Edge cloud resource collaborative scheduling method, cloud management platform and edge cloud node

Abstract

The present invention provides an edge cloud resource collaborative scheduling method, a cloud management platform and an edge cloud node. The edge cloud resource collaborative scheduling method applied to the cloud management platform includes: receiving the alarm information of the target resource sent by the first edge cloud node, the alarm information of the target resource is generated based on at least one of the following thresholds: the upper limit value of the resource allocation rate of the target resource, the lower limit value of the resource allocation rate, the upper limit value of the actual resource usage rate and the lower limit value of the actual resource usage rate; according to the alarm information of the target resource, the edge cloud resources are collaboratively scheduled. In the present invention, when performing the collaborative scheduling of cloud resources, the resource allocation problem and / or the actual resource usage problem of each preset resource in the edge cloud are taken into consideration, so that the resource waste problem in the edge cloud can be effectively reduced, and the load balancing between edge clouds can be improved, and the use of edge cloud resources can be more reasonably planned.

Description

Technical Field

[0001] Embodiments of the present invention relate to the field of edge cloud technology, and in particular to an edge cloud resource collaborative scheduling method, a cloud management platform, and an edge cloud node. Background Art

[0002] In order to reduce computing latency, the core cloud is moved down to meet customers' low-latency computing needs. Currently, the architecture of the edge cloud management system in the industry is mainly as follows: Figure 1 As shown in the figure, a cloud management platform (without distinguishing between MANO (Management and Orchestration, NFV management and orchestration), cloud management platform, etc.) is used to manage edge cloud resources. Appropriate edge cloud nodes are selected through the cloud management platform to deploy edge services and meet the business requirements for computing and latency.

[0003] The current edge cloud resource collaborative scheduling solution is that the cloud management platform queries the resource usage of edge cloud nodes through the northbound interface. When new services need to be deployed, nodes with sufficient resources are selected to deploy the services based on the remaining resources.

[0004] The existing edge cloud resource collaborative scheduling solutions have the following problems:

[0005] 1) The remaining resources in the edge cloud can only be understood manually or semi-automatically by the operator. Generally, the operator deploys related services by scheduling edge cloud nodes with sufficient remaining resources in the edge cloud. The operator cannot perceive the shortage of some resources in the edge cloud (such as bandwidth resources), which may cause the services deployed on the nodes to be unavailable, resulting in the failure of the final edge service deployment. However, the operator wastes resources to deploy the services throughout the life cycle, especially the sending of images occupies more network resources at the edge, which is unreasonable.

[0006] 2) Resource allocation problem: There are many types of resources in the edge cloud. In the same edge cloud, different resources are used. For example, storage is used up to 80%, but memory or CPU resources are only used about 40%. Edge resources themselves are relatively precious. For example, after the virtualization of content distribution network (CDN) services, there is a certain waste of resources, and resources need to be reasonably planned.

[0007] 3) The actual use of resources in the edge cloud: After the nodes are allocated resources, the actual use rate of resources by the business is too high, resulting in excessive load, or the actual use rate of resources by the business is low, resulting in less business load on some nodes, uneven distribution of resource usage, and failure to select nodes for deployment based on the actual use of cloud resource nodes. Summary of the invention

[0008] The embodiments of the present invention provide an edge cloud resource collaborative scheduling method, a cloud management platform and an edge cloud node, which are used to solve the problem that the existing edge cloud resource collaborative scheduling scheme is unreasonable.

[0009] In order to solve the above-mentioned technical problems, the present invention is achieved as follows:

[0010] In a first aspect, an embodiment of the present invention provides an edge cloud resource collaborative scheduling method, which is applied to a cloud management platform, including:

[0011] Receiving alarm information of a target resource sent by a first edge cloud node, wherein the alarm information of the target resource is generated based on at least one of the following thresholds: an upper limit value of a resource allocation rate of the target resource, a lower limit value of a resource allocation rate, an upper limit value of an actual resource usage rate, and a lower limit value of an actual resource usage rate;

[0012] According to the alarm information of the target resources, the edge cloud resources are coordinated and scheduled.

[0013] Optionally, according to the alarm information of the target resource, collaboratively scheduling resources in the edge cloud includes:

[0014] If the alarm information indicates that the resource allocation rate of the target resource is lower than the lower limit of its resource allocation rate, at least one of the following operations is performed on the first edge cloud node:

[0015] No new services are deployed for the first edge cloud node;

[0016] Generate a shutdown suggestion for shutting down the first edge cloud node and send it to a management system;

[0017] Generate a list of devices recommended to be powered off and notify the management system, where the devices recommended to be powered off are devices in the first edge cloud node that are not running any services;

[0018] Calculate the feasibility of business migration of the first edge cloud node, and if feasible, generate a business migration list and notify the management system;

[0019] If a service migration notification sent by the management system is received, a second edge cloud node that meets the resource requirements of the service to be migrated is selected, and the service to be migrated is deployed to the second edge cloud node;

[0020] Notify the management system that the migration of the to-be-migrated business has been completed, so that the management system shuts down the first edge cloud node.

[0021] Optionally, according to the alarm information of the target resource, collaboratively scheduling resources in the edge cloud includes:

[0022] If the alarm information indicates that the resource allocation rate of the target resource is higher than the upper limit of its resource allocation rate, no new services that meet the requirements of the target resource will be deployed for the first edge cloud node until the resource allocation rate of the target resource of the first edge cloud node is lower than the upper limit of its resource allocation rate.

[0023] Optionally, according to the alarm information of the target resource, collaboratively scheduling resources in the edge cloud includes:

[0024] If the alarm information indicates that the actual resource usage rate of the target resource is higher than the upper limit of the actual resource usage rate, selecting the service to be migrated of the first edge cloud node;

[0025] A second edge cloud node that meets the resource requirements of the service to be migrated is selected, and the service to be migrated is deployed to the second edge cloud node.

[0026] Optionally, selecting the service to be migrated of the first edge cloud node includes:

[0027] Selecting a service that meets a first preset condition from all or part of the services as the service to be migrated;

[0028] Among them, the first preset condition is: Min(R 实际 ≥|R 总实际 -R 总 *R1|);

[0029] Among them, R 实际 is the actual resource consumption value of the target resource used by the service, R 总实际 is the actual total resource usage of the first edge cloud, R 总 is the total resource amount of the first edge cloud, and R1 is the upper limit of the actual resource utilization rate of the target resource.

[0030] Optionally, the method further includes:

[0031] If no service that meets the first preset condition is selected from all or part of the services, the service that occupies the largest amount of the target resources in the first edge cloud node will be used as the service to be migrated until the actual utilization rate of the target resource is lower than the upper limit of its actual utilization rate.

[0032] Optionally, according to the alarm information of the target resource, collaboratively scheduling resources in the edge cloud includes:

[0033] If the alarm information indicates that the actual resource usage rate of the target resource is lower than the lower limit of the actual resource usage rate, select the unreasonable business of the first edge cloud node, and recommend downsizing, redeployment and / or migration of the unreasonable business;

[0034] Scale down the unreasonable business that needs to be scaled down, redeploy the unreasonable business on the first edge cloud node, and / or select a second edge cloud node that meets the resource requirements of the unreasonable business to be migrated, and deploy the business to be migrated to the second edge cloud node.

[0035] Optionally, the unreasonable business of selecting the first edge cloud node includes:

[0036] Select a business that meets the following criteria: 实际 ≤R 申请 *R2, where R 实际 is the actual resource consumption value of the target resource by the service, R 申请 is the resource value of the target resource applied for by the service in the first edge cloud node, and R2 is the lower limit of the actual resource usage rate of the target resource;

[0037] The selected services are sorted from high to low according to actual resource usage rates, and at least one service with a high actual resource usage rate is regarded as an unreasonable service.

[0038] Optionally, the second edge cloud node meets a second preset condition, and the second preset condition includes at least one of the following:

[0039] (R 1实际 +R 2实际 ) / R 2总 ≤R1, where R 1实际 is the resource consumption value of the target resource by the service at the first edge cloud node, R 2实际 is the actual resource consumption value of the target resource on the second edge cloud node, R 2总 is the total amount of resources of the second edge cloud node, and R1 is the upper limit of the actual resource utilization rate of the target resource;

[0040] The resource allocation rate in the second edge cloud node for the target resource is ≤ the upper limit of the resource allocation rate of the target resource;

[0041] D 2实际 +S≤D1*D 2可分配 , where D 2实际 is the actual resource allocation value in the second edge cloud node of the target resource, S is the required resource amount of the service to be migrated, D1 is the upper limit of the resource allocation rate of the target resource, and D 2可分配 is the actual allocatable amount of the target resource of the second edge cloud node;

[0042] Other resources except the target resource Other resources except the target resource meet at least one of the following corresponding thresholds: resource allocation rate upper limit value, resource allocation rate lower limit value, resource actual usage rate upper limit value and resource actual usage rate lower limit value.

