Multi-dimensional resource dependency graph cross-cloud collaborative scheduling method and device, equipment and storage medium

Abstract

The application discloses a multi-dimensional resource dependency graph cross-cloud collaborative scheduling method and device, equipment and a storage medium, relates to the technical field of cross-cloud resource scheduling and management, and comprises the following steps: acquiring a cross-cloud resource dependency graph and a business scheduling demand parameter value; finding a resource combination satisfying the business scheduling demand parameter value in the cross-cloud resource dependency graph based on a path search algorithm to obtain candidate resource information, wherein the candidate resource information comprises regional preference resource information, performance index resource information and cost constraint resource information; generating corresponding resource scheduling configurations according to the regional preference resource information, the performance index resource information and the cost constraint resource information, determining scheduling strategy information; based on the scheduling strategy information, instantiating, creating entity resources or adjusting entity resources, generating cross-cloud collaborative resource scheduling information, and completing cross-cloud collaborative scheduling based on the cross-cloud collaborative resource scheduling information. Efficient scheduling of cross-cloud resources is realized.

Description

Technical Field

[0001] This application relates to the field of cross-cloud resource scheduling and management technology, and in particular to cross-cloud collaborative scheduling methods, devices, equipment and storage media based on multi-dimensional resource dependency graphs. Background Technology

[0002] With the increasing adoption of multi-cloud strategies by enterprises, the demand for cross-cloud resource scheduling has surged. In a multi-cloud environment, resources are distributed across heterogeneous platforms from different vendors, requiring collaboration to complete complex business processes.

[0003] Currently, cross-cloud resource scheduling relies on manual dependency configuration, with cross-cloud resource dependencies being manually recorded using Excel or documents. However, this approach is costly in terms of heterogeneous resource interoperability, requiring manual adaptation of cross-cloud permission configuration and network connectivity, which is prone to configuration errors and has a long debugging cycle. Furthermore, the implicit nature of dependencies leads to blind scheduling, which is based solely on resource performance indicators without considering dependency constraints such as permissions, network, and data synchronization, resulting in cross-cloud call failures. In addition, there is a lack of real-time awareness of cross-cloud resource status changes, leading to long fault recovery times. Therefore, how to achieve efficient collaborative scheduling of cross-cloud resources and improve resource utilization, response efficiency, and cost control capabilities has become an urgent problem to be solved.

[0004] The above content is only used to help understand the technical solution of this application and does not represent an admission that the above content is prior art. Summary of the Invention

[0005] The main objective of this application is to provide a cross-cloud collaborative scheduling method, apparatus, device, and storage medium based on a multi-dimensional resource dependency graph, aiming to solve the technical problem of how to achieve efficient collaborative scheduling of cross-cloud resources and improve resource utilization, response efficiency, and cost control capabilities.

[0006] To achieve the above objectives, this application proposes a cross-cloud collaborative scheduling method based on a multi-dimensional resource dependency graph. The method includes:

[0007] Obtain cross-cloud resource dependency graphs and business scheduling requirement parameter values;

[0008] Based on the path search algorithm, resource combinations that meet the service scheduling requirement parameter values ​​in the cross-cloud resource dependency graph are found to obtain candidate resource information, which includes regional preference resource information, performance indicator resource information and cost constraint resource information.

[0009] Based on the regional preference resource information, the performance index resource information, and the cost constraint resource information, a corresponding resource scheduling configuration is generated, and scheduling strategy information is determined.

[0010] Instantiate based on the scheduling strategy information, create entity resources or adjust entity resources, generate cross-cloud collaborative resource scheduling information, and complete cross-cloud collaborative scheduling based on the cross-cloud collaborative resource scheduling information.

[0011] In one embodiment, the step of obtaining the cross-cloud resource dependency graph includes:

[0012] Obtain resource metadata information;

[0013] Map the target instance resources in the resource metadata information to the corresponding entity node resources to determine the mapped entity node information;

[0014] Based on the mapped entity node information, a corresponding target dependency relationship is constructed, and dependency edge information is determined. The dependency edge information includes network dependency edge information, permission dependency edge information, and traffic dependency edge information.

[0015] The mapping entity node information and the dependency edge information are stored in a preset graph database to generate a cross-cloud resource dependency graph.

[0016] In one embodiment, the step of storing the mapped entity node information and the dependency edge information into a preset graph database to generate a cross-cloud resource dependency graph includes:

[0017] Obtain monitoring feedback information;

[0018] Based on the monitoring feedback information, abnormal resource load and abnormal dependency edges are detected, and the detection results are determined.

[0019] Based on the detection results, update the attribute values ​​of the mapped entity node information and the attribute values ​​of the dependent edge information to determine the update information;

[0020] The updated information is stored in a preset graph database to obtain a cross-cloud resource dependency graph.

[0021] In one embodiment, the step of finding resource combinations in the cross-cloud resource dependency graph that satisfy the service scheduling requirement parameter values ​​based on the path search algorithm to obtain candidate resource information includes:

[0022] Based on the path search algorithm, find the resource combinations in the cross-cloud resource dependency graph that meet the service scheduling requirement parameter values, and determine the resource combinations to be detected;

[0023] Dependency conflicts are detected for each resource combination in the resource combination to be detected, and the dependency conflict detection results are determined.

[0024] Based on the dependency conflict detection results, target resources without dependency conflicts in the resource combination to be detected are selected to obtain candidate resource information.

[0025] In one embodiment, after the step of finding resource combinations that satisfy the service scheduling requirement parameter values ​​in the cross-cloud resource dependency graph based on the path search algorithm to obtain candidate resource information, the method further includes:

[0026] Calculate the corresponding objective function value based on the objective optimization function and the candidate resource information, and determine the overall function value;

[0027] Based on the overall function value, the resource combinations that meet the service scheduling requirement parameter values ​​are sorted, and the resource combination with the highest ranking is taken as the candidate resource information.

[0028] In one embodiment, the step of calculating the corresponding objective function value based on the objective optimization function and the candidate resource information, and determining the overall function value, includes:

[0029] The objective function value corresponding to each resource combination in the candidate resource information is calculated based on the objective optimization function, and the resource utilization function value, response time function value, and cost function value are determined.

[0030] The overall function value is obtained based on the resource utilization function value, the response time function value, and the cost function value.

[0031] In one embodiment, the step of instantiating, creating or adjusting entity resources based on the scheduling policy information, and generating cross-cloud collaborative resource scheduling information includes:

[0032] Obtain interoperability resource scheduling information;

[0033] Based on the scheduling strategy information, entity resources are instantiated and created to determine the first cross-cloud collaborative resource scheduling information;

[0034] Based on the interoperability resource scheduling information, adjust the entity resources in the first cross-cloud collaborative resource scheduling information to determine the second cross-cloud collaborative resource scheduling information.

[0035] Cross-cloud collaborative resource scheduling information is obtained based on the first cross-cloud collaborative resource scheduling information and the second cross-cloud collaborative resource scheduling information.

[0036] Furthermore, to achieve the above objectives, this application also proposes a cross-cloud collaborative scheduling device for a multi-dimensional resource dependency graph, wherein the cross-cloud collaborative scheduling device for the multi-dimensional resource dependency graph includes:

[0037] The acquisition module is used to obtain cross-cloud resource dependency graphs and business scheduling requirement parameter values;

[0038] The processing module is used to find resource combinations that meet the service scheduling requirement parameter values ​​in the cross-cloud resource dependency graph based on the path search algorithm, and obtain candidate resource information, which includes regional preference resource information, performance index resource information and cost constraint resource information.

[0039] The processing module is also used to generate corresponding resource scheduling configurations based on the regional preference resource information, the performance index resource information, and the cost constraint resource information, and to determine scheduling strategy information;

[0040] The execution module is used to instantiate, create or adjust entity resources based on the scheduling strategy information, generate cross-cloud collaborative resource scheduling information, and complete cross-cloud collaborative scheduling based on the cross-cloud collaborative resource scheduling information.

[0041] Furthermore, to achieve the above objectives, this application also proposes a cross-cloud collaborative scheduling device for a multi-dimensional resource dependency graph. The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor. The computer program is configured to implement the steps of the cross-cloud collaborative scheduling method for the multi-dimensional resource dependency graph as described above.

