WO2020207266A1

WO2020207266A1 - Network system, instance management method, device, and storage medium

Info

Publication number: WO2020207266A1
Application number: PCT/CN2020/081570
Authority: WO
Inventors: 王广芳
Original assignee: 阿里巴巴集团控股有限公司
Priority date: 2019-04-08
Filing date: 2020-03-27
Publication date: 2020-10-15
Also published as: CN111800282B; CN111800282A; CN116170316A

Abstract

Embodiments of the present application provide a network system, an instance management method, a device, and a storage medium. In the embodiments of the application, the concept of edge computing is used to place cloud computing capabilities closer to a terminal at an edge location, thereby providing a network system comprising edge cloud nodes. In the network system, instances capable of providing a cloud computing service to a service demand side are deployed in an edge cloud node, and can provide such services to a service demand side under the control of a central control device, thereby achieving the purpose of providing services to users using resources located in edge cloud nodes. The invention thus realizes "relocating cloud computing processing closer to a terminal by processing in an edge cloud node," and helps reduce response delays and bandwidth costs.

Description

Network system, instance management and control method, equipment and storage medium

This application claims the priority of the Chinese patent application filed on April 8, 2019 with the application number 201910277465.4 and the invention title "network system, instance control method, equipment and storage medium", the entire content of which is incorporated into this application by reference .

Technical field

This application relates to the field of computer technology, and in particular to a network system, instance management and control method, equipment, and storage medium.

Background technique

At present, the concept of cloud computing is based on centralized resource management and control. Even if multiple data centers are used for interconnection, all software and hardware resources are still treated as unified resources for management, scheduling, and sales. With the advent of the era of 5G and the Internet of Things and the gradual increase of cloud computing applications, the terminal side has higher and higher requirements for cloud resources in terms of latency and bandwidth. The centralized cloud network can no longer meet the increasing demand on the terminal side. Cloud resource requirements.

Summary of the invention

Various aspects of the present application provide a network system, instance management and control method, device, and storage medium to reduce service response delay and bandwidth cost.

The embodiment of the present application provides an instance management and control method, including: determining at least one instance deployed in at least one edge cloud node in a network system, the at least one instance can provide cloud computing services for the service demander; The instance is managed, so that the at least one instance provides cloud computing services for the service demander.

An embodiment of the present application also provides a network system, including: a central management and control device, and at least one edge cloud node; at least one instance is deployed in the at least one edge cloud node, and the at least one instance can provide a cloud for service demanders Computing services; the central management and control device is used to manage and control the at least one instance, so that the at least one instance provides cloud computing services for the service demander.

An embodiment of the present application also provides a central management and control device, including: a memory and a processor; the memory is used to store a computer program; when the computer program is executed by the processor, the processor is caused to implement the application The steps in the example management method provided in the embodiment.

The embodiment of the present application also provides a computer-readable storage medium storing a computer program. When the computer program is executed by one or more processors, the one or more processors are caused to implement the Examples of steps in the control method.

In the embodiments of the present application, in combination with the concept of edge computing, the ability of cloud computing is considered to be placed on the edge side closer to the terminal, so a network system including edge cloud nodes is provided. In the network system, the edge cloud node Instances that provide cloud computing services are deployed in the central control equipment. Under the control of the central control equipment, these instances can provide cloud computing services. This achieves the purpose of providing services to users with the help of resources in edge cloud nodes, so that "put cloud computing to a distance "Processing in edge cloud nodes closer to the terminal" has become a reality, which is conducive to reducing service response delay and bandwidth costs.

Description of the drawings

The drawings described here are used to provide a further understanding of the application and constitute a part of the application. The exemplary embodiments and descriptions of the application are used to explain the application and do not constitute an improper limitation of the application. In the attached picture:

Fig. 1a is a schematic structural diagram of a network system provided by an exemplary embodiment of this application;

FIG. 1b is a schematic structural diagram of a central management and control device and an edge management and control device provided by an exemplary embodiment of this application;

FIG. 1c is a schematic structural diagram of another network system provided by an exemplary embodiment of this application;

FIG. 2a is a schematic flowchart of an example management and control method provided by an exemplary embodiment of this application;

FIG. 2b is a schematic flowchart of an example upgrade method provided by an exemplary embodiment of this application;

FIG. 2c is a schematic flowchart of an example migration method provided by an exemplary embodiment of this application;

FIG. 3 is a schematic structural diagram of a central management and control device provided by an exemplary embodiment of this application.

detailed description

In order to make the purpose, technical solutions, and advantages of the present application clearer, the technical solutions of the present application will be described clearly and completely in conjunction with specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative work shall fall within the protection scope of this application.

In view of the technical problem that the existing centralized cloud network can no longer meet the terminal's increasing demand for cloud resources, in some embodiments of this application, in combination with the concept of edge computing, it is considered to place the cloud computing capability on the edge closer to the terminal Therefore, a network system including edge cloud nodes is provided. In this network system, instances that provide cloud computing services are deployed in the edge cloud nodes. Under the control of the central control device, these instances can provide cloud computing services to achieve With the purpose of providing services to users with the help of resources in edge cloud nodes, it becomes a reality to "place cloud computing in edge cloud nodes closer to the terminal for processing", which is conducive to reducing service response delays and reducing bandwidth costs.