[0043] Optionally, the method further includes:

[0044] If no edge cloud node that meets the second preset condition is screened out, a backup edge cloud node is enabled as the second edge cloud node, and the backup edge cloud node is an edge cloud node in a shutdown state.

[0045] Optionally, the method further includes:

[0046] Collecting actual resource usage and / or resource allocation rate of each preset resource of the edge cloud node;

[0047] Perform alarm analysis based on the collected actual resource usage rate and / or resource allocation rate of each preset resource within the preset time.

[0048] In a second aspect, an embodiment of the present invention provides an edge cloud resource collaborative scheduling method, which is applied to an edge cloud node, including:

[0049] Monitor the resource allocation rate and / or actual resource utilization rate of the preset resources in the edge cloud node in real time;

[0050] Generate alarm information for a target resource in the preset resources according to at least one of an upper limit value of a resource allocation rate, a lower limit value of a resource allocation rate, an upper limit value of an actual resource usage rate, and a lower limit value of an actual resource usage rate of each preset resource;

[0051] Send the target resource warning information to the cloud management platform.

[0052] Optionally, the preset resources include at least one of the following: device resources within the edge cloud node, network resources within the edge cloud, and firewall resources within the edge cloud.

[0053] Optionally, generating the alarm information of the target resource according to at least one of the upper limit value of the resource allocation rate, the lower limit value of the resource allocation rate, the upper limit value of the actual resource usage rate, and the lower limit value of the actual resource usage rate of each preset resource includes:

[0054] If at least one of the following is true, an alarm message is generated for the target resource:

[0055] The resource allocation rate of the target resource in the preset resources is lower than the lower limit of the resource allocation rate;

[0056] The resource allocation rate of the target resource in the preset resources is higher than the upper limit of the resource allocation rate;

[0057] The actual resource usage rate of the target resource in the preset resources is higher than the upper limit value of the actual resource usage rate;

[0058] The actual resource usage rate of the target resource in the preset resources is lower than the lower limit of the actual resource usage rate.

[0059] In a third aspect, an embodiment of the present invention provides a cloud management platform, including:

[0060] The client edge cloud resource monitoring module is used to receive the alarm information of the target resource sent by the first edge cloud node, wherein the alarm information of the target resource is generated based on at least one of the following thresholds: the upper limit value of the resource allocation rate of the target resource, the lower limit value of the resource allocation rate, the upper limit value of the actual resource usage rate, and the lower limit value of the actual resource usage rate;

[0061] The collaborative scheduling module is used to collaboratively schedule edge cloud resources according to the alarm information of the target resources.

[0062] Optionally, the collaborative scheduling module is configured to perform at least one of the following operations on the first edge cloud node if the alarm information indicates that the resource allocation rate of the target resource is lower than the lower limit of its resource allocation rate:

[0063] No new services are deployed for the first edge cloud node;

[0064] Generate a shutdown suggestion for shutting down the first edge cloud node and send it to a management system;

[0065] Generate a list of devices recommended to be powered off and notify the management system, where the devices recommended to be powered off are devices in the first edge cloud node that are not running any services;

[0066] Calculate the feasibility of business migration of the first edge cloud node, and if feasible, generate a business migration list and notify the management system;

[0067] If a service migration notification sent by the management system is received, a second edge cloud node that meets the resource requirements of the service to be migrated is selected, and the service to be migrated is deployed to the second edge cloud node;

[0068] Notify the management system that the migration of the to-be-migrated business has been completed, so that the management system shuts down the first edge cloud node.

[0069] Optionally, the collaborative scheduling module is used to no longer deploy new services that require the target resources for the first edge cloud node if the alarm information indicates that the resource allocation rate of the target resource is higher than the upper limit of its resource allocation rate, until the resource allocation rate of the target resource of the first edge cloud node is lower than the upper limit of its resource allocation rate.

[0070] Optionally, the collaborative scheduling module is used to select the service to be migrated of the first edge cloud node if the alarm information indicates that the actual resource utilization rate of the target resource is higher than the upper limit value of the actual resource utilization rate;

[0071] A second edge cloud node that meets the resource requirements of the service to be migrated is selected, and the service to be migrated is deployed to the second edge cloud node.

[0072] Optionally, the collaborative scheduling module is used to select a service that meets a first preset condition from all or part of the services as the service to be migrated;

[0073] Wherein, the first preset condition is:

[0074] Min(R 实际 ≥|R 总实际 -R 总 *R1|).

[0075] Among them, R 实际 is the actual resource consumption value of the target resource used by the service, R 总实际 is the actual total resource usage of the first edge cloud, R 总 is the total resource amount of the first edge cloud, and R1 is the upper limit of the actual resource utilization rate of the target resource.

[0076] Optionally, the collaborative scheduling module is used to select the business that occupies the largest amount of the target resources in the first edge cloud node as the business to be migrated if no business that meets the first preset condition is selected from all or part of the businesses, until the actual resource utilization rate of the target resource is lower than the upper limit of its actual resource utilization rate.

[0077] Optionally, the collaborative scheduling module is used to select the unreasonable business of the first edge cloud node if the alarm information indicates that the actual resource utilization rate of the target resource is lower than the lower limit of the actual resource utilization rate, and recommend a reduction specification, a redeployment recommendation and / or a migration recommendation for the unreasonable business;

[0078] Scale down the unreasonable business that needs to be scaled down, redeploy the unreasonable business on the first edge cloud node, and / or select a second edge cloud node that meets the resource requirements of the unreasonable business to be migrated, and deploy the business to be migrated to the second edge cloud node.

[0079] Optionally, the unreasonable business of selecting the first edge cloud node includes:

[0080] Select a business that meets the following criteria: 实际 ≤R 申请 *R2, where R实际 is the actual resource consumption value of the target resource by the service, R 申请 is the resource value of the target resource applied for by the service in the first edge cloud node, and R2 is the lower limit of the actual resource usage rate of the target resource;

[0081] The selected services are sorted from high to low according to actual resource usage rates, and at least one service with a high actual resource usage rate is regarded as an unreasonable service.

[0082] Optionally, the second edge cloud node meets a second preset condition, and the second preset condition includes at least one of the following:

[0083] (R 1实际 +R 2实际 ) / R 2总 ≤R1, where R 1实际 is the resource consumption value of the target resource by the service at the first edge cloud node, R 2实际 is the actual resource consumption value of the target resource on the second edge cloud node, R 2总 is the total amount of resources of the second edge cloud node, and R1 is the upper limit of the actual resource utilization rate of the target resource;

[0084] The resource allocation rate in the cloud for the target resource is ≤ the upper limit of the resource allocation rate of the target resource;

[0085] D 2实际 +S≤D1*D 2可分配 , where D 2实际 is the actual resource allocation value in the second edge cloud node of the target resource, S is the required resource amount of the service to be migrated, D1 is the upper limit of the resource allocation rate of the target resource, and D 2可分配 is the actual allocatable amount of the target resource of the second edge cloud node;

[0086] Other resources except the target resource Other resources except the target resource meet at least one of the following corresponding thresholds: resource allocation rate upper limit value, resource allocation rate lower limit value, resource actual usage rate upper limit value and resource actual usage rate lower limit value.

[0087] Optionally, the collaborative scheduling module is also used to enable a backup edge cloud node as the second edge cloud node if an edge cloud node that meets the second preset condition is not screened out, and the backup edge cloud node is an edge cloud node in a shutdown state.

[0088] Optionally, the collaborative scheduling module is also used to collect the actual resource utilization rate and / or resource allocation rate of each preset resource of the edge cloud node; and perform alarm analysis based on the collected actual resource utilization rate and / or resource allocation rate of each preset resource within a preset time.