[0042] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the cross-cloud collaborative scheduling method for the multi-dimensional resource dependency graph described above.

[0043] One or more technical solutions proposed in this application have at least the following technical effects:

[0044] This embodiment proposes a cross-cloud collaborative scheduling method based on a multi-dimensional resource dependency graph. The method involves obtaining a cross-cloud resource dependency graph and business scheduling requirement parameter values; using a path search algorithm to find resource combinations in the cross-cloud resource dependency graph that satisfy the business scheduling requirement parameter values, obtaining candidate resource information, including regional preference resource information, performance indicator resource information, and cost constraint resource information; generating corresponding resource scheduling configurations based on the regional preference resource information, performance indicator resource information, and cost constraint resource information, and determining scheduling strategy information; instantiating entity resources based on the scheduling strategy information, creating or adjusting entity resources, generating cross-cloud collaborative resource scheduling information, and completing cross-cloud collaborative scheduling based on the cross-cloud collaborative resource scheduling information. This application constructs a multi-dimensional resource dependency graph to explicitly represent cross-cloud resource dependencies, filters the dependency graph through business requests to determine candidate resource information, thereby creating or adjusting entity resources, achieving efficient cross-cloud resource collaborative scheduling, improving the efficiency of cross-cloud resource configuration, significantly shortening configuration time, reducing costs in multi-cloud environments through intelligent optimization strategies, and significantly enhancing system reliability and dynamic adaptability. Attached Figure Description

[0045] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0046] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is a flowchart illustrating an embodiment of the cross-cloud collaborative scheduling method for multi-dimensional resource dependency graphs in this application.

[0048] Figure 2 This is a schematic diagram of the cross-cloud resource dependency graph for the multi-dimensional resource dependency graph cross-cloud collaborative scheduling method of this application.

[0049] Figure 3 This is a flowchart illustrating Embodiment 2 of the cross-cloud collaborative scheduling method for multi-dimensional resource dependency graphs in this application.

[0050] Figure 4 A simplified flowchart illustrating the cross-cloud collaborative scheduling method for a multi-dimensional resource dependency graph provided in this application embodiment;

[0051] Figure 5This is a schematic diagram of the module structure of the cross-cloud collaborative scheduling device for a multi-dimensional resource dependency graph according to an embodiment of this application;

[0052] Figure 6 This is a schematic diagram of the device structure of the hardware operating environment involved in the cross-cloud collaborative scheduling method for multi-dimensional resource dependency graphs in the embodiments of this application.

[0053] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0054] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0055] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0056] The main solution of this application embodiment is as follows: Obtain a cross-cloud resource dependency graph and service scheduling requirement parameter values; use a path search algorithm to find resource combinations in the cross-cloud resource dependency graph that satisfy the service scheduling requirement parameter values, obtaining candidate resource information, which includes regional preference resource information, performance indicator resource information, and cost constraint resource information; generate corresponding resource scheduling configurations based on the regional preference resource information, the performance indicator resource information, and the cost constraint resource information, and determine scheduling strategy information; instantiate based on the scheduling strategy information, create entity resources or adjust entity resources, generate cross-cloud collaborative resource scheduling information, and complete cross-cloud collaborative scheduling based on the cross-cloud collaborative resource scheduling information.

[0057] In this embodiment, for ease of description, the following description will focus on a cross-cloud collaborative scheduling device that identifies multi-dimensional resource dependency graphs.

[0058] Due to the high cost of heterogeneous resource interoperability in existing technologies, cross-cloud permission configuration and network connectivity require manual adaptation, which can easily lead to configuration errors, long debugging cycles, and implicit dependencies that result in blind scheduling. Scheduling is based solely on resource performance indicators without considering dependencies such as permissions, networks, and data synchronization, which can lead to cross-cloud call failures. Furthermore, there is a lack of real-time awareness of cross-cloud resource status changes, resulting in long fault recovery times.

[0059] This application provides a solution. The overall architecture of this application adopts a three-layer architecture of graph construction, intelligent scheduling, and dynamic feedback. It may also include a multi-dimensional resource dependency graph construction module, an intelligent scheduling engine, and a dynamic feedback module. The multi-dimensional resource dependency graph construction module is used to collect cross-cloud resource metadata, define static / dynamic dependency relationships, and store them in a graph database. The intelligent scheduling engine is used to generate cross-cloud scheduling strategies based on the graph and includes three sub-modules: demand parsing, graph matching, and strategy optimization. The dynamic feedback module is used to monitor resource status in real time, update the graph, and trigger strategy adjustments.

[0060] As can be seen from the above embodiments, this application constructs a multi-dimensional resource dependency graph to make the dependencies between cross-cloud resources explicit. By filtering the dependency graph through business requests to determine candidate resource information, entity resources can be created or adjusted, thereby achieving efficient collaborative scheduling of cross-cloud resources, improving the efficiency of cross-cloud resource configuration, significantly shortening configuration time, reducing costs in a multi-cloud environment through intelligent optimization strategies, and significantly enhancing the reliability and dynamic adaptability of the system.

[0061] Based on this, embodiments of this application provide a cross-cloud collaborative adjustment method for a multi-dimensional resource dependency graph, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the cross-cloud collaborative adjustment method for multi-dimensional resource dependency graphs in this application.

[0062] In this embodiment, the cross-cloud collaborative adjustment method for the multi-dimensional resource dependency graph includes steps S10 to S40:

[0063] Step S10: Obtain the cross-cloud resource dependency graph and business scheduling requirement parameter values;

[0064] It should be noted that the cross-cloud resource dependency graph is a data model that models and expresses resources and their interdependencies in a cross-cloud environment in the form of a graph structure. The service scheduling requirement parameter values ​​can be regional preference values, performance index values, and cost constraint values. Among them, regional preference refers to the geographical location of the resource, such as the geographical information of different regions. Performance index refers to the specific requirements for resource performance, such as CPU utilization, memory capacity, and network bandwidth. Cost constraint can be a budget limit, such as requiring that resource costs be reduced as much as possible while meeting performance requirements.

[0065] It is understood that the cross-cloud resource dependency graph can abstract resources on different cloud platforms into nodes for visualization and structured presentation in the form of a graph, such as compute instances, storage resources, and network resources. It also visualizes and structures the dependencies between resources as edge dependencies, such as network connectivity dependencies, permission dependencies, and data synchronization dependencies. By constructing a cross-cloud resource dependency graph, resource combinations that meet business requests can be quickly and accurately identified, avoiding the tedious process of manually sorting out dependencies and significantly improving the efficiency of cross-cloud resource scheduling. Furthermore, the explicit modeling of the dependency graph makes resource dependencies clearly visible, and by filtering them in conjunction with business request information, scheduling errors caused by implicit dependencies can be effectively avoided, improving the accuracy of cross-cloud resource scheduling. Figure 2 As shown, Figure 2 This is a schematic diagram of the cross-cloud resource dependency graph of the multi-dimensional resource dependency graph of this application. Resource entities and dependency edges are stored in a graph database to form a visual dependency graph. The cross-cloud resource dependency graph can clearly show the dedicated network dependency relationship between EC2 instances and ECS instances, as well as the permission dependency relationship of EC2 instances obtaining OSS write permissions through RAM roles.

[0066] Additionally, it should be noted that business request information can be used to extract business scheduling requirement parameter values. The business request information refers to the user's or system's request information for cross-cloud resource scheduling, which makes resource scheduling more accurate, reduces manual intervention and trial-and-error costs, and significantly improves scheduling efficiency.

[0067] In one feasible implementation, step S10 may include steps A11 to A14:

[0068] Step A11: Obtain resource metadata information;

[0069] It should be noted that the resource metadata information is data collected from various resources in a cross-cloud environment.

[0070] It is understood that the resource metadata information is obtained by calling the API interfaces provided by the cloud provider. For example, the `describe-instances` API interface can be used to query EC2 instance information, and the `list-buckets` interface can be used to obtain S3 bucket information. The S3 bucket is a storage container in the object storage service provided by AWS, used to store objects. Each object can be a file or a piece of data. Similarly, the `describeInstances` API interface can be used to query ECS instances, and the `listBuckets` interface can be used to obtain OSS bucket information. The OSS bucket is a container in the object storage service OSS provided by the cloud provider used to store data. Users can create a storage space before uploading files to OSS for storing and managing files. That is, by calling the API interfaces provided by the cloud provider, basic attributes of the resource can be obtained, such as resource ID, type, performance specifications, and network configuration. Dynamic information such as the resource's running status and performance indicators can also be obtained. In practical application scenarios, it is necessary to ensure that access permissions and authentication information are correctly configured to securely access the resource metadata information.