The technical solutions provided by the embodiments of the present application will be described in detail below with reference to the accompanying drawings.

Fig. 1a is a schematic structural diagram of a network system provided by an exemplary embodiment of this application. As shown in FIG. 1a, the network system 100 includes: a central management and control device 101 and at least one edge cloud node 102; at least one edge cloud node 102 is connected to the central management and control device 101 in a network.

The network system 100 in this embodiment is a cloud computing platform built on edge infrastructure based on cloud computing technology and edge computing capabilities, and is a cloud platform with computing, network, storage, and security capabilities at the edge.

Corresponding to a central cloud or a traditional cloud computing platform, the network system 100 of this embodiment can be regarded as an edge cloud network system. Edge cloud is a relative concept. Edge cloud refers to a cloud computing platform that is relatively close to the terminal. In other words, it is different from central cloud or traditional cloud computing platform. Central cloud or traditional cloud computing platform can include large-scale resources and centralized locations. Data centers, and edge cloud nodes cover a wider network range, and therefore have the characteristics of being closer to the terminal. The resource scale of a single edge cloud node is small, but the number of edge cloud nodes is large, and multiple edge cloud nodes constitute the original Part of the edge cloud in the embodiment. The terminal in this embodiment refers to the demand side of cloud computing services, for example, it may be a terminal or a user side in the Internet, or a terminal or a user side in the Internet of Things. The edge cloud network is a network based on the infrastructure between the central cloud or the traditional cloud computing system and the terminal. Wherein, the network system 100 includes at least one edge cloud node 102, and each edge cloud node 102 includes a series of edge infrastructures. These edge infrastructures include, but are not limited to: distributed data centers (DC), wireless computer rooms or clusters, operations Communication networks, core network equipment, base stations, edge gateways, home gateways, computing devices and/or storage devices and other edge devices and corresponding network environments. It is explained here that the locations, capabilities, and included infrastructure of different edge cloud nodes 102 may be the same or different.

Among them, the network system 100 of this embodiment is combined with a central cloud or a traditional cloud computing platform and other central networks and terminals to form a "cloud edge-end three-body coordination" network architecture. In this network architecture, the network can be forwarded and stored. Tasks such as computing and/or intelligent data analysis are processed in each edge cloud node 102 in the network system 100. Since each edge cloud node 102 is closer to the terminal, the response delay can be reduced and the central cloud or traditional cloud The pressure on the computing platform reduces bandwidth costs.

How to reasonably schedule the resources of multiple edge cloud nodes and how to manage and control multiple edge cloud nodes to perform cloud computing services with correct and stable logic is an important challenge. In the network system 100 of this embodiment, a central management and control device 101 is deployed. The central management and control device 101 uses the edge cloud node 102 as the management and control object for resource scheduling, image management, instance management and control, operation and maintenance, network, security, etc. At least one edge cloud node 102 in the network system 100 is uniformly managed and controlled, so that cloud computing services are placed in each edge cloud node 102 for processing. In terms of deployment and implementation, the central control device 101 can be deployed in one or more cloud computing data centers, or it can be deployed in one or more traditional data centers, and the central control device 101 can also be connected to at least one edge cloud under its control. The nodes jointly constitute an edge cloud network, which is not limited in this embodiment.

For an edge cloud node 102, various resources may be provided externally, such as computing resources such as CPU and GPU, storage resources such as memory and hard disk, and network resources such as bandwidth. In addition, the edge cloud node 102 can also create a corresponding instance based on the image, and provide various cloud computing services externally through the instance. Among them, the image is the basic file needed to create an instance in the edge cloud node. For example, it can be an image file such as an operating system, application, or operation configuration required to provide users with cloud computing services, and it can be in line with edge cloud node computing deployment Requirements, according to a specific series of documents in a certain format made into documents. In addition, there are various forms of images, which can be virtual machine (VM) image files, container (Docker) image files, or various types of application packaging files, etc. The image form can be compatible with the virtualization technology used by cloud computing services. Regarding, this embodiment does not limit this. Corresponding to the image, the implementation form of the instance can be a virtual machine, container, or application.

In combination with the above, in this embodiment, the central management and control device 101 can perform resource scheduling on at least one edge cloud node 102 according to resource requirements, or can perform image management and distribution for at least one edge cloud node 102 according to image requirements. Of course, It is possible to perform resource scheduling on at least one edge cloud node 102 and provide mirroring for at least one edge cloud node 102 according to cloud computing service requirements. Among them, cloud computing service requirements include resource requirements and mirroring requirements. Optionally, the central management and control device 101 may provide a requirement submission portal to the outside, and the requirement submission portal may be a web page, an application page, or a command window. The role of the requirement submission portal is for the requirement to submit its own requirement description information to the central control device 101.