[0089] In a fourth aspect, an embodiment of the present invention provides an edge cloud node, including:

[0090] The agent-side edge cloud resource monitoring module is used to monitor the resource allocation rate and / or actual resource utilization rate of the preset resources in the edge cloud node in real time; generate alarm information for the target resources in the preset resources based on at least one of the upper limit value of the resource allocation rate, the lower limit value of the resource allocation rate, the upper limit value of the actual resource utilization rate and the lower limit value of the actual resource utilization rate of each preset resource; and send the alarm information of the target resources to the cloud management platform.

[0091] Optionally, the preset resources include at least one of the following: device resources within the edge cloud node, network resources within the edge cloud, and firewall resources within the edge cloud.

[0092] Optionally, the proxy-side edge cloud resource monitoring module is used to generate alarm information of the target resource if at least one of the following is met:

[0093] The resource allocation rate of the target resource in the preset resources is lower than the lower limit of the resource allocation rate;

[0094] The resource allocation rate of the target resource in the preset resources is higher than the upper limit of the resource allocation rate;

[0095] The actual resource usage rate of the target resource in the preset resources is higher than the upper limit value of the actual resource usage rate;

[0096] The actual resource usage rate of the target resource in the preset resources is lower than the lower limit of the actual resource usage rate.

[0097] In an embodiment of the present invention, when performing collaborative scheduling of cloud resources, the resource allocation problem and / or the actual resource usage problem of each preset resource in the edge cloud is taken into consideration, and the edge cloud resources are collaboratively scheduled, and the business is redeployed or migrated, thereby effectively reducing the problem of resource waste in the edge cloud, improving the load balancing between edge clouds, and more reasonably planning the use of edge cloud resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0098] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0099] Figure 1 It is a schematic diagram of the architecture of an edge cloud management system in the prior art;

[0100] Figure 2 A schematic diagram of a process flow of an edge cloud resource collaborative scheduling method applied to a cloud management platform according to an embodiment of the present invention;

[0101] Figure 3 A schematic diagram of a flow chart of an edge cloud resource collaborative scheduling method applied to an edge cloud node according to an embodiment of the present invention;

[0102] Figure 4 This is a schematic diagram of the structure of a cloud management platform according to an embodiment of the present invention;

[0103] Figure 5 A schematic diagram of the structure of an edge cloud node according to an embodiment of the present invention;

[0104] Figure 6 A schematic diagram of the architecture of an edge cloud management system according to an embodiment of the present invention;

[0105] Figure 7 A schematic diagram of the interaction process of the edge cloud resource collaborative scheduling method according to an embodiment of the present invention;

[0106] Figure 8 A schematic diagram of an interactive process of an edge cloud resource collaborative scheduling method according to another embodiment of the present invention;

[0107] Fig. 9 A schematic diagram of an interactive process of an edge cloud resource collaborative scheduling method according to another embodiment of the present invention;

[0108] Fig.10 A schematic diagram of an interactive process of an edge cloud resource collaborative scheduling method according to another embodiment of the present invention;

[0109] Fig.11 A schematic diagram of the structure of a cloud management platform according to another embodiment of the present invention;

[0110] Fig.12 This is a schematic diagram of the structure of an edge cloud node according to another embodiment of the present invention. DETAILED DESCRIPTION

[0111] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0112] Please refer to Figure 2The embodiment of the present invention provides an edge cloud resource collaborative scheduling method, which is applied to a cloud management platform. The method includes:

[0113] Step 21: receiving alarm information of a target resource sent by a first edge cloud node, wherein the alarm information of the target resource is generated based on at least one of the following thresholds: an upper limit value of a resource allocation rate of the target resource, a lower limit value of a resource allocation rate, an upper limit value of an actual resource usage rate, and a lower limit value of an actual resource usage rate;

[0114] In an embodiment of the present invention, a cloud management platform is connected to multiple edge cloud nodes for collaboratively scheduling resources of the multiple edge cloud nodes, and the first edge cloud node is any edge cloud node connected to the cloud management platform for management.

[0115] Each edge cloud node has a variety of resources, each resource has a corresponding threshold, and the threshold includes at least one of the following: an upper limit value of resource allocation rate, a lower limit value of resource allocation rate, an upper limit value of actual resource usage rate, and a lower limit value of actual resource usage rate;

[0116] For example, the threshold of a resource is as follows:

[0117] Resource allocation rate upper limit (D1): 100%;

[0118] Resource allocation rate lower limit (D2): 30%;

[0119] The upper limit of actual resource utilization (R1): 80%;

[0120] The lower limit of actual resource utilization rate (R2): 50%.

[0121] The thresholds of various resources can be the same or different. The thresholds of various resources can be adjusted as needed.

[0122] The resource allocation rate of each resource refers to: the resource allocation amount of the resource / the total resource amount of the resource in the edge cloud node.

[0123] The actual resource usage rate of each resource refers to: the actual usage of the resource / the total resource usage of the resource in the edge cloud node.

[0124] Based on at least one threshold of a resource, it can be determined whether an alarm condition is met. If the alarm condition is met, the resource is used as a target resource, and the edge cloud node generates alarm information of the target resource and sends it to the cloud management platform.

[0125] Step 22: Coordinated scheduling of edge cloud resources based on the alarm information of the target resources.

[0126] In an embodiment of the present invention, when performing collaborative scheduling of cloud resources, the resource allocation problem and / or the actual resource usage problem of each preset resource in the edge cloud is taken into consideration, and the edge cloud resources are collaboratively scheduled, and the business is redeployed or migrated, thereby effectively reducing the problem of resource waste in the edge cloud, improving the load balancing between edge clouds, and more reasonably planning the use of edge cloud resources.

[0127] In the embodiment of the present invention, the cloud management platform can coordinate the scheduling of edge cloud resources based on the following resource scheduling scheme principles:

[0128] Principle 1: Avoid running at no load or very low resource allocation rate

[0129] For resource allocation, the resource allocation rate of each resource shall not be lower than the lower limit of the resource allocation rate D2. Otherwise, it is recommended that the administrator set the edge cloud node to a state to be shut down and no longer deploy new services for the first edge cloud node; calculate the feasibility of business migration of the first edge cloud node, and if feasible, generate a business migration list; power off some unused devices; and, through manual evaluation of whether it is necessary to continue to maintain the edge cloud node or whether there is an unreasonable resource planning problem with the node, rectify the node, set the node after rectification to a shutdown state, and start the service on demand according to the resource usage, that is, shut down some or all machines in the edge cloud nodes with lower activity, and then power on the machines as appropriate according to the utilization of edge cloud resources.

[0130] The resource allocation rate of each resource of the edge cloud node can reach its maximum resource allocation rate upper limit D1. For an edge cloud node after a certain resource reaches its resource allocation rate upper limit D1, no new edge services with this resource demand will be deployed until the resource allocation rate of the target resource of the first edge cloud node is lower than its resource allocation rate upper limit D1. Then, it can continue to enter the edge cloud scheduling system with this resource for new service deployment.

[0131] Principle 2: Avoid resource overload

[0132] When there are too many services deployed on an edge cloud node, through dynamic monitoring, it is found that the actual resource usage rate of a certain resource actually used by the services deployed in the edge cloud node rises to above the upper limit of the actual resource usage rate R1. The services deployed in the edge cloud node are analyzed to find the services with higher actual resource usage rates for the resource, and the services to be migrated are evaluated. The evaluation method is as follows:

[0133] Selecting a service that meets a first preset condition from all or part of the services as the service to be migrated;

[0134] Among them, the first preset condition is: Min(R实际 ≥|R 总实际 -R 总 *R1|);

[0135] Among them, R 实际 is the actual resource consumption value of the target resource used by the service, R 总实际 is the actual total resource usage of the first edge cloud, R 总 is the total resource amount of the first edge cloud, and R1 is the upper limit of the actual resource utilization rate of the target resource.

[0136] If there is a business that meets the first preset condition, after finding the business, record the actual resource usage indicator parameters of the business, filter the edge cloud cluster for the business in the edge cloud system, and filter out the edge cloud nodes that meet the basic resource and location requirements of the business in the system. The selection method of the edge cloud nodes is as follows:

[0137] Filtering an edge cloud node that meets a second preset condition as the second edge cloud node;

[0138] The second preset condition is:

[0139] (the resource consumption value of the target resource by the service at the first edge cloud node + the actual resource consumption value of the target resource at the second edge cloud node) / the total resource amount of the second edge cloud node ≤ the upper limit value of the actual resource usage rate of the target resource;

[0140] The resource allocation rate in the second edge cloud node for the target resource is ≤ the upper limit of the resource allocation rate of the target resource;

[0141] The actual resource allocation value in the second edge cloud node of the target resource + the required resource amount of the service to be migrated ≤ the upper limit of the resource allocation rate of the target resource * the actual allocatable amount of the target resource of the second edge cloud node;

[0142] Other resources except the target resource Other resources except the target resource meet at least one of the following corresponding thresholds: resource allocation rate upper limit value, resource allocation rate lower limit value, resource actual usage rate upper limit value and resource actual usage rate lower limit value.