[0071] In a specific embodiment, resource information, such as resource metadata information like EC2 instances and ECS instances, can be obtained from cloud vendor APIs. For EC2 instances, the region information, performance specifications, and IAM role information can be obtained by calling the AWS describe-instances API interface. The region information can be us-west-2, the performance specifications can be 4 vCPUs and 16GB RAM, and the IAM role information can be arn:aws:iam::123456789012:role / EC2Role. For ECS instances, the image ID, RAM role, and network configuration can be obtained by calling the DescribeInstances API interface. The image ID can be m-bp, the RAM role can be acs:ram::123456789012:role / ECSRole, and the network configuration can be IP address, subnet ID, or security group rules.

[0072] Step A12: Map the target instance resource in the resource metadata information to the corresponding entity node resource to determine the mapped entity node information;

[0073] It should be noted that the mapped entity node information is the resource metadata information obtained from the cloud vendor's API mapped into standardized graph node format information.

[0074] It is understood that the mapped entity node information refers to the standardization of resource metadata information obtained from cloud vendor APIs. This standardization process can include resource type standardization, attribute information standardization, dependency standardization, and geographic location standardization. The resource type standardization maps resource types from different cloud vendors to a unified resource type identifier. For example, compute nodes; EC2 instances and ECS instances are both mapped as compute nodes and stored in the graph as the same type of resource. The attribute information standardization converts resource attribute information into a unified format. For example, AWS's 4 vCPU, 16GB RAM and other cloud vendors' 4 vCPU, 16GB RAM are uniformly mapped to CPU: 4, Memory: The 16GB dependency standardization converts resource dependencies into a unified dependency edge format. For example, it maps AWS's IAM role and cloud vendors' RAM role to "permission dependency edges" and records their attributes, such as credential validity periods. The geolocation standardization maps resource geolocation information to standardized geolocation identifiers. This geolocation information is obtained with the user's permission or consent. In other words, when this application is applied to specific products or technologies, user permission is required to acquire and process relevant data, and the processing of this data must comply with relevant national and regional laws, regulations, and regulatory standards. For example, when obtaining a user's current geolocation, a location acquisition prompt can be displayed on the user's terminal. After receiving confirmation from the user regarding the location acquisition prompt, the terminal can acquire the user's geolocation information and obfuscate the collected resource geolocation information. For example, AWS's us-west-2a and cloud vendors' cn-hangzhou-a are uniformly mapped to Region: us-west-2a, Zone: cn-hangzhou-a refers to the use of standardized processing to uniformly map computing resources from different cloud vendors into common entities of computing nodes. This enables resources from different cloud platforms to be managed in a consistent format across the cloud resource dependency graph. For example, although AWS EC2 instances and cloud vendor ECS instances have different descriptions and attribute fields in their native APIs, through mapping, they are both abstracted into "computing nodes" and have unified attribute descriptions, such as the number of CPU cores, memory size, and network configuration. This significantly simplifies the complexity of resource management and enables the system to handle resource collaboration in a multi-cloud environment more efficiently.

[0075] Additionally, it should be noted that the target instance resource is a specific resource instance obtained from the APIs of cloud vendors such as AWS and cloud providers, for example, an AWS EC2 instance or a cloud vendor's ECS resource instance. The resource instance is an actual computing resource existing in the cross-cloud environment, used to support the operation of various business applications. The entity node resource is a standardized graph node used to represent the target instance resource in the cross-cloud resource dependency graph. When constructing the dependency graph, the metadata information of the target instance resource is extracted and converted into a unified format so that it can be represented and managed in a consistent manner in the graph. For example, although AWS EC2 instances and cloud vendor ECS instances come from different cloud platforms, they are both mapped as the common entity node resource of compute nodes in the dependency graph, enabling resources from different cloud platforms to be effectively managed in the same graph.

[0076] Step A13: Construct the corresponding target dependency relationship based on the mapped entity node information, and determine the dependency edge information, which includes network dependency edge information, permission dependency edge information and traffic dependency edge information;

[0077] It should be noted that the dependency edge information is the directed edge information used in the cross-cloud resource dependency graph to represent the dependency relationship between resource entity nodes. Through the form of directed edges, the dependency type of one resource on another is clearly indicated. It can characterize the network connectivity, access permissions and traffic interaction relationship between resources, enabling the system to manage and optimize resource collaboration in a multi-cloud environment more efficiently.

[0078] It is understood that the network dependency edge information is the dependency relationship where a resource needs to communicate with another resource through a specific network path, the permission dependency edge information is the dependency relationship where a resource needs specific permissions to access another resource, and the traffic dependency edge information is the dependency relationship where changes in the load of one resource will affect the load of another resource.

[0079] Additionally, it should be noted that the target dependency can be a network dependency, a permission dependency, or a traffic dependency. The network dependency can be network latency, network bandwidth, network path, and security rules. Network latency represents the maximum allowed latency of a network connection; network bandwidth represents the minimum bandwidth requirement for a network connection; network path represents the network connection path between resources, such as a dedicated line or virtual private cloud connection; and security rules represent allowed or restricted network traffic rules, such as specific port ranges or protocol types. In this case, if a network dependency is established between an EC2 instance and an ECS instance, with the source node being EC2 instance i-0123456789abcdef0 and the target node being ECS ​​instance i-bp, and the network dependency edge attributes being latency not exceeding 50ms, bandwidth not less than 100Mbps, and the network path being a dedicated line connection... The security group rule allows TCP port 22. The permission dependency can be permission type, credential validity period, permission role, and permission policy. Permission type indicates the type of permission for resource access, such as read, write, or management permissions. Credential validity period indicates the validity period of the permission credential, such as 6 hours. Permission role indicates the permission role on which resource access depends, such as an AWS IAM role or a cloud provider's RAM role. Permission policy indicates the specific permission policy, such as the scope of allowed or restricted operations. In this case, if a permission dependency is established for an EC2 instance to access a cloud provider's OSS storage bucket, with the source node being the EC2 instance i-0123456789abcdef0 and the target node being the cloud provider's OSS storage bucket oss-bucket-name, and the permission dependency's side attribute having a write permission type, a 6-hour credential validity period, and a permission role of acs:ram::123456789012, then... :role / ECSRole, the permission policy is to allow write operations. The traffic dependency can be traffic weight, traffic direction, data synchronization frequency, and data volume. Traffic weight can represent the proportional relationship of traffic interaction. For example, 0.8 means that when the load of the source node increases by 10%, the load of the target node increases by 8%. Traffic direction represents the direction of traffic flow, such as from one resource to another. Data synchronization frequency represents the frequency of data synchronization, such as once per minute. Data volume represents the amount of data synchronized or interacted each time, such as 10MB. At this time, if a traffic dependency is established between an EC2 instance and an ECS instance, the source node is EC2 instance i-0123456789abcdef0, the target node is ECS instance i-bp, the traffic weight in the traffic dependency edge attribute is 0.8, the traffic direction is from EC2 instance to ECS instance, the data synchronization frequency is once per minute, and the data volume is 10MB.

[0080] In a specific implementation, AWS EC2 instances and cloud provider ECS instances can be connected via dedicated lines to establish network dependencies, ensuring efficient and stable network communication between the two. The network dependency edge attribute of the dedicated line connection includes latency, requiring a latency of no more than 50 milliseconds to guarantee low latency in data transmission. EC2 instances can obtain write permissions to the cloud provider's OSS storage bucket through the cloud provider's RAM role, establishing permission dependencies. The attributes of the permission dependency edge include the validity period of the STS certificate, such as setting it to 6 hours, for more secure and flexible resource access, allowing data write operations within a limited time. Based on historical data analysis, when the load of the EC2 instance increases by 10%, the load of the ECS instance will also increase by 8% synchronously, establishing traffic dependencies with a weight of 0.8, representing the load correlation between the two instances. For example, as user traffic increases, the load of the EC2 instance rises, and the ECS instance also needs to increase resources accordingly to cope with the increased business demand, ensuring the stable operation of the entire system.