For the resource demander, the resource demand description information can be submitted to the central management and control device 101 through the above demand submission entry. The resource demand description information includes: edge cloud node selection parameters and resource selection parameters; edge cloud node selection parameters include scheduling domains and/or For the performance requirements of edge cloud nodes, the resource selection parameters include resource type, resource quantity, and performance requirements for resource equipment. The central management and control device 101 may perform resource scheduling on at least one edge cloud node according to the resource requirement description information. Optionally, a resource scheduling method includes: the central management and control device 101 determines the scheduled target edge cloud node and the scheduled target edge cloud node from at least one edge cloud node 102 of the network system 100 according to resource demand description information Resource information; according to the resource information, the corresponding resource device in the target edge cloud node is controlled to allocate or reserve resources.

For the mirroring demand side, the mirroring demand description information can be submitted to the central management and control device 101 through the above demand submission entry. The mirroring demand description information can point to the mirror that needs to be used, which can be the mirror itself, or the name, ID and other identification types of the mirror. The information can also be some function description information of the cloud computing service, which can reflect the required image. The central management and control device 101 can obtain the image according to the description information of the image demand; provide the image to the edge cloud node in the network system 100 that needs the image, so that the edge cloud node creates a corresponding instance based on the image, and the instance provides the corresponding cloud externally Computing services.

For the cloud computing service demander, the service demand description information can be submitted to the central management and control device 101 through the above demand submission portal. The service demand description information includes resource demand description information and mirroring demand description information. For resource requirement description information and mirroring requirement description information, please refer to the previous description, which will not be repeated here. It is worth noting that the resource requirement description information and the mirroring requirement description information in the service requirement description information can be submitted together or separately. The central management and control device 101 can perform resource scheduling on at least one edge cloud node 102 in the network system 100 according to the service demand description information; provide a mirror image of the scheduled resources in the at least one edge cloud node 102 to use the The scheduled resources provide corresponding cloud computing services.

For the detailed process of the foregoing resource scheduling and image management and distribution, refer to the following embodiments, which will not be described in detail here.

In this embodiment, the central management and control device 101 can not only provide a mirror image for at least one edge cloud node 102 for the edge cloud node 102 to create a corresponding instance, but can also manage and control the instances in at least one edge cloud node 102. There may be at least one instance in the at least one edge cloud node 102, that is, one or more instances. The instances in the edge cloud node may be created based on the image provided by the central management and control device 101, or based on other images, or may be migrated from other edge cloud nodes or other systems, which is not limited. At least one instance in the edge cloud node 102 can provide cloud computing services for the service demander, where the service demander can be any device, application, system, or another service that needs to use the cloud computing service provided by the instance in the edge cloud node . Taking the system as an example, the service demander can be but not limited to: online video system, risk management system, customer information management system, data distribution system, etc. The central management and control device 101 can manage and control at least one instance of at least one edge cloud node 102, so that these instances can provide cloud computing services for service demanders.

The central management and control device 101 can perform various management and control on at least one instance, for example, it can include at least one of upgrade, migration, shutdown, restart, and release, but is not limited thereto. The instance upgrade and migration will be described in detail below.

Instance upgrade:

In actual applications, as business requirements change or the mirror version is updated, it is possible to upgrade the mirror or the corresponding instance. Among them, the central management and control equipment 101 performs upgrade management and control of instances mainly including:

The central management and control device 101 determines the instance to be upgraded from at least one instance. The instance to be upgraded can be one or more; it sends an upgrade request to the service demander, so that the service demander can determine the instance to be upgraded based on the business situation of the instance to be upgraded Upgrade strategy. The upgrade request carries the identification information of the instance to be upgraded, such as the ID and name of the instance to be upgraded. It can also be the ID and name of the service corresponding to the instance to be upgraded. It can also be the ID, name and other information of the image corresponding to the instance to be upgraded. . After receiving the upgrade request, the service demander can determine the instance to be upgraded according to the upgrade request, and combine the business conditions on the instance to be upgraded, such as the business request on the instance to be upgraded and the response status of the business request, to determine whether the instance to be upgraded is It is suitable for upgrading, when is suitable for upgrading, what method is used for upgrading, etc., and then an upgrade strategy can be generated for the instance to be upgraded and returned to the central control device 101. The central management and control device 101 receives the upgrade strategy sent by the service demander, and upgrades the instance to be upgraded according to the upgrade strategy.

In an optional embodiment, the service demander can combine the business conditions on the instance to be upgraded, such as the number of business requests that have been received and not yet completed (referred to as inventory business requests), and whether there are any new business requests ( Incremental service request), etc., to determine when the instance to be upgraded can be upgraded, that is, the upgrade strategy can include the upgrade time. If all existing service requests on the instance to be upgraded have been responded, and there are no incremental service requests, in this case, the upgrade service request of the instance to be upgraded will not be interrupted and will not affect the user experience, then it is considered OK Upgrade the instance to be upgraded. For the service demander, when it considers that the instance to be upgraded can be upgraded, it can return an upgrade notification to the central management and control device 101, and the upgrade notification carries the time to instruct the central management and control device 101 to upgrade the instance to be upgraded after receiving the upgrade notification. Information, the way that the upgrade notification carries the time information can be explicit or implicit. For the central management and control device 101, after receiving the upgrade notification, the instance to be upgraded can be upgraded.