[0143] After the screening is completed, the service is deployed on the edge cloud node that meets the second preset condition;

[0144] If no edge cloud node that meets the second preset condition is screened out, a backup edge cloud node is enabled as the second edge cloud node, and the backup edge cloud node is an edge cloud node in a shutdown state.

[0145] If no service that meets the first preset condition is selected from all or part of the services, the service that occupies the largest amount of the target resources in the first edge cloud node will be used as the service to be migrated until the actual utilization rate of the target resource is lower than the upper limit of its actual utilization rate.

[0146] Principle 3: Avoid low actual resource utilization

[0147] When the actual resource utilization rate of some resources in the services deployed on the edge cloud node is lower than the lower limit of the actual resource utilization rate, it means that the edge cloud node has a waste of service allocation resources in resource allocation. It is necessary to screen the services with serious resource waste and scale them down or redeploy them on the edge cloud node according to the recommended resource allocation or migrate them. The following conditions should be followed in the screening method of services:

[0148] The actual resource consumption value of the target resource by the service ≤ the resource value of the target resource applied by the service in the first edge cloud node * the lower limit of the actual resource utilization rate of the target resource. Filter out unreasonable services and sort the services from high to low according to the actual resource utilization rate.

[0149] Principle 4: Screening edge cloud nodes

[0150] When selecting edge cloud nodes, comprehensive scheduling is required based on the edge cloud resource allocation and actual resource usage. In addition to meeting the requirements for edge cloud node latency and basic resources, the node resources should also meet the following requirements:

[0151] The resource allocation rate of the target resource is ≤ the upper limit of the resource allocation rate of the target resource;

[0152] The required resource volume of the service to be migrated is ≤ the total resource volume of the second edge cloud node*the upper limit of the actual resource usage rate of the target resource-the actual usage volume of the target resource of the second edge cloud node.

[0153] In the embodiment of the present invention, optionally, the actual resource consumption value may be an average monthly usage value, and the value calculation method is not limited.

[0154] In the embodiment of the present invention, when edge cloud nodes are screened, some resource-complementary services may be migrated to the own edge cloud services and third-party services to achieve load balancing of resource usage.

[0155] In the embodiment of the present invention, the solution of using core cloud storage for part of the services of edge cloud nodes is also within the scope of this solution, and the scheduling effect is the same as the scheduling without core cloud storage.

[0156] Based on the above principles, in some embodiments of the present invention, optionally, according to the alarm information of the target resource, the collaborative scheduling of the resources in the edge cloud includes:

[0157] If the alarm information indicates that the resource allocation rate of the target resource is lower than the lower limit of its resource allocation rate, at least one of the following operations is performed on the first edge cloud node:

[0158] No new services are deployed for the first edge cloud node;

[0159] Generate a shutdown suggestion for shutting down the first edge cloud node and send it to a management system;

[0160] Generate a list of devices recommended to be powered off and notify the management system, where the devices recommended to be powered off are devices in the first edge cloud node that are not running any services;

[0161] Calculate the feasibility of business migration of the first edge cloud node, and if feasible, generate a business migration list and notify the management system;

[0162] If a service migration notification sent by the management system is received, a second edge cloud node that meets the resource requirements of the service to be migrated is selected, and the service to be migrated is deployed to the second edge cloud node;

[0163] Notify the management system that the migration of the to-be-migrated business has been completed, so that the management system shuts down the first edge cloud node.

[0164] In an embodiment of the present invention, the management system may be a cloud management system, an edge cloud management system, a hardware management system, or the like.

[0165] In some embodiments of the present invention, optionally, coordinating the resources in the edge cloud according to the alarm information of the target resource includes:

[0166] If the alarm information indicates that the resource allocation rate of the target resource is higher than the upper limit of its resource allocation rate, no new services that meet the requirements of the target resource will be deployed for the first edge cloud node until the resource allocation rate of the target resource of the first edge cloud node is lower than the upper limit of its resource allocation rate.

[0167] In some embodiments of the present invention, optionally, coordinating the resources in the edge cloud according to the alarm information of the target resource includes:

[0168] If the alarm information indicates that the actual resource usage rate of the target resource is higher than the upper limit of the actual resource usage rate, selecting the service to be migrated of the first edge cloud node;

[0169] A second edge cloud node that meets the resource requirements of the service to be migrated is selected, and the service to be migrated is deployed to the second edge cloud node.

[0170] Optionally, selecting the service to be migrated of the first edge cloud node includes:

[0171] Selecting a service that meets a first preset condition from all or part of the services as the service to be migrated;

[0172] Among them, the first preset condition is: Min(R 实际 ≥|R 总实际 -R 总 *R1|).

[0173] Among them, R 实际 is the actual resource consumption value of the target resource used by the service, R 总实际 is the actual total resource usage of the first edge cloud, R 总 is the total resource amount of the first edge cloud, and R1 is the upper limit of the actual resource utilization rate of the target resource.

[0174] Optionally, if a service that meets the first preset condition is not selected from all or part of the services, the service that occupies the largest amount of the target resources in the first edge cloud node is used as the service to be migrated.

[0175] In some embodiments of the present invention, optionally, coordinating the resources in the edge cloud according to the alarm information of the target resource includes:

[0176] If the alarm information indicates that the actual resource usage rate of the target resource is lower than the lower limit of the actual resource usage rate, select the unreasonable business of the first edge cloud node, and recommend downsizing, redeployment and / or migration of the unreasonable business;

[0177] Scale down the unreasonable business that needs to be scaled down, redeploy the unreasonable business on the first edge cloud node, and / or select a second edge cloud node that meets the resource requirements of the unreasonable business to be migrated, and deploy the business to be migrated to the second edge cloud node.

[0178] Optionally, the unreasonable business of selecting the first edge cloud node includes:

[0179] Select a business that meets the following criteria: 实际 ≤R 申请 *R2, where R 实际 is the actual resource consumption value of the target resource by the service, R 申请is the resource value of the target resource applied for by the service in the first edge cloud node, and R2 is the lower limit of the actual resource usage rate of the target resource;

[0180] The selected services are sorted from high to low according to actual resource usage rates, and at least one service with a high actual resource usage rate is regarded as an unreasonable service.

[0181] In the above embodiments, optionally, the second edge cloud node meets a second preset condition, and the second preset condition includes at least one of the following:

[0182] (R 1实际 +R 2实际 ) / R 2总 ≤R1, where R 1实际 is the resource consumption value of the target resource by the service at the first edge cloud node, R 2实际 is the actual resource consumption value of the target resource on the second edge cloud node, R 2总 is the total amount of resources of the second edge cloud node, and R1 is the upper limit of the actual resource utilization rate of the target resource;

[0183] The resource allocation rate in the second edge cloud node for the target resource is ≤ the upper limit of the resource allocation rate of the target resource;

[0184] D 2实际 +S≤D1*D 2可分配 , where D 2实际 is the actual resource allocation value in the second edge cloud node of the target resource, S is the required resource amount of the service to be migrated, D1 is the upper limit of the resource allocation rate of the target resource, and D 2可分配 is the actual allocatable amount of the target resource of the second edge cloud node;

[0185] Other resources except the target resource Other resources except the target resource meet at least one of the following corresponding thresholds: resource allocation rate upper limit value, resource allocation rate lower limit value, resource actual usage rate upper limit value and resource actual usage rate lower limit value.

[0186] Optionally, the method further includes:

[0187] If no edge cloud node that meets the second preset condition is screened out, a backup edge cloud node is enabled as the second edge cloud node, and the backup edge cloud node is an edge cloud node in a shutdown state.

[0188] Optionally, the method further includes:

[0189] Collecting actual resource usage and / or resource allocation rate of each preset resource of the edge cloud node;

[0190] Perform alarm analysis based on the collected actual resource usage rate and / or resource allocation rate of each preset resource within the preset time.