[0081] Step A14: Store the mapping entity node information and the dependency edge information into a preset graph database to generate a cross-cloud resource dependency graph.

[0082] It is understood that the graph database is a database system used to store and query graph-structured data. It stores data in the form of nodes and edges, naturally representing dependencies between resources. Each node represents a resource entity, and each edge represents a dependency between resources. For example, EC2 instances and ECS instances can be stored as nodes, and the network dependency edges between them can be stored as edges connecting these two nodes. It supports complex query operations, enabling fast traversal and querying of paths, subgraphs, and other structures within the graph. It can efficiently handle complex dependencies and multi-dimensional relational data, supporting rapid querying and analysis, for example, through the graph database's query language. It can quickly find all ECS instances connected to a specific EC2 instance via dedicated lines, and can dynamically add, delete, or modify nodes and edges. It supports real-time graph updates to reflect changes in resource status in a timely manner, and supports dynamic scheduling and optimization. For example, when the load on an EC2 instance increases, it can dynamically update the attributes of its traffic-dependent edges, such as traffic weight or data synchronization frequency. It has high-performance storage and query engines, can handle large-scale graph data, supports horizontal scaling, and can adapt to the ever-increasing number of resources and dependencies. For example, it can use graph databases such as Neo4j and OrientDB to process large-scale graph data and support complex business scenarios.

[0083] In one feasible implementation, step A14 may include steps B11 to B14:

[0084] Step B11: Obtain monitoring feedback information;

[0085] It should be noted that the monitoring feedback information is real-time data on resource operating status and performance indicators collected from the cloud platform's monitoring system.

[0086] It is understandable that the monitoring feedback information can be used to detect changes in resource status in a timely manner. For example, when the CPU utilization of a certain resource exceeds a threshold, the system can trigger an automatic expansion operation.

[0087] In a specific embodiment, the system integrates monitoring tools such as Prometheus and CloudWatch to collect resource performance metrics and operational status data in real time. The Prometheus tool is used to monitor resources in the Kubernetes environment, while the CloudWatch tool is used to monitor resources in the AWS cloud platform, providing real-time resource load information, network performance metrics, and resource health status data, thereby obtaining monitoring feedback information.

[0088] Step B12: Based on the monitoring feedback information, detect abnormal resource load and abnormal dependency edges, and determine the detection results;

[0089] It should be noted that the detection results are specific detection results regarding abnormal resource load and abnormal dependency edges, obtained after analyzing the monitoring feedback information.

[0090] It is understood that the abnormal resource load can be due to abnormal CPU utilization, i.e., the CPU utilization of a certain resource is detected to exceed a preset threshold. For example, real-time data obtained by the system through monitoring tools shows that the CPU utilization of a certain ECS instance has been continuously exceeding 90% in the past 5 minutes, indicating that the resource load is too high. It can also be due to abnormal memory utilization, i.e., the memory utilization of a certain resource is detected to exceed a preset threshold. For example, the memory utilization of a certain EC2 instance has been continuously exceeding 80% in the past 10 minutes and there is no trend of release. It can also be due to abnormal network bandwidth, i.e., the network bandwidth utilization of a certain resource is detected to exceed a preset threshold. For example, the network bandwidth of a certain ECS instance has been continuously exceeding 95% in the past 5 minutes, resulting in increased network transmission latency. It can also be due to abnormal disk I / O, i.e., the disk I / O operation rate of a certain resource is detected to exceed a preset threshold. For example, the disk I / O operation rate of a certain EC2 instance has been continuously exceeding 1000 times / second in the past 10 minutes, resulting in increased disk response time.

[0091] Additionally, it should be noted that the aforementioned dependency edge anomalies can be network dependency edge anomalies, permission dependency edge anomalies, and traffic dependency edge anomalies. Specifically, network dependency edge anomalies can be network latency anomalies, meaning the latency of a network dependency edge is detected to exceed a preset threshold. For example, the network latency between an EC2 instance and an ECS instance has consistently exceeded 50ms for the past 5 minutes, resulting in reduced data transmission efficiency. They can also be insufficient network bandwidth, meaning the bandwidth of a network dependency edge is detected to be below a preset threshold. For example, the network bandwidth between an EC2 instance and an ECS instance has consistently been below 100Mbps for the past 10 minutes, resulting in slow data transmission speeds. Furthermore, they can be network path failures, meaning a failure is detected in the path of a network dependency edge. For example, the dedicated line connection between an EC2 instance and an ECS instance is interrupted, causing a complete network communication interruption. Permission dependency edge anomalies can be permission invalidation, meaning the credentials for a permission dependency edge are detected to be expired or invalid. For example, an EC2 instance obtains write permissions for a cloud provider's OSS storage bucket through a RAM role, but the permission credentials are detected to be expired, preventing normal data writing. Alternatively, they can be insufficient permissions, meaning insufficient permissions are detected for a permission dependency edge. For example, an EC2 instance needs write permissions to write data to the cloud provider's OSS storage bucket, but if the current permissions are only read permissions, the write operation will fail. The traffic dependency edge anomaly can be an abnormal traffic weight, that is, a change in the traffic weight of a certain traffic dependency edge is detected. For example, the traffic weight between the EC2 instance and the ECS instance changes from 0.8 to 0.5, indicating a change in the load correlation between the two instances. It can also be an abnormal data synchronization frequency, that is, a change in the data synchronization frequency of a certain traffic dependency edge is detected. For example, the data synchronization frequency between the EC2 instance and the ECS instance changes from 1 minute / time to 30 seconds / time, indicating an increase in the data synchronization frequency. It can also be an abnormal data volume, that is, a change in the data volume of a certain traffic dependency edge is detected. For example, the data volume between the EC2 instance and the ECS instance changes from 10MB to 20MB, indicating an increase in the amount of data for each synchronization or interaction.

[0092] In a specific embodiment, after obtaining monitoring feedback information, abnormalities in dependencies can be detected. For example, when the monitoring system detects that the load of the cloud provider's ECS instance exceeds 90%, the system will automatically trigger a series of operations. When the set threshold is exceeded or the set conditions are not met, the system can adjust the network configuration or switch to a backup path. That is, if the ECS load is detected to exceed 90%, the graph update is automatically triggered, a backup EC2 instance is scheduled, and the data synchronization frequency is adjusted.

[0093] Step B13: Update the attribute values ​​of the mapped entity node information and the attribute values ​​of the dependent edge information based on the detection results, and determine the update information;

[0094] It should be noted that the update information is the specific update information after adjusting the mapping entity node information and dependent edge information based on the detection results.

[0095] It is understood that updating the attribute values ​​of the mapped entity node information is equivalent to adjusting the attribute values ​​of the resources. If the CPU utilization of a resource exceeds a preset threshold, the CPU utilization attribute value of the resource node will be updated. If the memory utilization of a resource exceeds a preset threshold, the memory utilization attribute value of the resource node will be updated. If the network bandwidth utilization of a resource exceeds a preset threshold, the network bandwidth attribute value of the resource node will be updated. If the disk I / O operation rate of a resource exceeds a preset threshold, the disk I / O attribute value of the resource node will be updated.

[0096] Additionally, it should be noted that updating the attribute values ​​of the dependency edge information is equivalent to adjusting the dependency relationships between resources. If the latency, bandwidth, and path of a certain network dependency edge are detected to exceed a preset threshold, the system will update the latency, bandwidth, and path attribute values ​​of that dependency edge accordingly. If the credentials of a certain permission dependency edge are detected to be expired, invalid, or have insufficient permissions, the system will regenerate valid permission credentials, adjust the permission policy, and update the credential attribute values ​​of that dependency edge. Traffic weight update: If the traffic weight or data volume of a certain traffic dependency edge is detected to have changed, the system will update the traffic weight attribute value or data volume of that dependency edge accordingly.