Of course, in addition to the above-mentioned method, the service demander can also estimate the appropriate upgrade time in combination with the business conditions on the instance to be upgraded, and send the upgrade time to the central control device 101 in the upgrade notification. After receiving the upgrade notification, the central management and control device 101 obtains the upgrade time therefrom, and starts to upgrade the instance to be upgraded at the upgrade time.

The upgrade strategy may include an upgrade time, which is determined by the service demander in combination with the business situation on the instance to be upgraded. Of course, the upgrade strategy may not include the upgrade time. The upgrade time can be determined by the central management and control device 101 according to factors such as the status of the instance to be upgraded, the load situation of the central management and control device 101, and the like. In addition, the upgrade strategy may include an upgrade method, where the upgrade method refers to the method used to upgrade the instance to be upgraded, which can be determined by the service demander in combination with the business situation on the instance to be upgraded. Depending on the image type, the upgrade method is also different. If the upgrade strategy includes the upgrade time, the central control device 101 can start to upgrade the instance to be upgraded at the upgrade time specified in the upgrade strategy; if the upgrade strategy includes the upgrade method, the central control device 101 can use the upgrade method specified in the upgrade strategy to treat The upgrade instance is upgraded; if the upgrade strategy includes an upgrade time and an upgrade method, the central management and control device 101 can adopt the upgrade method specified in the upgrade strategy, and start to upgrade the instance to be upgraded at the upgrade time specified in the upgrade strategy.

Optionally, the upgrade of the instance can be initiated by the central control device 101. For example, the central management and control device 101 can monitor the version information of the mirror corresponding to each instance, and when a new version of the mirror is found, it can determine that the instance corresponding to the new version of the mirror needs to be upgraded; or, it can also monitor the running status of each instance , Life cycle and other information. When problems such as loopholes, instability, insufficiency, excessive consumption of CPU or memory resources are found in the running process of the instance, it can be determined that the instance with these problems needs to be upgraded.

Optionally, upgrading the instance can also be initiated by the service demander. For example, when an instance needs to be upgraded according to business requirements, the service demander can send upgrade description information to the central management and control device 101. The upgrade description information includes instance filter conditions. Based on the instance filter conditions, the instances to be upgraded can be filtered out Instance. The instance filter condition can be the identification information of the instance to be upgraded, such as the ID and name of the instance to be upgraded, or the ID and name of the image corresponding to the instance to be upgraded, or the ID and name of the service corresponding to the instance to be upgraded. Determine the instance to be upgraded. Or, if it is necessary to upgrade all the instances, the instance filter condition may also be identifying information indicating to upgrade all the instances, such as “all”, “1”, etc. The identifying information can be flexibly set. For the central control device 101, the upgrade description information sent by the service demander can be received, the instance filter condition is obtained from the upgrade description information, and the instance to be upgraded is determined from at least one instance according to the instance filter condition; The party sends an upgrade request to request the service demander to determine an upgrade strategy for the upgraded instance based on the business situation on the instance to be upgraded; after the service demander returns the upgrade strategy of the instance to be upgraded, the instance to be upgraded can be upgraded according to the upgrade strategy.

Among them, upgrading the instance to be upgraded mainly refers to: shutting down the instance to be upgraded, updating the instance to be upgraded according to the mirror image of the corresponding version (generally, the new version), and restarting the instance after the update. The image version required for the upgrade of the instance to be upgraded can be determined by the central management and control device 101, for example, the latest version of the corresponding image is used as the image version required for the upgrade, or it can be specified by the service demander. Optionally, the service demander may carry the image version required for the upgrade in the upgrade description information and provide it to the central management and control device 101. For example, the upgrade description information may include "all or specified instances from mirror version A to mirror version B). Upgrade" and other information. Based on this, the central management and control device 101 can obtain the image version required for the upgrade from the upgrade description information, and then, according to the upgrade strategy, use the image corresponding to the image version to upgrade the instance to be upgraded. When all the instances to be upgraded are upgraded, the instance upgrade process ends.

Instance migration:

In practical applications, the instance needs to be migrated in some cases. For example, in the case that the entire edge cloud node is faulty or unavailable, the instances in the edge cloud node need to be migrated to other edge cloud nodes. For another example, in the case of a failure or downtime of a physical machine hosting an instance, the instance on the physical machine needs to be migrated to another physical machine. For another example, it may be necessary to migrate one or some instances from one edge cloud node to other edge cloud nodes due to business needs. For another example, when resources need to be merged, one or some instances need to be migrated. Under the management and control of the central control device 101, instances in the edge cloud nodes can be migrated, and the migration process mainly includes:

The central management and control device 101 determines the instance to be migrated from at least one instance. There may be one or more instances to be migrated; if there are multiple instances to be migrated, the multiple instances to be migrated can be deployed in the same edge cloud node or in different edge cloud nodes.