[0191] For example, the cloud management platform collects information about edge cloud nodes, such as CPU usage of 40%, CPU allocation of 20%, and network bandwidth usage of 80%. Usually, part of the usage is real-time usage, which requires further calculation, such as calculating the monthly (or weekly or daily) average usage, and obtaining the monthly (or weekly or daily) average usage and (or weekly or daily) average allocation rate, and issuing alarms based on the calculation results.

[0192] Please refer to Figure 3 The embodiment of the present invention further provides an edge cloud resource collaborative scheduling method, which is applied to an edge cloud node. The method includes:

[0193] Step 31: Monitor the resource allocation rate and / or actual resource usage rate of the preset resources in the edge cloud node in real time;

[0194] Step 32: Generate alarm information of a target resource in the preset resources according to at least one of the upper limit value of the resource allocation rate, the lower limit value of the resource allocation rate, the upper limit value of the actual resource usage rate, and the lower limit value of the actual resource usage rate of each preset resource;

[0195] Step 33: Send the alarm information of the target resource to the cloud management platform.

[0196] In the embodiment of the present invention, each edge cloud node is monitored in real time, which solves the problem of no one or few people on duty in the edge cloud, and provides a more accurate understanding of the real-time information of resources in the edge cloud, which is more conducive to the rational use of widely dispersed edge cloud resources.

[0197] In the embodiment of the present invention, the implementation method of the edge cloud node is not limited, for example, it can be OpenStack, Kubernetes or VMware EXSi.

[0198] In an embodiment of the present invention, optionally, the preset resources include at least one of the following: device resources in edge cloud nodes, network resources in edge cloud, and firewall resources in edge cloud. The device resources include physical resources and virtual resources, and may include at least one of the following and / or are not limited to the following: CPU (central processing unit), memory, GPU (graphics processing unit), and network port, etc. The network resources in the edge cloud may include at least one of the following and / or are not limited to the following: node export form, node export bandwidth, etc.

[0199] The embodiments of the present invention solve the following problems: there is no monitoring module for various resources in the edge cloud. Currently, the monitoring of resources in the edge cloud is mainly carried out at the virtualization platform level to monitor the network bandwidth, CPU, storage and other aspects of a single host. The scope of resource monitoring needs to be improved. For example, there are certain defects in server hardware, switch network resources, edge cloud export bandwidth and other aspects.

[0200] In the embodiment of the present invention, the resources in the edge cloud are fully acquired, which requires that the edge cloud node can dynamically collect the resource usage of the host, virtual machine, container, bare metal and / or edge cloud network in the edge cloud.

[0201] In the embodiment of the present invention, optionally, generating the alarm information of the target resource according to at least one of the upper limit value of the resource allocation rate, the lower limit value of the resource allocation rate, the upper limit value of the actual resource usage rate, and the lower limit value of the actual resource usage rate of each preset resource includes:

[0202] If at least one of the following is true, an alarm message is generated for the target resource:

[0203] The resource allocation rate of the target resource in the preset resources is lower than the lower limit of the resource allocation rate;

[0204] The resource allocation rate of the target resource in the preset resources is higher than the upper limit of the resource allocation rate;

[0205] The actual resource usage rate of the target resource in the preset resources is higher than the upper limit value of the actual resource usage rate;

[0206] The actual resource usage rate of the target resource in the preset resources is lower than the lower limit of the actual resource usage rate.

[0207] Please refer to Figure 4 , an embodiment of the present invention provides a cloud management platform, including:

[0208] The client edge cloud resource monitoring module 41 is used to receive the alarm information of the target resource sent by the first edge cloud node, where the alarm information of the target resource is generated based on at least one of the following thresholds: the upper limit value of the resource allocation rate of the target resource, the lower limit value of the resource allocation rate, the upper limit value of the actual resource usage rate, and the lower limit value of the actual resource usage rate;

[0209] The collaborative scheduling module 41 is used to collaboratively schedule edge cloud resources according to the alarm information of the target resources.

[0210] Optionally, the collaborative scheduling module 42 is configured to perform at least one of the following operations on the first edge cloud node if the alarm information indicates that the resource allocation rate of the target resource is lower than the lower limit of its resource allocation rate:

[0211] No new services are deployed for the first edge cloud node;

[0212] Generate a shutdown suggestion for shutting down the first edge cloud node and send it to a management system;

[0213] Generate a list of devices recommended to be powered off and notify the management system, where the devices recommended to be powered off are devices in the first edge cloud node that are not running any services;

[0214] Calculate the feasibility of business migration of the first edge cloud node, and if feasible, generate a business migration list and notify the management system;

[0215] If a service migration notification sent by the management system is received, a second edge cloud node that meets the resource requirements of the service to be migrated is selected, and the service to be migrated is deployed to the second edge cloud node;

[0216] Notify the management system that the migration of the to-be-migrated business has been completed, so that the management system shuts down the first edge cloud node.

[0217] Optionally, the collaborative scheduling module 42 is used to no longer deploy new services that require the target resources for the first edge cloud node if the alarm information indicates that the resource allocation rate of the target resource is higher than the upper limit of its resource allocation rate, until the resource allocation rate of the target resource of the first edge cloud node is lower than the upper limit of its resource allocation rate.

[0218] Optionally, the collaborative scheduling module 42 is configured to select the service to be migrated of the first edge cloud node if the alarm information indicates that the actual resource usage rate of the target resource is higher than the upper limit value of the actual resource usage rate;

[0219] A second edge cloud node that meets the resource requirements of the service to be migrated is selected, and the service to be migrated is deployed to the second edge cloud node.

[0220] Optionally, the collaborative scheduling module is used to select a service that meets a first preset condition from all or part of the services as the service to be migrated;

[0221] Among them, the first preset condition is: Min(R 实际 ≥|R 总实际 -R 总 *R1|).

[0222] Among them, R 实际 is the actual resource consumption value of the target resource used by the service, R 总实际 is the actual total resource usage of the first edge cloud, R 总 is the total resource amount of the first edge cloud, and R1 is the upper limit of the actual resource utilization rate of the target resource.

[0223] Optionally, the collaborative scheduling module is used to select the business that occupies the largest amount of the target resources in the first edge cloud node as the business to be migrated if no business that meets the first preset condition is selected from all or part of the businesses, until the actual resource utilization rate of the target resource is lower than the upper limit of its actual resource utilization rate.

[0224] Optionally, the collaborative scheduling module 42 is configured to select an unreasonable service of the first edge cloud node if the alarm information indicates that the actual resource utilization rate of the target resource is lower than the lower limit of the actual resource utilization rate thereof, and to provide a recommended downsizing specification, a recommended redeployment, and / or a recommended migration for the unreasonable service;

[0225] Scale down the unreasonable business that needs to be scaled down, redeploy the unreasonable business on the first edge cloud node, and / or select a second edge cloud node that meets the resource requirements of the unreasonable business to be migrated, and deploy the business to be migrated to the second edge cloud node.

[0226] Optionally, the unreasonable business of selecting the first edge cloud node includes:

[0227] Select a business that meets the following criteria: 实际 ≤R 申请 *R2, where R 实际 is the actual resource consumption value of the target resource by the service, R 申请 is the resource value of the target resource applied for by the service in the first edge cloud node, and R2 is the lower limit of the actual resource usage rate of the target resource;

[0228] The selected services are sorted from high to low according to actual resource usage rates, and at least one service with a high actual resource usage rate is regarded as an unreasonable service.

[0229] Optionally, the second edge cloud node meets a second preset condition, and the second preset condition includes at least one of the following:

[0230] (R 1实际 +R 2实际 ) / R 2总 ≤R1, where R 1实际 is the resource consumption value of the target resource by the service at the first edge cloud node, R 2实际 is the actual resource consumption value of the target resource on the second edge cloud node, R 2总 is the total amount of resources of the second edge cloud node, and R1 is the upper limit of the actual resource utilization rate of the target resource;

[0231] The actual resource allocation rate in the second edge cloud node of the target resource is ≤ the upper limit value of the resource allocation rate of the target resource;

[0232] D 2实际 +S≤D1*D 2可分配 , where D 2实际 is the actual resource allocation value in the second edge cloud node of the target resource, S is the required resource amount of the service to be migrated, D1 is the upper limit of the resource allocation rate of the target resource, and D 2可分配 is the actual allocatable amount of the target resource of the second edge cloud node;

[0233] Other resources except the target resource Other resources except the target resource meet at least one of the following corresponding thresholds: resource allocation rate upper limit value, resource allocation rate lower limit value, resource actual usage rate upper limit value and resource actual usage rate lower limit value.

