 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 method for intelligent capacity planning in a hybrid cloud environment, characterized by: The following steps are involved: Step S1: Resource modeling Use a unified declarative description language to model heterogeneous resources in the hybrid cloud, including computing, storage, networking, and middleware. Define the target state, configuration parameters, dependencies, and lifecycle hooks of heterogeneous resources. Describe the target state and configuration intent of resources through structured documents. Step S2: Dependency parsing and directed acyclic graph (DAG) generation To ensure the order and correctness of complex resources during deployment and change, after parsing the modeling data, the dependencies between resources are automatically identified. Based on these dependencies, a directed acyclic graph (DAG) is constructed using a topological sorting algorithm to clarify the order of resource creation and change. Step S3: Task scheduling and execution control Resource operation tasks are batched based on a directed acyclic graph (DAG). Non-dependent tasks are executed in parallel using a distributed scheduler. The concurrency and execution pace of tasks are dynamically adjusted based on system load and cloud platform limitations. The details are as follows: The distributed scheduler executes tasks in batches according to the DAG topology, ensuring the order of resource dependencies while executing them in parallel between independent nodes to maximize execution efficiency. Adaptively and dynamically adjust task concurrency and execution rate based on the cloud platform's API rate limit, resource type processing characteristics, and current system load to achieve refined scheduling of resources and interfaces. Supports remote deployment of execution nodes. The distributed scheduler dispatches tasks to the nearest cloud platform or region based on the resource, avoiding delays and failures caused by cross-region calls. Each resource operation task has an idempotent mechanism, with customizable timeout thresholds and exponential backoff retry strategies to ensure that the task can self-recover from brief exceptions. The distributed scheduler has a status feedback capability and can determine in real time whether the task is successfully advanced. If a status blockage or error is found, it will actively trigger a rescheduling and compensation mechanism. Step S4: State synchronization and consistency check To ensure that the system view is consistent with the actual cloud resource status, the operating status of each resource is synchronized in real time during resource operations. A mechanism combining event-driven and polling is used to obtain status change information, and a state machine is used to perform status consistency verification and anomaly marking. Step S5: Anomaly detection and automatic rollback When an exception occurs in a resource operation, the exception type is automatically identified, a retry mechanism is triggered for transient exceptions, a rollback process is triggered for serious exceptions, and related resources are destroyed in reverse topological sorting order to ensure the integrity of dependency relationships; Step S6: Execution feedback and continuous optimization After the resource orchestration process is completed, task execution data is automatically collected, including execution time, failure rate, and retry count indicators; After aggregating data from the entire process, performance bottleneck nodes and high-risk resources are identified based on data analysis, and the directed acyclic graph (DAG) structure and scheduling strategy are automatically adjusted to improve execution stability. Use machine learning algorithms to model historical task execution data, build resource behavior models and deployment time estimation models, and support optimal path prediction and concurrency setting before task scheduling; Based on historical successful experiences, it automatically recommends the optimal resource combination, application programming interface (API) call parameters, and concurrency control strategies to achieve automatic process tuning. It also automatically writes the optimization results into the model template and scheduling strategy library, forming the system's self-learning and strategy self-evolution capabilities to support business scale expansion.

2. The method for intelligent capacity planning in a hybrid cloud environment according to claim 1, characterized in that: In step S1, the data serialization format YAML or JSON format standard language is used to encapsulate the interface parameters of the cloud platform, support unified description of various types of resources, including computing, network, storage, security group, load balancing, database and middleware, support model parameterization, templating, public module reference and resource combination modeling; Model parameterization and templating refer to the reuse, dynamic adjustment, and flexible expansion of resource models through variable definition, reference to common modules, and nested combination. Introducing resource abstraction adapters to connect to the resource application programming interface (API) interfaces of different cloud vendors, automatically converting resource parameters, and achieving cross-platform modeling and unified delivery; Build a version management and audit mechanism to uniformly store all resource description files in a centralized version management system, supporting configuration comparison, change approval, version backtracking, and audit tracking to meet security compliance and team collaboration requirements; It supports embedding lifecycle hook functions, compliance check items, resource quota control, and permission tags in models to enhance the integrity, security, and governance capabilities of resource delivery.

3. The method for intelligent capacity planning in a hybrid cloud environment according to claim 1, characterized in that: In step S2, the multi-level dependency identification technology automatically analyzes explicit dependencies and implicit dependencies, constructs a complete dependency graph, uses a topological sorting algorithm to convert the directed acyclic graph (DAG) into an ordered execution plan, and customizes the division of concurrent batches to ensure the correctness of dependency relationships while improving scheduling efficiency. During the DAG generation phase, the system automatically identifies critical path nodes that affect the overall execution time, prioritizes their scheduling, optimizes the critical path, and dynamically evaluates task execution costs using a prediction model to optimize the order of resource usage. Supports pruning or expanding the directed acyclic graph (DAG) based on the actual operating status of resources to improve processing performance and scenario adaptability; The generated directed acyclic graph (DAG) is visualized through a graphical interface for administrators to review, adjust, and simulate the execution path, thereby reducing the risk of execution failure due to modeling errors.