In some application scenarios, the central management and control device 101 may monitor the state of at least one instance deployed in at least one edge cloud node 102, and obtain the failed instance and/or the instance in which a specified event occurs during operation according to the state of the at least one instance. The instance to be migrated, and then the instance to be migrated is migrated. Among them, a failed instance refers to an instance that cannot run normally, for example, it can be an instance on a physical machine that is down, or an instance that itself is down. Such instances need to be migrated in order to continue to serve the demand side. Provide cloud computing services. The designated event mainly refers to some events that the instance can still run normally after occurrence, which can be flexibly set according to application requirements, and there is no restriction on this. For example, the specified event can be some early warning or alarm events, etc. Although some early warning or alarm events occur, the instance does not produce actual problems and can still run (that is, no failure), but there are hidden dangers of failure, which can be carried out in time before failure Migration to avoid problems such as service interruption caused by faults. In addition, the central management and control device 101 maintains the information of each edge cloud node and the information of each instance deployed in each edge cloud node. Based on this, the edge cloud node to which the instance to be migrated belongs can be determined. For ease of description and distinction, the The edge cloud node to which the instance belongs before the migration is recorded as the first edge cloud node.

In other application scenarios, as time accumulates, there will be some resource fragments in the edge cloud node, or an instance with a larger resource specification needs to be deployed, but the resource equipment in the edge cloud node may not meet the resource specification requirements In these cases, the resources in the edge cloud nodes can be merged through instance migration, which can make full use of resource fragments, and then produce larger resource blocks, which is beneficial to improve resource utilization. Based on this, the central management and control device 101 can determine the instance to be migrated from at least one instance according to resource merging requirements, and then migrate the instance to be migrated. Among them, resource merging is mainly the process of integrating resource fragments through instance migration. After integration, the resource fragments in edge cloud nodes will be reduced or even nonexistent, which is conducive to improving resource utilization in edge cloud nodes. It is worth noting that resource merging requirements can be system-level or node-level. System-level resource merging refers to the integration of resource fragments in the entire network system through instance migration; node-level resource merging refers to the consideration from the dimension of edge cloud nodes, through instance migration to edge cloud The resource fragments in the node are integrated.

Optionally, the resource consolidation requirement may be provided by the service demander. For example, when a service demander needs to deploy a new instance, if the available resources on each resource device in the edge cloud node it serves are not enough to carry the new instance, it can request the migration of the instance in the edge cloud node. Resource integration, so as to provide sufficient resources for new instances. Alternatively, the resource merging requirement may also be a regular behavior of the resource scheduling module of the central management and control device 101. For example, the resource scheduling module of the central management and control device 101 periodically performs resource fragmentation checks. When it is found that the fragmentation rate reaches a certain threshold and instance migration can be performed, it integrates the resource fragments in each edge cloud node to improve the resources in the edge cloud node. Utilization rate.

Among them, the resource consolidation requirements contain information related to resource consolidation. For example, the resource merging requirements may include information about instances that need to be migrated to achieve the purpose of resource merging. Based on this, the instances to be migrated can be directly determined according to the resource merging requirements. For another example, the resource merging requirements may include information about edge cloud nodes that need to be merged. Based on this, the edge cloud node that needs to be merged can be determined according to the resource merging requirements. In this embodiment, the edge cloud node that needs to be merged is called the first edge cloud node; in turn, the resources in the first edge cloud node can be combined The remaining available resources on the device and the resources required by each instance in the first edge cloud node determine the instance to be migrated.

Regardless of the application scenario, after determining the instance to be migrated, the central management and control device 101 can determine whether the first edge cloud node to which the instance to be migrated belongs meets the intra-node migration condition; if the first edge cloud node meets the intra-node migration condition, then The instance to be migrated is migrated within the edge cloud node; if the first edge cloud node does not meet the intra-node migration condition, the instance to be migrated is migrated across edge cloud nodes.

Optionally, the central management and control device 101 may determine whether the first edge cloud node is currently available; if the first edge cloud node is currently available, determine whether the available resources of the first edge cloud node are sufficient to carry the instance to be migrated; The available resources of an edge cloud node are sufficient to carry the instance to be migrated, and it is determined that the first edge cloud node meets the migration conditions within the node; if the first edge cloud node is currently in an unavailable state, or the available resources of the first edge cloud node are insufficient to carry the instance In the migration instance, it is determined that the first edge cloud node does not meet the intra-node migration condition. In the embodiments of the present application, the migration of instances is divided into two types: intra-node migration and cross-node migration. Among them, the available resources of the first edge cloud node mainly refer to the available resources on each resource device in the first edge cloud node; accordingly, judging whether the available resources of the first edge cloud node are sufficient to carry the instance to be migrated mainly refers to judging Whether there is a resource device with sufficient resources available in the first edge cloud node to carry the instance to be migrated.