[0234] Optionally, the collaborative scheduling module is also used to enable a backup edge cloud node as the second edge cloud node if an edge cloud node that meets the second preset condition is not screened out, and the backup edge cloud node is an edge cloud node in a shutdown state.

[0235] Optionally, the collaborative scheduling module is also used to collect the actual resource utilization rate and / or resource allocation rate of each preset resource of the edge cloud node; and perform alarm analysis based on the collected actual resource utilization rate and / or resource allocation rate of each preset resource within a preset time.

[0236] Please refer to Figure 5 , an embodiment of the present invention provides an edge cloud node 50, including:

[0237] The agent-side edge cloud resource monitoring module 51 is used to monitor the resource allocation rate and / or actual resource utilization rate of the preset resources in the edge cloud node in real time; generate alarm information of the target resources in the preset resources according to at least one of the upper limit value of the resource allocation rate, the lower limit value of the resource allocation rate, the upper limit value of the actual resource utilization rate and the lower limit value of the actual resource utilization rate of each preset resource; and send the alarm information of the target resources to the cloud management platform.

[0238] Optionally, the preset resources include at least one of the following and / or are not limited to the following items: device resources within the edge cloud node, network resources within the edge cloud, and firewall resources within the edge cloud.

[0239] Optionally, the proxy-side edge cloud resource monitoring module is used to generate alarm information of the target resource if at least one of the following is met:

[0240] The resource allocation rate of the target resource in the preset resources is lower than the lower limit of the resource allocation rate;

[0241] The resource allocation rate of the target resource in the preset resources is higher than the upper limit of the resource allocation rate;

[0242] The actual resource usage rate of the target resource in the preset resources is higher than the upper limit value of the actual resource usage rate;

[0243] The actual resource usage rate of the target resource in the preset resources is lower than the lower limit of the actual resource usage rate.

[0244] Please refer to Figure 6 , Figure 6 The schematic diagram of the architecture of an edge cloud management system according to an embodiment of the present invention includes: a cloud management platform and multiple edge cloud nodes.

[0245] Each edge cloud node includes:

[0246] The agent-side edge cloud resource monitoring module 61 is used to monitor the resource allocation rate and / or actual resource utilization rate of the preset resources in the edge cloud node in real time; generate alarm information of the target resources in the preset resources according to at least one of the upper limit value of the resource allocation rate, the lower limit value of the resource allocation rate, the upper limit value of the actual resource utilization rate and the lower limit value of the actual resource utilization rate of each preset resource; and send the alarm information of the target resources to the cloud management platform.

[0247] The cloud management platform includes:

[0248] The client edge cloud resource monitoring module is used to receive the alarm information of the target resource sent by the edge cloud node, and the alarm information of the target resource is generated based on at least one of the following thresholds: the upper limit value of the resource allocation rate of the target resource, the lower limit value of the resource allocation rate, the upper limit value of the actual resource usage rate and the lower limit value of the actual resource usage rate, and read the data of each resource in the edge cloud node;

[0249] The edge cloud resource scheduling and calculation module is a module used for edge cloud resource scheduling and calculation. It is mainly used to calculate the data read by the client edge cloud resource monitoring module, give the selection of edge cloud nodes, reduce the probability of business deployment failure and balance resource usage distribution, etc.

[0250] Edge cloud service deployment / migration module: This module can migrate or redeploy the selected related services based on the threshold alarm function given by the client edge cloud resource monitoring module.

[0251] The edge cloud resource scheduling calculation module and the edge cloud service deployment / migration module can realize the various functions used by the collaborative scheduling module in the above-mentioned embodiment, which will not be described one by one here.

[0252] The following describes the edge cloud resource collaborative scheduling method of an embodiment of the present invention in combination with the above-mentioned edge cloud management system.

[0253] Embodiment 1

[0254] Scenario 1: Ultra-low load operation (no longer maintained)

[0255] Solve the problem of low resource allocation rate, monitor and avoid idle operation of devices in edge cloud nodes, and adjust some unreasonable overall resource situations.

[0256] Principles involved: Principle 1: avoid no-load and ultra-low resource allocation rate operation; Principle 4: screen edge cloud nodes.

[0257] Please refer to Figure 7 , the edge cloud resource collaborative scheduling method of the embodiment of the present invention is as follows:

[0258] Step 1: The proxy edge cloud resource monitoring module in the first edge cloud node detects that the resource allocation rate of the target resource is lower than the lower limit D2 of its resource allocation rate, generates an alarm message and reports it to the client edge cloud resource monitoring module in the cloud management platform.

[0259] Step 2: The client edge cloud resource monitoring module reports shutdown suggestions to the edge cloud computing module.

[0260] Step 3: After calculation, the edge cloud computing module finds that there are devices in the first edge cloud node that are not running services, and generates a list of recommended devices to be powered off. If the first edge cloud node carries fewer services, the feasibility of service migration is calculated.

[0261] Step 4: Send a shutdown suggestion to the management system, along with a list of recommended devices to be powered off and a list of services to be migrated.

[0262] Step 5: The management system powers off the devices according to the recommended power-off device list, and evaluates that it is unnecessary to continue to maintain the first edge cloud node according to the recommended power-off device list and the service migration list.

[0263] Step 6: The management system sends a notification of unnecessary maintenance to the business deployment / migration module, and sends a business migration notification to the business deployment / migration module.

[0264] Step 7: The service deployment / migration module sends a service migration location calculation request to the edge cloud computing module.

[0265] Step 8: The edge cloud resource calculation module calculates and selects the second edge cloud node that meets the business resource requirements (involving Principle 4).

[0266] Step 9: The edge cloud resource calculation module returns the edge location where the service can be migrated.

[0267] Step 10: The service deployment / migration module migrates the service and deploys it to the second edge cloud node until all services are migrated.

[0268] Step 11: The service deployment / migration module notifies the management system that all service migrations have been completed.

[0269] Step 12: The management system shuts down the first edge cloud node and performs rectification on the first edge cloud node.

[0270] Scenario 2: Ultra-low load operation (the management system decides to continue maintenance)

[0271] Solve the problem of low resource allocation rate, monitor and avoid idle operation of devices in edge cloud nodes.

[0272] Principles involved: Principle 1: Avoid no-load and ultra-low resource allocation rate operation.

[0273] Please refer to Figure 8 , the edge cloud resource collaborative scheduling method of the embodiment of the present invention is as follows:

[0274] Step 1: The proxy edge cloud resource monitoring module in the first edge cloud node detects that the resource allocation rate of the target resource is lower than the lower limit D2 of its resource allocation rate, generates an alarm message and reports it to the client edge cloud resource monitoring module in the cloud management platform.

[0275] Step 2: The client edge cloud resource monitoring module reports shutdown suggestions to the edge cloud computing module.

[0276] Step 3: After calculation, the edge cloud computing module finds that there are devices in the first edge cloud node that are not running services, and generates a list of recommended devices to be powered off. If the first edge cloud node carries fewer services, the feasibility of service migration is calculated.

[0277] Step 4: Send a shutdown suggestion to the management system, along with a list of recommended devices to be powered off and a list of services to be migrated.

[0278] Step 5: The management system powers off the devices according to the recommended power-off device list, and based on the recommended power-off device list and the service migration list, the management system evaluates that it is necessary to continue to maintain the first edge cloud node, so no operation is performed, and the first edge cloud node remains in the scheduled edge cloud pool.

[0279] Scenario 3: The utilization rate of a target resource on the edge cloud node is overloaded

[0280] Resolve resource overload, monitor and avoid overload of devices in edge cloud nodes, and avoid overload of single sites.

[0281] Principles involved: Principle 2, avoid overuse and overloading of resources; Principle 4, screen edge cloud nodes.

[0282] Please refer to Fig. 9 , the edge cloud resource collaborative scheduling method of the embodiment of the present invention is as follows:

[0283] Step 1: The proxy segment edge cloud resource monitoring module in the first edge cloud node detects that the actual resource usage rate of the target resource in the edge cloud node is higher than the upper limit value R1 of its actual resource usage rate, generates an alarm message and reports it.

[0284] Step 2: The client edge cloud resource monitoring module of the management platform triggers calculation and reports relevant resource information to the edge cloud resource computing module;

[0285] Step 3: The edge cloud resource calculation module calculates and selects the services to be migrated of the first edge cloud node. For the selection of services to be migrated, refer to Principle 2.