[0097] In a specific embodiment, the system obtains the latest resource load data from the monitoring tool and updates the corresponding node attributes in the graph database. For example, it updates the CPU utilization of the ECS instance to 90%. Furthermore, the system checks the dependent edges related to the ECS instance and updates the status and attributes of the dependent edges. For example, if the high load of the ECS instance leads to increased network latency, the system updates the latency attribute of the network dependent edge to obtain the detection result. Based on the detection result, the system updates the attribute values ​​of the mapped entity node information and the attribute values ​​of the dependent edge information to determine the update information. That is, the system evaluates the current resource load and dependencies based on the updated graph information, automatically triggers the scheduling strategy, and updates the attribute values ​​of the mapped entity node information and the attribute values ​​of the dependent edge information to obtain the update information.

[0098] Step B14: Store the updated information in a preset graph database to obtain a cross-cloud resource dependency graph.

[0099] Understandably, resource scheduling strategies can be optimized. Based on real-time load and performance data, the system can dynamically adjust resource allocation to ensure efficient system operation.

[0100] In a specific implementation, the system will search for a spare AWS EC2 instance from the graph and, according to a preset scheduling policy, transfer some of the load from the ECS instance to the spare EC2 instance. For example, the system can start a new EC2 instance and add it to the load balancing group to share the pressure on the ECS instance. The system will adjust the data synchronization frequency between the ECS instance and related resources. For example, the data synchronization frequency between the ECS instance and the OSS bucket will be increased from once per minute to once every 30 seconds. The system will adjust the data transmission schedule task according to the new synchronization frequency.

[0101] Step S20: Based on the path search algorithm, find the resource combinations in the cross-cloud resource dependency graph that meet the service scheduling requirement parameter values ​​to obtain candidate resource information. The candidate resource information includes regional preference resource information, performance index resource information, and cost constraint resource information.

[0102] It should be noted that the candidate resource information is a set of resources that have been screened and verified by the system, can meet business needs, and have no dependency conflicts. For example, in the Double 11 e-commerce scenario, the candidate resource information represents the interaction information between ECS instances and EC2 instances, and the resources are connected through a dedicated line, with no network or permission dependency conflicts.

[0103] It is understood that the regional preference resource information refers to the geographical location of the resource, such as the geographical information of different regions; the performance index resource information refers to the specific requirements for resource performance, such as CPU utilization, memory capacity, and network bandwidth; and the cost constraint resource information can be a budget constraint, such as requiring that resource costs be reduced as much as possible while meeting performance requirements.

[0104] In a specific implementation, the system receives business requests from users, such as the need to expand computing resources by 500 during the Double 11 shopping festival, while requiring user access latency to be less than 100 milliseconds. In this case, the system extracts business scheduling requirement parameters, such as regional preferences, performance indicators, and cost constraints. It then uses Dijkstra's algorithm to perform path search in a multi-dimensional resource dependency graph to find resource combinations that meet the business request conditions. For example, for the aforementioned Double 11 expansion requirement, the system will search for a resource combination scheme, selecting ECS ​​instances from cloud providers as the primary computing resources and AWS EC2 instances as backup resources in a backup region to achieve redundancy and load balancing. This ensures that if the primary resource fails, it can quickly switch to the backup resource, guaranteeing business continuity. After finding a resource combination that initially meets the conditions, the system will perform dependency conflict detection on the resource combination, excluding EC2-ECS combinations without dedicated line connections, and checking whether the permissions between resource nodes match. For example, if an EC2 instance does not have sufficient permissions to access the storage resources that an ECS instance depends on, it is excluded, thus obtaining candidate resource information.

[0105] In one feasible implementation, step S20 may include steps C11-C13:

[0106] Step C11: Based on the path search algorithm, find the resource combinations in the cross-cloud resource dependency graph that meet the service scheduling requirement parameter values, and determine the resource combinations to be detected;

[0107] It should be noted that the combination of resources to be detected is a set of resources that meets the business scheduling constraints and is initially screened by the path search algorithm.

[0108] It is understood that the resource combination to be detected also exists in the graph in the form of nodes and edges, which meets the requirements of regional preferences, performance indicators and cost constraints in business requests. In the complex cross-cloud resource dependency graph, resource combinations that initially meet the scheduling constraints can be quickly found, which significantly improves the efficiency of resource screening, avoids unnecessary calculation and detection of a large number of resource combinations that do not meet the conditions, optimizes the overall scheduling process and reduces computing costs.

[0109] Step C12: Detect dependency conflicts for each resource combination in the resource combination to be detected, and determine the dependency conflict detection result;

[0110] It should be noted that the dependency conflict detection results are obtained after checking the dependencies of the initially screened combination of resources to be detected.

[0111] It is understood that the dependency conflict detection result is used to detect dependency conflicts in resource combinations, which helps to select feasible resource combinations, avoid scheduling failures caused by dependency issues, optimize resource utilization, and reduce resource waste.

[0112] Step C13: Based on the dependency conflict detection results, select target resources that do not have dependency conflicts in the resource combination to be detected, and obtain candidate resource information.

[0113] Understandably, by analyzing the dependency conflict detection results, resource combinations with dependency conflicts can be eliminated, thereby ensuring that the selected resource combinations are feasible in terms of dependencies, improving the reliability of cross-cloud resource scheduling, adapting to the needs of different business scenarios, and enhancing the versatility and adaptability of the system.

[0114] Step S30: Generate a corresponding resource scheduling configuration based on the regional preference resource information, the performance index resource information, and the cost constraint resource information, and determine the scheduling strategy information;

[0115] It should be noted that the scheduling strategy information is generated based on regional preference resource information, performance indicator resource information, and cost constraint resource information, and is used to characterize the specific configuration of cross-cloud resource scheduling.

[0116] Understandably, based on the geographical preferences in business scheduling needs, resources with suitable geographical locations can be prioritized. For example, if the business requires resources to be close to the user's location to reduce latency, the scheduling strategy will prioritize cloud resources close to the user's location. Based on performance requirements, resources that can meet performance needs will be selected. For example, if the business requires high computing power, the scheduling strategy will prioritize high-performance computing instances. Under the premise of meeting performance and geographical requirements, the resource combination with the lowest cost will be selected. For example, by comparing the resource prices of different cloud vendors, the resource with the highest cost performance will be selected.

[0117] Step S40: Instantiate based on the scheduling strategy information, create entity resources or adjust entity resources, generate cross-cloud collaborative resource scheduling information, and complete cross-cloud collaborative scheduling based on the cross-cloud collaborative resource scheduling information.

[0118] It should be noted that the cross-cloud collaborative resource scheduling information is a set of detailed information generated by the system after completing resource scheduling. It is used to record and describe the specific content, status and results of resource scheduling. Instantiation is the process of transforming the abstract resource scheduling plan into a specific and operable entity resource based on the scheduling strategy information. That is, using the scheduling strategy to create or adjust actual cloud resources to meet business needs.

[0119] It is understood that the cross-cloud collaborative resource scheduling information may include resource allocation details, dependency configuration, scheduling status, and performance metrics. The resource allocation details are records of resources being allocated or adjusted on the cloud platform. The dependency configuration is a record of the configuration and adjustment of dependencies between resources. The scheduling status is a record of the execution of scheduling operations. The performance metrics are records of the performance metrics of the resources after scheduling.

[0120] Additionally, it should be noted that the aforementioned physical resources are actual computing, storage, and network resources, which are the direct objects of cross-cloud resource scheduling. These physical resources can come from different cloud platforms and may include computing resources, storage resources, and network resources. Computing resources may be AWS EC2 instances and Alibaba Cloud ECS instances, providing computing power for running applications and services. Storage resources may be AWS S3 buckets and Alibaba Cloud OSS buckets, used for storing data. Network resources may be virtual private clouds and dedicated line connections, used to ensure network communication between different cloud platforms. These resources are uniformly managed and scheduled through the cross-cloud resource scheduling system to meet the needs of different business scenarios. The cloud platform interface is an application programming interface provided by the cloud service provider, used to manage and operate cloud resources. Different cloud platforms have different API interfaces. By calling the cloud platform interface, the cross-cloud resource scheduling system can achieve unified management and scheduling of resources from different cloud platforms, ensuring efficient collaborative work across cloud resources.

[0121] In one feasible implementation, step S40 may include steps D11-D14:

[0122] Step D11: Obtain interoperability resource scheduling information;

[0123] It should be noted that the interoperability resource scheduling information is used to coordinate and synchronize the resource scheduling of different cloud platforms when performing resource collaborative operations between multiple cloud platforms.