4. The method for intelligent capacity planning in a hybrid cloud environment according to claim 1, characterized in that: In step S4, a dual-channel mechanism based on the cloud platform's native event bus and timed polling is adopted to achieve low-latency status updates by subscribing to event notifications from cloud vendors; for resource types that do not support event push, a differentiated polling mechanism is enabled for status compensation; Build a standard resource state transition model, clearly define the state paths from "creating" to "running" and from "failed" to "rolling back", and support state-triggered automatic processing; Combined with a distributed consistency verification strategy, it compares the resource description model with the actual status fields to identify status drift and configuration deviations, and automatically triggers synchronization, repair, or alarm processes to ensure synchronization timeliness and integrity. Use a distributed storage system to store state information to ensure multiple copies of data, transaction consistency, and high concurrent reading and writing; Record a snapshot of resource status before a critical operation. If the operation fails, perform reverse operations or compensation processing by comparing the rollback point status to enhance the state consistency closed loop.

5. The method for intelligent capacity planning in a hybrid cloud environment according to claim 1, wherein: In step S5, the anomalies are divided into transient, structural and strategic categories, and a custom response strategy is formulated for each category; Transient anomalies include network jitter and interface congestion, structural anomalies include missing dependencies and parameter errors, and policy anomalies include quota exceeding and insufficient permissions. Retry and skip mechanism: For transient exceptions, retries are automatically performed using an exponential backoff strategy. For custom-selected non-critical resources that tolerate failure, a user-configured skip strategy is used to ensure the overall process progress. When a task fails and cannot be recovered, the rollback task is executed in the reverse topological order of the directed acyclic graph (DAG) to avoid resource residue and configuration drift caused by uncleared factor resources. Supports users to declare rollback hook functions in the model and inject custom recovery scripts or notification mechanisms to implement custom compensation logic to meet differentiated scenarios; All exception information, execution stack information, context variables, retry history, and rollback results are written into the audit log to achieve comprehensive audit records for post-problem review and intelligent optimization analysis.

6. The method for intelligent capacity planning in a hybrid cloud environment according to claim 1, characterized in that: In step S6, the task execution data collected includes the following dimensions: task completion rate, average / maximum time consumption, concurrent utilization, application programming interface (API) failure rate, retry distribution, failure type distribution and critical path length.

7. A system for intelligent capacity planning in a hybrid cloud environment, characterized by: The method for implementing any one of claims 1 to 6 comprises: The declarative resource modeling module is responsible for modeling heterogeneous resources in the hybrid cloud, including computing, storage, networking, and middleware, using a unified declarative description language. It defines the target state, configuration parameters, dependencies, and lifecycle hooks of heterogeneous resources, and describes the target state and configuration intent of resources through structured documents. The dependency parsing module is responsible for automatically identifying the dependencies between resources after parsing the modeling data; The DAG generation module is responsible for constructing a DAG based on the dependencies between resources using a topological sorting algorithm to clarify the order of resource creation and modification. The distributed task scheduling module is responsible for batching resource operation tasks based on the directed acyclic graph (DAG) and executing independent tasks in parallel through the distributed scheduler. The execution control and state synchronization module is responsible for dynamically adjusting the concurrency and execution rhythm of tasks based on system load and cloud platform limitations. It also synchronizes the operating status of each resource in real time during resource operations. It uses a combination of event-driven and polling mechanisms to obtain state change information and performs state consistency verification and anomaly marking through a state machine. The exception handling and automated rollback module is responsible for automatically identifying the exception type when an exception occurs in a resource operation, triggering a retry mechanism for transient exceptions, triggering a rollback process for severe exceptions, and destroying related resources in reverse topological sorting order to ensure the integrity of dependency relationships; The execution feedback and optimization module is responsible for automatically collecting task execution data after the resource orchestration process is completed, including execution time, failure rate, and retry count indicators. After summarizing the data of the entire process, it combines the built-in analysis engine to perform performance evaluation and strategy optimization of resource operations to achieve intelligent evolution.

8. A device for intelligent capacity planning in a hybrid cloud environment, characterized by: The method comprises a memory and a processor; the memory is used to store a computer program, and the processor is used to implement the method according to any one of claims 1 to 6 when executing the computer program.

9. A readable storage medium, characterized in that: The readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images