It is worth noting that in the resource merging scenario, the instance migration for resource merging is mainly intra-node migration, and of course, it can also be cross-node migration. Optionally, in the process of determining the instance to be migrated according to the remaining available resources on each resource device in the first edge cloud node and the resources required by each instance in the first edge cloud node, the instance to be migrated may also be determined A resource device, where the resource device is a resource device whose remaining available resources in the first edge cloud node can carry the instance to be migrated. Of course, if there are no remaining available resources in the first edge cloud node that can carry resource devices of the instance to be migrated, cross-node migration can be performed for the instance to be migrated. In view of the purpose of resource merging, in the process of cross-node migration for the instances to be migrated, priority is given to migrating the instances to be migrated to other edge cloud nodes that have been used and the remaining available resources can carry the resource devices of the instances to be migrated; Further, in the case that there are multiple resources that have been used and remaining available resources can carry the resource equipment of the instance to be migrated, the principle of minimum resource fragmentation can be used to select the matching degree between the remaining available resources and the resources required by the instance to be migrated. Higher resource equipment, try to produce less resource fragments or no resource fragments.

For intra-node migration: Optionally, the continuity of the cloud computing service provided by the instance can be ensured through the hot migration technology. For the hot migration technology, please refer to the prior art, which will not be repeated here.

For cross-node migration: the central management and control device 101 can select a second edge cloud node from at least one edge cloud node, the second edge cloud node is different from the first edge cloud node, and the available resources in the second edge cloud node are sufficient to carry the migration Instance, that is, enough resources; migrate the instance to be migrated to the second edge cloud node, and send the attribute information of the instance to be migrated in the second edge cloud node to the service demander, so that the service demander can base on the attribute information Perform business scheduling for the instances to be migrated. Among them, the attribute information of the instance to be migrated in the second edge cloud node means that after the instance to be migrated is migrated to the second edge cloud node, an external (for example, a service demander or a third party authorized by the service demander) conducts an operation on the instance to be migrated. Information required for service scheduling may include, but is not limited to, for example, information such as the area where the second edge cloud node is located, operator information, and/or public network IP. Taking the service demander as an example, it is possible to determine whether to request the service according to the area and operator information of the second edge cloud node in the above attribute information, combined with the operator information and area of the terminal that initiated the service request. Assigned to the instance to be migrated in the second edge cloud node; if it is determined to assign the service request to the instance to be migrated in the second edge cloud node, the public network IP in the above attribute information can be provided to The terminal, the terminal's request can access the instance to be migrated in the second edge cloud node, achieving the purpose of scheduling the service request of the terminal to the instance to be migrated in the second edge cloud node.

Optionally, when selecting the second edge cloud node, the following methods can be used but not limited to:

Method 1: According to the distance between other edge cloud nodes and the first edge cloud node, select the edge cloud node whose distance from the first edge cloud node is less than the set distance threshold, or select the closest distance to the first edge cloud node An edge cloud node, or an edge cloud node arbitrarily selected from the N edge cloud nodes closest to the first edge cloud node as the second edge cloud node. In Manner 1, the second edge cloud node is closest or relatively close to the first edge cloud node, which can save data transmission time and help improve migration efficiency. Optionally, the distance between other edge cloud nodes and the first edge cloud node may be the average distance between other edge cloud nodes and the first edge cloud node, or the distance between other edge cloud nodes and the first edge cloud node. The distance between the centers may also be the distance between other edge cloud nodes and the closest outer edge of the first edge cloud node, etc., which can be adaptively defined according to requirements.

Method 2: You can select edge cloud nodes with relatively sufficient bandwidth resources according to the bandwidth resources of other edge cloud nodes. For example, select the edge cloud node with the largest bandwidth resource, or select the bandwidth resource greater than the set bandwidth threshold, or select the bandwidth utilization rate lower The edge cloud node serves as the second edge cloud node. In method 2, the bandwidth resources of the second edge cloud node are sufficient, which can increase the data transmission rate, which is beneficial to improve the migration efficiency.

Method 3: According to the current load situation of other edge cloud nodes, select the edge cloud node with relatively light load, for example, select the edge cloud node with the smallest load, or select the edge cloud node with the load less than the set load threshold as the second Edge cloud node. In mode 3, the load of the second edge cloud node is lighter, it has sufficient resources and can handle instance migration in time, which is beneficial to improve migration efficiency.

Optionally, when migrating the instance to be migrated to the second edge cloud node, the central management and control device 101 may reserve or allocate resources for the instance to be migrated in the second edge cloud node according to the resource requirements of the instance to be migrated; After the resource reservation or allocation is successful, the instance to be migrated is migrated to the resources reserved or allocated in the second edge cloud node. For example, the resource requirements of the instances to be migrated can be combined to determine the type of resources, the amount of resources and/or the performance requirements of the resource equipment required by the instances to be migrated, and resource reservation or allocation can be performed in the second edge cloud node based on this information , Which can provide resource guarantee for successful instance migration. Regarding the process of the central management and control device 101 in the second edge cloud node to reserve or allocate resources for the instances to be migrated, please refer to the content of the subsequent resource scheduling part, which will not be repeated here.