[0286] Step 4: The edge cloud resource calculation module returns the screened services to be migrated to the service deployment / migration module;

[0287] Step 5: The service deployment / migration module sends a service request for scheduling and screening service resource usage to the client edge cloud resource monitoring module;

[0288] Step 6: The client edge cloud resource monitoring module returns the service resource usage;

[0289] Step 7: The service deployment / migration module sends a request to the edge cloud resource computing module to select edge clouds or edge computing machines;

[0290] Step 8: The edge cloud resource calculation module selects the second edge cloud node that meets the business resource requirements; (involving Principle 4)

[0291] Step 9: The edge cloud resource computing module notifies the service deployment / migration module that the service can be migrated to the edge location;

[0292] Step 10: The service deployment / migration module migrates and deploys the service to the second edge cloud node.

[0293] Scenario 4: Actual resource utilization is lower than R2

[0294] Solve the problem of low-load resource usage, monitor and avoid low-load operation of devices in edge cloud nodes, and migrate or reduce the capacity of related low-load businesses.

[0295] Principles involved: Principle three, avoid low actual resource utilization; Principle four, screen edge cloud nodes.

[0296] Please refer to Fig.10 , the edge cloud resource collaborative scheduling method of the embodiment of the present invention is as follows:

[0297] Step 1: The proxy-side edge cloud resource monitoring module in the first edge cloud node detects that the actual resource usage rate of the target resource in the edge cloud node is lower than its resource usage rate lower limit R2, generates an alarm message and reports it.

[0298] Step 2: The client edge cloud resource monitoring module of the cloud management platform sends a request to the edge cloud computing module to filter out services with unreasonable resource allocation.

[0299] Step 3: The edge cloud resource calculation module calculates and obtains a list of unreasonable services in the first edge cloud node and recommended reduction specifications.

[0300] Step 4: The cloud resource computing module notifies the service deployment / migration module or the third-party deployment module to migrate or reduce the capacity of the service, carrying a list of services to be migrated or reduced and recommended reduction specifications.

[0301] Step 5: The service deployment / migration module shrinks the services in the shrinking service list (the relevant operation process of the third-party deployment module is omitted).

[0302] Step 6: If there is a business that needs to be migrated, the business deployment / migration module queries the edge cloud resource computing module for the edge cloud nodes that can be migrated.

[0303] Step 7: The edge cloud resource calculation module calculates the recommended downsizing specifications for the business.

[0304] Step 8: The edge cloud resource calculation module selects the second edge cloud node that meets the business resource requirements. (Involving Principle 4)

[0305] Step 9: The edge cloud resource calculation module notifies the service deployment / migration module that the service can be migrated to the edge location.

[0306] Step 10: The service deployment / migration module migrates the services and deploys them to the second edge cloud node until all services that need to be migrated are completely migrated.

[0307] Please refer to Fig.11 The embodiment of the present invention further provides a cloud management platform 110, including a processor 111, a memory 112, and a computer program stored in the memory 112 and executable on the processor 111. When the computer program is executed by the processor 111, each process of the embodiment of the edge cloud resource collaborative scheduling method applied to the cloud management platform is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be described here.

[0308] Please refer to Fig.12The embodiment of the present invention further provides an edge cloud node 120, including a processor 121, a memory 122, and a computer program stored in the memory 122 and executable on the processor 121. When the computer program is executed by the processor 121, each process of the embodiment of the edge cloud resource collaborative scheduling method applied to the edge cloud node is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be described here.

[0309] The embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, each process of the above-mentioned edge cloud resource collaborative scheduling method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it is not repeated here. The computer-readable storage medium is, for example, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0310] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0311] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a terminal (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0312] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims, all of which are within the protection of the present invention.

Claims

1. A method for collaborative scheduling of edge cloud resources, applied to a cloud management platform, characterized in that: include: Receiving alarm information of a target resource sent by a first edge cloud node, wherein the alarm information of the target resource is generated based on at least one of the following thresholds: an upper limit value of a resource allocation rate of the target resource, a lower limit value of a resource allocation rate, an upper limit value of an actual resource usage rate, and a lower limit value of an actual resource usage rate; Coordinated scheduling of edge cloud resources based on the alarm information of the target resources; These include: If the alarm information indicates that the resource allocation rate of the target resource is lower than the lower limit of its resource allocation rate, at least one of the following operations is performed on the first edge cloud node: No new services are deployed for the first edge cloud node; Generate a shutdown suggestion for shutting down the first edge cloud node and send it to a management system; Generate a list of devices recommended to be powered off and notify the management system, where the devices recommended to be powered off are devices in the first edge cloud node that are not running any services; Calculate the feasibility of business migration of the first edge cloud node, and if feasible, generate a business migration list and notify the management system; If a service migration notification sent by the management system is received, a second edge cloud node that meets the resource requirements of the service to be migrated is selected, and the service to be migrated is deployed to the second edge cloud node; Notifying the management system that the migration of the service to be migrated has been completed, so that the management system shuts down the first edge cloud node; Coordinating and scheduling edge cloud resources according to the alarm information of the target resource also includes: If the alarm information indicates that the actual resource usage rate of the target resource is lower than the lower limit of the actual resource usage rate, select the unreasonable business of the first edge cloud node, and recommend downsizing, redeployment and / or migration of the unreasonable business; Scale down the unreasonable business that needs to be scaled down, redeploy the unreasonable business on the first edge cloud node, and / or select a second edge cloud node that meets the resource requirements of the unreasonable business to be migrated, and deploy the business to be migrated to the second edge cloud node.

2. The method according to claim 1, characterized in that According to the alarm information of the target resource, the coordinated scheduling of the resources in the edge cloud includes: If the alarm information indicates that the resource allocation rate of the target resource is higher than the upper limit of its resource allocation rate, no new services that meet the requirements of the target resource will be deployed for the first edge cloud node until the resource allocation rate of the target resource of the first edge cloud node is lower than the upper limit of its resource allocation rate.

3. The method according to claim 1, characterized in that According to the alarm information of the target resource, the coordinated scheduling of the resources in the edge cloud includes: If the alarm information indicates that the actual resource usage rate of the target resource is higher than the upper limit of the actual resource usage rate, selecting the service to be migrated of the first edge cloud node; A second edge cloud node that meets the resource requirements of the service to be migrated is selected, and the service to be migrated is deployed to the second edge cloud node.

4. The method according to claim 3, characterized in that The services to be migrated from the first edge cloud node include: Selecting a service that meets a first preset condition from all or part of the services as the service to be migrated; Among them, the first preset condition is: Min(R 实际 ≥|R 总实际 -R 总 *R1|); Among them, R 实际 is the actual resource consumption value of the target resource used by the service, R 总实际 is the actual total resource usage of the first edge cloud, R 总 is the total resource amount of the first edge cloud, and R1 is the upper limit of the actual resource utilization rate of the target resource.

5. The method according to claim 4, characterized in that Also includes: If no service that meets the first preset condition is selected from all or part of the services, the service that occupies the largest amount of the target resources in the first edge cloud node will be used as the service to be migrated until the actual utilization rate of the target resource is lower than the upper limit of its actual utilization rate.

6. The method according to claim 1, characterized in that The unreasonable business of selecting the first edge cloud node includes: Select a business that meets the following criteria: 实际 ≤R 申请 *R2, where R 实际 is the actual resource consumption value of the target resource by the service, R 申请 is the resource value of the target resource applied for by the service in the first edge cloud node, and R2 is the lower limit of the actual resource usage rate of the target resource; The selected services are sorted from high to low according to actual resource usage rates, and at least one service with a high actual resource usage rate is regarded as an unreasonable service.

7. The method according to claim 1 or 3, characterized in that: The second edge cloud node meets a second preset condition, and the second preset condition includes at least one of the following: (R 1实际 +R 2实际 ) / R 2总 ≤R1, where R 1实际 is the resource consumption value of the target resource by the service at the first edge cloud node, R 2实际 is the actual resource consumption value of the target resource on the second edge cloud node, R 2总 is the total amount of resources of the second edge cloud node, and R1 is the upper limit of the actual resource utilization rate of the target resource; The actual resource allocation rate in the second edge cloud node of the target resource is ≤ the upper limit value of the resource allocation rate of the target resource; D 2实际 +S≤D1*D 2可分配 , where D 2实际 is the actual resource allocation value in the second edge cloud node of the target resource, S is the required resource amount of the service to be migrated, D1 is the upper limit of the resource allocation rate of the target resource, and D 2可分配 is the actual allocatable amount of the target resource of the second edge cloud node; The resources other than the target resource meet at least one of the following corresponding thresholds: an upper limit value of a resource allocation rate, a lower limit value of a resource allocation rate, an upper limit value of an actual resource usage rate, and a lower limit value of an actual resource usage rate.