[0124] It is understood that the interoperability resource scheduling information is used to coordinate resource scheduling between different cloud platforms, record in detail the availability and health status of resources as well as the dependencies between resources, so as to ensure the collaborative work between resources, monitor resource status in real time, discover and resolve potential problems in a timely manner, realize efficient collaborative scheduling of cross-cloud resources, and significantly improve resource utilization, response efficiency and cost control capabilities.

[0125] Step D12: Instantiate and create entity resources based on the scheduling strategy information, and determine the first cross-cloud collaborative resource scheduling information;

[0126] It should be noted that the first cross-cloud collaborative resource scheduling information is the initial scheduling information generated by the cross-cloud resource scheduling system when it calls the cloud platform interface to create or adjust resources.

[0127] It is understandable that the first cross-cloud collaborative resource scheduling information records in detail the specific content of the resource creation request, the detailed information of the interface call, the status of resource allocation, the initial dependency configuration, and the initial performance indicators and monitoring data of the resource. By using the first cross-cloud collaborative resource scheduling information, problems that occur during the resource creation process can be discovered and dealt with in a timely manner, achieving efficient resource scheduling, significantly shortening the business launch time, and improving user experience and satisfaction.

[0128] Step D13: Adjust the entity resources in the first cross-cloud collaborative resource scheduling information based on the interoperability resource scheduling information, and determine the second cross-cloud collaborative resource scheduling information;

[0129] It should be noted that the second cross-cloud collaborative resource scheduling information is detailed information generated by the cross-cloud resource scheduling system after adjusting and optimizing the created resources based on the interoperability resource scheduling information.

[0130] It is understandable that the second cross-cloud collaborative resource scheduling information records the specific content of resource adjustment, the adjusted resource status, the update of dependencies, and the changes in performance indicators. Using the second cross-cloud collaborative resource scheduling information, resources can be dynamically adjusted, significantly improving the system's adaptability and flexibility. Furthermore, during the resource adjustment process, potential problems can be detected and addressed in a timely manner, ensuring the high availability and reliability of resources. For example, by automatically switching to backup resources or reconfiguring network connections, the system can quickly recover in the event of a failure.

[0131] Step D14: Obtain cross-cloud collaborative resource scheduling information based on the first cross-cloud collaborative resource scheduling information and the second cross-cloud collaborative resource scheduling information.

[0132] Understandably, by using a unified API gateway to call cloud vendor interfaces to create resources and obtaining the first cross-cloud collaborative resource scheduling information, and injecting interoperability rules, such as generating temporary OSS writing credentials for EC2, the first cross-cloud collaborative resource scheduling information is adjusted to obtain the second cross-cloud collaborative resource scheduling information. This simplifies the cross-cloud resource creation process and significantly reduces resource creation time; for example, resource creation time is shortened to 5 minutes per batch.

[0133] This embodiment proposes a cross-cloud collaborative scheduling method based on a multi-dimensional resource dependency graph. The method involves obtaining a cross-cloud resource dependency graph and business scheduling requirement parameter values; using a path search algorithm to find resource combinations in the cross-cloud resource dependency graph that satisfy the business scheduling requirement parameter values, obtaining candidate resource information, which includes regional preference resource information, performance indicator resource information, and cost constraint resource information; generating corresponding resource scheduling configurations based on the regional preference resource information, performance indicator resource information, and cost constraint resource information, and determining scheduling strategy information; instantiating entity resources based on the scheduling strategy information, creating or adjusting entity resources, generating cross-cloud collaborative resource scheduling information, and completing cross-cloud collaborative scheduling based on the cross-cloud collaborative resource scheduling information. This invention addresses the technical challenge of achieving efficient collaborative scheduling of cross-cloud resources, thereby improving resource utilization, response efficiency, and cost control. Compared to existing technologies, this application constructs a multi-dimensional resource dependency graph, filters the dependency graph based on business requests to determine candidate resource information, and creates or adjusts entity resources. This enables efficient collaborative scheduling of cross-cloud resources, improves the efficiency of cross-cloud resource configuration, significantly shortens configuration time, reduces costs through dynamic monitoring and optimization strategies, enhances system reliability and adaptability, and greatly reduces manual intervention. It can more flexibly address complex business needs in multi-cloud environments.

[0134] Based on the first embodiment of this application, in the second embodiment of this application, the same or similar content as the first embodiment can be referred to the above description, and will not be repeated hereafter.

[0135] In this embodiment, refer to Figure 3 , Figure 3 This is a flowchart illustrating Embodiment 2 of the cross-cloud collaborative scheduling method for multi-dimensional resource dependency graphs in this application. After step S20, it further includes steps S21 to S22:

[0136] Step S21: Calculate the corresponding objective function value based on the objective optimization function and the candidate resource information, and determine the overall function value;

[0137] It should be noted that the overall function value is a quantitative indicator obtained by comprehensively evaluating each resource combination in the candidate resource information, and is used to characterize the overall performance and efficiency of the resource combination in meeting the parameter values ​​of business scheduling requirements.

[0138] It is understood that the overall function value can be calculated based on multiple objective function values, taking into account multiple factors such as resource utilization, response time and cost, and measuring the performance of the resource combination in different aspects, such as resource utilization, response time and cost, and is calculated by weighted summation, with the weights allocated according to the importance of business needs.

[0139] In one feasible implementation, step S21 may include steps E11-E12:

[0140] Step E11: Calculate the objective function value corresponding to each resource combination in the candidate resource information based on the objective optimization function, and determine the resource utilization function value, response time function value, and cost function value;

[0141] It should be noted that the resource utilization function value is a quantitative value that measures the effective utilization of resources in the actual use of the resource combination, the response time function value is a quantitative value that measures the response speed of the resource combination when processing requests, and the cost function value is a quantitative value that measures the cost-effectiveness of the resource combination in the use process.

[0142] Understandably, the higher the objective function value for each resource combination, the more efficient the resource utilization, the faster the response speed, and the lower the cost while meeting the corresponding business needs. The resource utilization function value is the ratio of the actual resource usage to the total allocated amount. For example, for a resource combination, if 100 CPU cores are allocated and 80 are actually used, the resource utilization function value is 0.8. The response time function value is 1 divided by the average response time. For example, if the average response time for a resource combination is 100ms, the response time function value is 0.01. The cost function value is 1 divided by the total cost. For example, if the total cost for a resource combination is 1000 yuan, the cost function value is 0.001.

[0143] Step E12: Obtain the overall function value based on the resource utilization function value, the response time function value, and the cost function value.

[0144] It is understood that the resource utilization function value is a quantitative value that measures the resource utilization rate of the resource combination when meeting business needs. High resource utilization rate indicates that resources are used effectively and waste is reduced. The response time function value is a quantitative value that measures the response time of the resource combination when meeting business needs. Low response time indicates that the system can respond to user requests quickly and improve user experience. The cost function value is a quantitative value that measures the cost of the resource combination when meeting business needs. Low cost indicates that the cost of resource usage is effectively controlled while meeting performance and geographical requirements.

[0145] Step S22: Based on the overall function value, sort the resource combinations that meet the service scheduling requirement parameter values, and take the resource combination with the highest ranking as candidate resource information.

[0146] Understandably, by calculating the resource utilization function value, response time function value, and cost function value for each candidate resource, and combining this with pre-configured calculation weights, the overall function value for each resource combination can be obtained. For example, if the resource utilization function value is configured to 0.4, the response time function value to 0.3, and the cost function value to 0.3, then the objective function value = 0.4 × resource utilization function value + 0.3 × response time function value + 0.3 × cost function value. In this case, for a resource combination with a resource utilization function value of 0.8, a response time function value of 0.01, and a cost function value of 0.001, the objective function value is 0.4 × 0.8 + 0.3 × 0.01 + 0.3 × 0.001 = 0.32 + 0.003 + 0.0003 = 0.3233. Using the calculated overall function value for each resource combination, each candidate resource can be evaluated and ranked. Finally, the resource combination with the highest overall function value is selected as the target ranked resource.