Optionally, if the instance to be migrated is a failed instance, that is, an instance that is not functioning normally, the central management and control device 101 may also notify the service demander of the migration event, so that the service demander can make appropriate response actions, such as Update the information of the instance in the service demander, or make a disaster recovery response to the downtime during the instance migration. Further, in the process of notifying the migration event, the attribute information of the instance to be migrated in the second edge cloud node may be provided to the service demander. Of course, after the instance to be migrated is successfully migrated to the second edge cloud node, the attribute information of the instance to be migrated in the second edge cloud node may be provided to the service demander.

Optionally, if the instance to be migrated is an instance in which a specified event occurs during operation, that is, an instance that can still run normally despite the occurrence of a specified event, the central control device 101 may also send a migration request to the service demander for the service demander Determine a migration strategy for the instance to be migrated in combination with the business situation on the instance to be migrated; receive the migration strategy sent by the service demander, and migrate the instance to be migrated to the second edge cloud node according to the migration strategy. The migration strategy mainly includes at least one information of whether to migrate, migration time, and migration mode. Optionally, the service demander can determine when to perform the migration based on the number of stock service requests and incremental service requests on the instance to be migrated and the response status. For example, all stock service requests on the instance to be migrated can be responded to. And if there are not many incremental business requests, determine the instance migration.

Further optionally, the central management and control device 101 may send the attribute information of the instance to be migrated in the second edge cloud node together with the migration request to the service demander. Alternatively, after the instance to be migrated is successfully migrated to the second edge cloud node, the attribute information of the instance to be migrated in the second edge cloud node may be provided to the service demander.

Further optionally, if the instance to be migrated is an instance that has a specified event but can still run normally, during the migration process, the instance to be migrated can continue to run on the first edge cloud node, so that the business during the migration process can continue to be scheduled To the instance to be migrated in the first edge cloud node, ensure business continuity. After the instance to be migrated is successfully migrated to the second edge cloud node, and the service demander is ensuring that all new business requests are scheduled to the second edge cloud node, and the business requests in the first edge cloud node are gradually reduced, there is no new The central management and control device can release the instance to be migrated in the first edge cloud node when there are no more business requests on the instance to be migrated running on the first edge cloud node. Optionally, after determining that the service requester no longer has any service requests on the instance to be migrated running in the first edge cloud node, and there is neither an inventory service request nor an incremental service request, it can send a release to the central control device 101 Notification; the central management and control device 101 receives the release notification sent by the service demander, and releases the instance to be migrated running in the first edge cloud node according to the release notification. Further, the central management and control device 101 may also synchronize the running state of the instance to be migrated running in the first edge cloud node to the instance to be migrated in the second edge cloud node.

Furthermore, regardless of the instance to be migrated, migrating the instance to be migrated to the second edge cloud node is mainly to control the corresponding resource equipment in the second edge cloud node to reserve or The process of creating an instance to be migrated on the allocated resources. Optionally, the central management and control device 101 may provide the corresponding resource device in the second edge cloud node with the application image or instance snapshot corresponding to the instance to be migrated, so that the corresponding resource device in the second edge cloud node can perform the pre-processing based on the application image or instance snapshot. Create an instance to be migrated on the reserved or allocated resources, but it is not limited to this.

Optionally, the central management and control device 101 of this embodiment may encapsulate its own instance upgrade, instance migration, and other management and control functions into a series of application programming interfaces (Application Programming Interface, API) and open them to service demanders. These open APIs are called Open APIs (OpenAPI), and the central management and control device 101 can interact with the service demander through OpenAPI.

It is worth noting that in the network system 100, the central management and control device 101 can directly control and schedule at least one edge cloud node 102, but it is not limited to this. As shown in 1b, in the network system 100, in addition to a central management and control device 101 and at least one edge cloud node 102, an edge management and control device 103 is also included. Among them, the number of edge management and control devices 103 may be one or multiple. In addition, the edge management and control device 103 may be deployed in one or more edge cloud nodes 102. In an optional embodiment, as shown in FIG. 1b, an edge management and control device 103 is separately deployed in each edge cloud node 102. Further, each edge cloud node includes one or more resource devices. Optionally, the edge management and control device 103 may be deployed on one resource device in a centralized manner, or may be deployed on multiple resource devices in a distributed manner. In addition, each edge cloud node can include one or more proprietary devices in addition to resource devices. The edge management and control device 103 can be deployed on one dedicated device or distributed on multiple dedicated devices. On the device. Among them, the proprietary device refers to the physical device used to deploy the edge management and control device 103, which is different from the resource device. In addition, the edge management and control device 103 can also be deployed with the central management and control device 101, which is not limited here. In addition, the central management and control device 101 may be deployed in one or more cloud computing data centers or traditional data centers, and may also be deployed in an edge cloud network together with at least one edge cloud node.