8. The method according to claim 7, characterized in that Also includes: If no edge cloud node that meets the second preset condition is screened out, a backup edge cloud node is enabled as the second edge cloud node, and the backup edge cloud node is an edge cloud node in a shutdown state.

9. The method according to claim 1, characterized in that Also includes: Collecting actual resource usage and / or resource allocation rate of each preset resource of the edge cloud node; Perform alarm analysis based on the collected actual resource usage rate and / or resource allocation rate of each preset resource within the preset time.

10. A method for collaborative scheduling of edge cloud resources, applied to edge cloud nodes, characterized in that: include: Monitor the resource allocation rate and / or actual resource utilization rate of the preset resources in the edge cloud node in real time; Generate alarm information for a target resource in the preset resources according to at least one of an upper limit value of a resource allocation rate, a lower limit value of a resource allocation rate, an upper limit value of an actual resource usage rate, and a lower limit value of an actual resource usage rate of each preset resource; Sending alarm information of the target resource to the cloud management platform; The cloud management platform is used to coordinate the scheduling of edge cloud resources according to the alarm information of the target resources; including: If the alarm information indicates that the resource allocation rate of the target resource is lower than the lower limit of its resource allocation rate, at least one of the following operations is performed on the edge cloud node: No new services are deployed for the edge cloud nodes; Generate a shutdown suggestion for shutting down the edge cloud node and send it to a management system; Generate a list of recommended powered-off devices and notify the management system, wherein the recommended powered-off devices are devices in the edge cloud node that are not running any services; Calculate the feasibility of business migration of the edge cloud node, and if feasible, generate a business migration list and notify the management system; If a service migration notification sent by the management system is received, a second edge cloud node that meets the resource requirements of the service to be migrated is selected, and the service to be migrated is deployed to the second edge cloud node; Notify the management system that the migration of the service to be migrated has been completed, so that the management system shuts down the edge cloud node; Coordinating and scheduling edge cloud resources according to the alarm information of the target resource also includes: If the alarm information indicates that the actual resource usage rate of the target resource is lower than the lower limit of the actual resource usage rate, select the unreasonable business of the edge cloud node, and recommend downsizing, redeployment and / or migration of the unreasonable business; Scale down the unreasonable business that needs to be scaled down, redeploy the unreasonable business on the edge cloud node, and / or select a second edge cloud node that meets the resource requirements of the unreasonable business to be migrated, and deploy the business to be migrated to the second edge cloud node.

11. The method according to claim 10, characterized in that The preset resources include at least one of the following: device resources within the edge cloud node, network resources within the edge cloud, and firewall resources within the edge cloud.

12. The method according to claim 11, characterized in that Generating the alarm information of the target resource according to at least one of the upper limit value of the resource allocation rate, the lower limit value of the resource allocation rate, the upper limit value of the actual resource usage rate, and the lower limit value of the actual resource usage rate of each preset resource includes: If at least one of the following is true, an alarm message is generated for the target resource: The resource allocation rate of the target resource in the preset resources is lower than the lower limit of the resource allocation rate; The resource allocation rate of the target resource in the preset resources is higher than the upper limit of the resource allocation rate; The actual resource usage rate of the target resource in the preset resources is higher than the upper limit value of the actual resource usage rate; The actual resource usage rate of the target resource in the preset resources is lower than the lower limit value of the actual resource usage rate.

13. A cloud management platform, characterized in that: include: The client edge cloud resource monitoring module is used to receive the alarm information of the target resource sent by the first edge cloud node, wherein the alarm information of the target resource is generated based on at least one of the following thresholds: the upper limit value of the resource allocation rate of the target resource, the lower limit value of the resource allocation rate, the upper limit value of the actual resource usage rate, and the lower limit value of the actual resource usage rate; A collaborative scheduling module, used to collaboratively schedule edge cloud resources according to the alarm information of the target resources; According to the alarm information of the target resource, the edge cloud resources are coordinated and scheduled, including: If the alarm information indicates that the resource allocation rate of the target resource is lower than the lower limit of its resource allocation rate, at least one of the following operations is performed on the first edge cloud node: No new services are deployed for the first edge cloud node; Generate a shutdown suggestion for shutting down the first edge cloud node and send it to a management system; Generate a list of devices recommended to be powered off and notify the management system, where the devices recommended to be powered off are devices in the first edge cloud node that are not running any services; Calculate the feasibility of business migration of the first edge cloud node, and if feasible, generate a business migration list and notify the management system; If a service migration notification sent by the management system is received, a second edge cloud node that meets the resource requirements of the service to be migrated is selected, and the service to be migrated is deployed to the second edge cloud node; Notifying the management system that the migration of the service to be migrated has been completed, so that the management system shuts down the first edge cloud node; Coordinating and scheduling edge cloud resources according to the alarm information of the target resource also includes: If the alarm information indicates that the actual resource usage rate of the target resource is lower than the lower limit of the actual resource usage rate, select the unreasonable business of the first edge cloud node, and recommend downsizing, redeployment and / or migration of the unreasonable business; Scale down the unreasonable business that needs to be scaled down, redeploy the unreasonable business on the first edge cloud node, and / or select a second edge cloud node that meets the resource requirements of the unreasonable business to be migrated, and deploy the business to be migrated to the second edge cloud node.

14. An edge cloud node, characterized in that: include: The proxy-side edge cloud resource monitoring module is used to monitor the resource allocation rate and / or actual resource usage rate of the preset resources in the edge cloud node in real time; Generate alarm information of a target resource in the preset resources according to at least one of the upper limit value of the resource allocation rate, the lower limit value of the resource allocation rate, the upper limit value of the actual resource usage rate, and the lower limit value of the actual resource usage rate of each preset resource; and send the alarm information of the target resource to the cloud management platform; The cloud management platform is also used to coordinate the scheduling of edge cloud resources according to the alarm information of the target resources; including: If the alarm information indicates that the resource allocation rate of the target resource is lower than the lower limit of its resource allocation rate, at least one of the following operations is performed on the edge cloud node: No new services are deployed for the edge cloud nodes; Generate a shutdown suggestion for shutting down the edge cloud node and send it to a management system; Generate a list of recommended powered-off devices and notify the management system, wherein the recommended powered-off devices are devices in the edge cloud node that are not running any services; Calculate the feasibility of business migration of the edge cloud node, and if feasible, generate a business migration list and notify the management system; If a service migration notification sent by the management system is received, a second edge cloud node that meets the resource requirements of the service to be migrated is selected, and the service to be migrated is deployed to the second edge cloud node; Notify the management system that the migration of the service to be migrated has been completed, so that the management system shuts down the edge cloud node; Coordinating and scheduling edge cloud resources according to the alarm information of the target resource also includes: If the alarm information indicates that the actual resource usage rate of the target resource is lower than the lower limit of the actual resource usage rate, select the unreasonable business of the edge cloud node, and recommend downsizing, redeployment and / or migration of the unreasonable business; Scale down the unreasonable business that needs to be scaled down, redeploy the unreasonable business on the edge cloud node, and / or select a second edge cloud node that meets the resource requirements of the unreasonable business to be migrated, and deploy the business to be migrated to the second edge cloud node.

15. A cloud management platform, characterized in that: include: A processor, a memory, and a program stored in the memory and executable on the processor, wherein when the program is executed by the processor, the steps of the edge cloud resource collaborative scheduling method as described in any one of claims 1 to 9 are implemented.

16. An edge cloud node, characterized in that: include: A processor, a memory, and a program stored in the memory and executable on the processor, wherein when the program is executed by the processor, the steps of the edge cloud resource collaborative scheduling method as described in any one of claims 10 to 12 are implemented.

17. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the edge cloud resource collaborative scheduling method as described in any one of claims 1 to 9; or, when executed by a processor, implements the steps of the edge cloud resource collaborative scheduling method as described in any one of claims 10 to 12.

Images