[0147] This embodiment proposes a cross-cloud collaborative scheduling method based on a multi-dimensional resource dependency graph. It calculates the corresponding objective function value based on the objective optimization function and the candidate resource information to determine the overall function value. Based on the overall function value, resource combinations that meet the business scheduling requirement parameters are ranked, and the resource combination ranked first is selected as the candidate resource information. This solves the technical problem of how to achieve efficient collaborative scheduling of cross-cloud resources, improving resource utilization, response efficiency, and cost control. Compared with existing technologies, this application introduces an objective optimization function to calculate the objective function values ​​of resource utilization, response time, and cost for each resource combination, determines the overall function value, and then ranks the resource combinations, selecting the resource combination ranked first as the candidate resource information. This achieves the optimal balance between resource utilization, response time, and cost, optimizes resource utilization, improves scheduling efficiency, enhances system reliability and dynamic adaptability, and ensures efficient cross-cloud resource scheduling.

[0148] For example, to help understand the implementation process of the cross-cloud collaborative scheduling method for the multi-dimensional resource dependency graph obtained in this embodiment combined with the above embodiment one, please refer to... Figure 4 , Figure 4 A simplified flowchart illustrating a cross-cloud collaborative scheduling method based on a multi-dimensional resource dependency graph is provided. Specifically:

[0149] Referring to Example 1, a cross-cloud resource dependency graph and service scheduling requirement parameter values ​​are obtained; a path search algorithm is used to find resource combinations in the cross-cloud resource dependency graph that satisfy the service scheduling requirement parameter values, obtaining candidate resource information, which includes regional preference resource information, performance indicator resource information, and cost constraint resource information; a corresponding resource scheduling configuration is generated based on the regional preference resource information, the performance indicator resource information, and the cost constraint resource information to determine scheduling strategy information; based on the scheduling strategy information, entity resources are instantiated, creating or adjusting entity resources to generate cross-cloud collaborative resource scheduling information; and cross-cloud collaborative scheduling is completed based on the cross-cloud collaborative resource scheduling information. Referring to Example 2, a corresponding objective function value is calculated based on the objective optimization function and the candidate resource information to determine the overall function value; based on the overall function value, resource combinations that satisfy the service scheduling requirement parameter values ​​are sorted, and the resource combination with the highest ranking is used as candidate resource information. This application defines a triplet model that includes resource attributes, static dependencies, and dynamic dependencies to achieve an explicit expression of cross-cloud resource dependencies. It also employs a dependency-aware intelligent scheduling algorithm, combined with graph neural networks and reinforcement learning, to enforce dependency constraints in the scheduling strategy and dynamically optimize multi-objective functions. Furthermore, it drives graph updates through real-time monitoring data, achieving a closed loop of monitoring-scheduling-feedback, supporting minute-level strategy iteration. Its beneficial effects are significant: cross-cloud resource configuration time is reduced from 2 hours to 15 minutes, dependency conflict troubleshooting efficiency is improved by 90%, costs are optimized, significantly reducing costs in multi-cloud environments, fault recovery time is reduced from 30 minutes to 5 minutes, and cross-cloud call success rate is increased to 99.5%, significantly enhancing reliability.

[0150] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the cross-cloud collaborative scheduling method of the multi-dimensional resource dependency graph of this application. Any simple transformations based on this technical concept are within the protection scope of this application.

[0151] This application also provides a cross-cloud collaborative scheduling device based on a multi-dimensional resource dependency graph. Please refer to [link / reference]. Figure 5 The cross-cloud collaborative scheduling device for the multi-dimensional resource dependency graph includes:

[0152] Module 10 is used to obtain cross-cloud resource dependency graphs and business scheduling requirement parameter values;

[0153] Processing module 20 is used to find resource combinations in the cross-cloud resource dependency graph that meet the service scheduling requirement parameter values ​​based on the path search algorithm, and obtain candidate resource information, which includes regional preference resource information, performance index resource information and cost constraint resource information;

[0154] The processing module 20 is further configured to generate a corresponding resource scheduling configuration based on the regional preference resource information, the performance index resource information and the cost constraint resource information, and determine the scheduling strategy information;

[0155] The execution module 30 is used to instantiate based on the scheduling strategy information, create entity resources or adjust entity resources, generate cross-cloud collaborative resource scheduling information, and complete cross-cloud collaborative scheduling based on the cross-cloud collaborative resource scheduling information.

[0156] The acquisition module 10 is also used to acquire resource metadata information;

[0157] Map the target instance resources in the resource metadata information to the corresponding entity node resources to determine the mapped entity node information;

[0158] Based on the mapped entity node information, a corresponding target dependency relationship is constructed, and dependency edge information is determined. The dependency edge information includes network dependency edge information, permission dependency edge information, and traffic dependency edge information.

[0159] The mapping entity node information and the dependency edge information are stored in a preset graph database to generate a cross-cloud resource dependency graph.

[0160] The acquisition module 10 is also used to acquire monitoring feedback information;

[0161] Based on the monitoring feedback information, abnormal resource load and abnormal dependency edges are detected, and the detection results are determined.

[0162] Based on the detection results, update the attribute values ​​of the mapped entity node information and the attribute values ​​of the dependent edge information to determine the update information;

[0163] The updated information is stored in a preset graph database to obtain a cross-cloud resource dependency graph.

[0164] The processing module 20 is also used to find resource combinations in the cross-cloud resource dependency graph that meet the service scheduling requirement parameter values ​​based on the path search algorithm, and determine the resource combination to be detected;

[0165] Dependency conflicts are detected for each resource combination in the resource combination to be detected, and the dependency conflict detection results are determined.

[0166] Based on the dependency conflict detection results, target resources without dependency conflicts in the resource combination to be detected are selected to obtain candidate resource information.

[0167] The processing module 20 is further configured to calculate the corresponding objective function value based on the objective optimization function and the candidate resource information, and determine the overall function value;

[0168] Based on the overall function value, the resource combinations that meet the service scheduling requirement parameter values ​​are sorted, and the resource combination with the highest ranking is taken as the candidate resource information.

[0169] The processing module 20 is further configured to calculate the objective function value corresponding to each resource combination in the candidate resource information based on the objective optimization function, and determine the resource utilization function value, response time function value and cost function value;

[0170] The overall function value is obtained based on the resource utilization function value, the response time function value, and the cost function value.

[0171] The execution module 30 is also used to obtain interoperability resource scheduling information;

[0172] Based on the scheduling strategy information, entity resources are instantiated and created to determine the first cross-cloud collaborative resource scheduling information;

[0173] Based on the interoperability resource scheduling information, adjust the entity resources in the first cross-cloud collaborative resource scheduling information to determine the second cross-cloud collaborative resource scheduling information.

[0174] Cross-cloud collaborative resource scheduling information is obtained based on the first cross-cloud collaborative resource scheduling information and the second cross-cloud collaborative resource scheduling information.

[0175] The cross-cloud collaborative scheduling device for multi-dimensional resource dependency graphs provided in this application, employing the cross-cloud collaborative scheduling method for multi-dimensional resource dependency graphs described in the above embodiments, can solve the technical problem of how to achieve efficient collaborative scheduling of cross-cloud resources and improve resource utilization, response efficiency, and cost control capabilities. Compared with the prior art, the beneficial effects of the cross-cloud collaborative scheduling device for multi-dimensional resource dependency graphs provided in this application are the same as those of the cross-cloud collaborative scheduling method for multi-dimensional resource dependency graphs provided in the above embodiments, and other technical features in the cross-cloud collaborative scheduling device for multi-dimensional resource dependency graphs are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0176] This application provides a cross-cloud collaborative scheduling device for a multi-dimensional resource dependency graph. The cross-cloud collaborative scheduling device for a multi-dimensional resource dependency graph includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the cross-cloud collaborative scheduling method for a multi-dimensional resource dependency graph in the above embodiment 1.

[0177] The following is for reference. Figure 6This document illustrates a structural diagram of a cross-cloud collaborative scheduling device suitable for implementing a multi-dimensional resource dependency graph in the embodiments of this application. The cross-cloud collaborative scheduling device for the multi-dimensional resource dependency graph in the embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), and in-vehicle terminals (e.g., in-vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. Figure 6 The cross-cloud collaborative scheduling device with a multi-dimensional resource dependency graph shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0178] like Figure 6 As shown, the cross-cloud collaborative scheduling device for multi-dimensional resource dependency graphs may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes based on programs stored in ROM (Read Only Memory) 1002 or programs loaded from storage device 1003 into RAM (Random Access Memory) 1004. RAM 1004 also stores various programs and data required for the operation of the cross-cloud collaborative scheduling device for multi-dimensional resource dependency graphs. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via bus 1005. Input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the cross-cloud collaborative scheduling device with a multi-dimensional resource dependency graph to communicate wirelessly or wiredly with other devices to exchange data. Although the figure shows a cross-cloud collaborative scheduling device with a multi-dimensional resource dependency graph of various systems, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems can be implemented or possessed alternatively.

[0179] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0180] The cross-cloud collaborative scheduling device based on a multi-dimensional resource dependency graph provided in this application, employing the cross-cloud collaborative scheduling method based on a multi-dimensional resource dependency graph as described in the above embodiments, can solve the technical problem of how to achieve efficient collaborative scheduling of cross-cloud resources and improve resource utilization, response efficiency, and cost control capabilities. Compared with the prior art, the beneficial effects of the cross-cloud collaborative scheduling device based on a multi-dimensional resource dependency graph provided in this application are the same as those of the cross-cloud collaborative scheduling method based on a multi-dimensional resource dependency graph provided in the above embodiments, and other technical features in this cross-cloud collaborative scheduling device based on a multi-dimensional resource dependency graph are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0181] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0182] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0183] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, which are used to execute the cross-cloud collaborative scheduling method for the multi-dimensional resource dependency graph in the above embodiments.

[0184] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0185] The aforementioned computer-readable storage medium may be included in a cross-cloud collaborative scheduling device for a multi-dimensional resource dependency graph; or it may exist independently and not be assembled into a cross-cloud collaborative scheduling device for a multi-dimensional resource dependency graph.

[0186] The aforementioned computer-readable storage medium carries one or more programs. When these programs are executed by a cross-cloud collaborative scheduling device with a multi-dimensional resource dependency graph, the cross-cloud collaborative scheduling device with a multi-dimensional resource dependency graph performs the following actions: acquires a cross-cloud resource dependency graph and business scheduling requirement parameter values; searches for resource combinations in the cross-cloud resource dependency graph that satisfy the business scheduling requirement parameter values ​​based on a path search algorithm, obtaining candidate resource information, which includes regional preference resource information, performance indicator resource information, and cost constraint resource information; generates corresponding resource scheduling configurations based on the regional preference resource information, the performance indicator resource information, and the cost constraint resource information, determining scheduling strategy information; instantiates based on the scheduling strategy information, creating or adjusting entity resources, generating cross-cloud collaborative resource scheduling information, and completes cross-cloud collaborative scheduling based on the cross-cloud collaborative resource scheduling information.

[0187] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0188] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0189] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0190] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the cross-cloud collaborative scheduling method based on the multi-dimensional resource dependency graph described above. This solves the technical problem of how to achieve efficient collaborative scheduling of cross-cloud resources and improve resource utilization, response efficiency, and cost control capabilities. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the cross-cloud collaborative scheduling method based on the multi-dimensional resource dependency graph provided in the above embodiments, and will not be elaborated upon here.

[0191] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A cross-cloud collaborative scheduling method based on a multi-dimensional resource dependency graph, characterized in that, The method includes: Obtain cross-cloud resource dependency graphs and business scheduling requirement parameter values; Based on the path search algorithm, resource combinations that meet the business scheduling requirement parameter values ​​in the cross-cloud resource dependency graph are found to obtain candidate resource information, which includes regional preference resource information, performance indicator resource information and cost constraint resource information. Based on the regional preference resource information, the performance index resource information, and the cost constraint resource information, a corresponding resource scheduling configuration is generated, and scheduling strategy information is determined. Instantiate based on the scheduling strategy information, create entity resources or adjust entity resources, generate cross-cloud collaborative resource scheduling information, and complete cross-cloud collaborative scheduling based on the cross-cloud collaborative resource scheduling information.

2. The method as described in claim 1, characterized in that, The steps for obtaining the cross-cloud resource dependency graph include: Obtain resource metadata information; Map the target instance resources in the resource metadata information to the corresponding entity node resources to determine the mapped entity node information; Based on the mapped entity node information, a corresponding target dependency relationship is constructed, and dependency edge information is determined. The dependency edge information includes network dependency edge information, permission dependency edge information, and traffic dependency edge information. The mapping entity node information and the dependency edge information are stored in a preset graph database to generate a cross-cloud resource dependency graph.

3. The method as described in claim 2, characterized in that, The step of storing the mapped entity node information and the dependency edge information into a preset graph database to generate a cross-cloud resource dependency graph includes: Obtain monitoring feedback information; Based on the monitoring feedback information, abnormal resource load and abnormal dependency edges are detected, and the detection results are determined. Based on the detection results, update the attribute values ​​of the mapped entity node information and the attribute values ​​of the dependent edge information to determine the update information; The updated information is stored in a preset graph database to obtain a cross-cloud resource dependency graph.

4. The method as described in claim 1, characterized in that, The step of finding resource combinations that satisfy the service scheduling requirement parameter values ​​in the cross-cloud resource dependency graph based on the path search algorithm to obtain candidate resource information includes: Based on the path search algorithm, find the resource combinations in the cross-cloud resource dependency graph that meet the service scheduling requirement parameter values, and determine the resource combinations to be detected; Dependency conflicts are detected for each resource combination in the resource combination to be detected, and the dependency conflict detection results are determined. Based on the dependency conflict detection results, target resources without dependency conflicts in the resource combination to be detected are selected to obtain candidate resource information.

5. The method as described in claim 1, characterized in that, After the step of finding resource combinations that meet the service scheduling requirement parameter values ​​in the cross-cloud resource dependency graph based on the path search algorithm to obtain candidate resource information, the method further includes: Calculate the corresponding objective function value based on the objective optimization function and the candidate resource information, and determine the overall function value; Based on the overall function value, the resource combinations that meet the service scheduling requirement parameter values ​​are sorted, and the resource combination with the highest ranking is taken as the candidate resource information.

6. The method as described in claim 5, characterized in that, The step of calculating the corresponding objective function value based on the objective optimization function and the candidate resource information, and determining the overall function value, includes: The objective function value corresponding to each resource combination in the candidate resource information is calculated based on the objective optimization function, and the resource utilization function value, response time function value, and cost function value are determined. The overall function value is obtained based on the resource utilization function value, the response time function value, and the cost function value.

7. The method as described in claim 1, characterized in that, The step of instantiating, creating or adjusting entity resources based on the scheduling strategy information, and generating cross-cloud collaborative resource scheduling information includes: Obtain interoperability resource scheduling information; Based on the scheduling strategy information, entity resources are instantiated and created to determine the first cross-cloud collaborative resource scheduling information; Based on the interoperability resource scheduling information, adjust the entity resources in the first cross-cloud collaborative resource scheduling information to determine the second cross-cloud collaborative resource scheduling information. Cross-cloud collaborative resource scheduling information is obtained based on the first cross-cloud collaborative resource scheduling information and the second cross-cloud collaborative resource scheduling information.

8. A cross-cloud collaborative scheduling device for a multi-dimensional resource dependency graph, characterized in that, The device includes: The acquisition module is used to obtain cross-cloud resource dependency graphs and business scheduling requirement parameter values; The processing module is used to find resource combinations that meet the service scheduling requirement parameter values ​​in the cross-cloud resource dependency graph based on the path search algorithm, and obtain candidate resource information, which includes regional preference resource information, performance index resource information and cost constraint resource information. The processing module is also used to generate corresponding resource scheduling configurations based on the regional preference resource information, the performance index resource information, and the cost constraint resource information, and to determine scheduling strategy information; The execution module is used to instantiate, create or adjust entity resources based on the scheduling strategy information, generate cross-cloud collaborative resource scheduling information, and complete cross-cloud collaborative scheduling based on the cross-cloud collaborative resource scheduling information.

9. A cross-cloud collaborative scheduling device based on a multi-dimensional resource dependency graph, characterized in that, The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the cross-cloud collaborative scheduling method for a multi-dimensional resource dependency graph as described in any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the cross-cloud collaborative scheduling method for the multi-dimensional resource dependency graph as described in any one of claims 1 to 7.

Images