It is explained here that the central management and control device of this embodiment can be a logical device with the capabilities of resource scheduling and image management. These functions can be implemented on one physical machine or virtual machine, or distributed in multiple devices. On a physical machine or a virtual machine. Of course, the central management and control device in this embodiment may also be one or more physical devices with capabilities such as resource scheduling and image management. The embodiment of the present application does not limit the implementation structure of the central management and control device 101, and any device structure with the foregoing capabilities is applicable to the embodiment of the present application.

Similar to the central management and control device 101, the edge management and control device 103 can also be a logical device, which has the ability to deploy a physical machine (for example, a resource device or a proprietary device in an edge cloud node) or a virtual machine. It can also be deployed on multiple physical machines (such as resource devices or proprietary devices in edge cloud nodes) or virtual machines in a decentralized manner. Of course, the edge control device can also be one or more physical devices with corresponding capabilities. The embodiments of the present application do not limit the implementation structure of the edge management and control device 103, and any device structure with corresponding capabilities is applicable to the embodiments of the present application.

In this embodiment, the edge management and control device 103 can assist and cooperate with the central management and control device 101 to manage and control at least one edge cloud node 102. With the assistance of the edge management and control device 103, the central management and control device 101 can manage and schedule at least one edge cloud node 102 more conveniently and efficiently, thereby achieving the purpose of making full use of edge resources.

Among them, the central management and control device 101 and the edge management and control device 103 can establish a secure and encrypted communication channel, and interact based on the communication channel. The communication channel includes a control interface and a data interface, and the central management and control device 101 interacts with the edge management and control device 103 on the control plane and the data plane based on the control interface and the data interface to complete the scheduling and management of the edge cloud node 102. The data interface is used for data transmission between the central management and control device 101 and the edge management and control device 103. The control interface has but not limited to the following functions:

1. Resource scheduling capability: The central control device 101 can perform resource scheduling on edge cloud nodes from multiple dimensions through a control interface with resource scheduling capabilities (can be referred to as resource scheduling interface for short). The edge cloud node is the central control device 101 for resource scheduling Object;

2. Image management and distribution capabilities: The central management and control device 101 can provide images to edge cloud nodes through a control interface with image management and distribution capabilities (referred to as image management interfaces), so that the edge cloud nodes can create images based on the received images Corresponding examples, providing corresponding cloud computing services through examples;

3. Operation and maintenance management capability: The central control device 101 performs operation and maintenance management on edge cloud nodes through a control interface with operation and maintenance management capabilities (referred to as the operation and maintenance management interface). The operation and maintenance management includes but is not limited to: control edge cloud nodes Application, virtualization software, etc., monitor the status, resource usage and infrastructure of the instance.

Corresponding to the capabilities of the aforementioned control interface, the central management and control device 101 of this embodiment has but not limited to the following functions:

1. Information can be described according to service requirements, such as the specifications of cloud computing services, the areas where cloud computing services need to be deployed, the distribution of operator networks, network delays, load conditions, bandwidth costs, required resource types and/or resource equipment Performance requirements, etc., to schedule edge cloud nodes;

2. The image required for cloud computing services can be obtained, and the image can be provided to the corresponding resource equipment in the edge cloud node for configuration and installation, so that the corresponding resource equipment can create corresponding instances to provide cloud computing services;

3. Operation and maintenance management and control of edge cloud nodes can be performed, including but not limited to: management and control of applications, virtualized components, instance status, resource usage and/or infrastructure conditions in edge cloud nodes, to achieve remote operation and maintenance, logs Management etc.

In addition to the above functions, the central control equipment can also have other functions, such as security assurance functions, involving the security of the central control equipment, the link security between the central control equipment and the edge control equipment, and the edge cloud nodes. Security of cloud nodes; responsible for maintaining networking information in the network system.

In the network system 100, at least one edge cloud node 102 can form a resource pool, and each edge cloud node 102 serves as a scheduling object, and provides various resources or cloud computing services externally under the scheduling of the central management and control device 101. Among them, the central management and control device 101 and the edge management and control device 102 cooperate with each other to perform resource scheduling on at least one edge cloud node 102, and can also perform mirror management and distribution for at least one edge cloud node 102. Of course, it can also perform resource scheduling on at least one edge cloud node 102. The edge cloud node 102 performs resource scheduling and provides a mirror image for at least one edge cloud node 102. In addition to resource scheduling and image management and distribution for the edge cloud node 102, the management and control of the instances in the edge cloud node 102 is also a problem that the network system 100 needs to solve. Successfully solving this problem is also "putting cloud computing in the distance The basis of processing in the edge cloud node closer to the terminal. To this end, the central management and control device 101 and the edge management and control device 103 cooperate with each other, and can also manage and control instances in at least one edge cloud node 102, such as at least one of upgrade, migration, shutdown, restart, and release.

Optionally, in the process of upgrading the instance to be upgraded, the edge management and control device 103 may assist the central management and control device 101 to upgrade the instance to be upgraded by using the mirror corresponding to the mirror version according to the upgrade strategy. For example, the central management and control device 101 may send the upgrade policy and the image corresponding to the image version to the edge management and control device 103, and the edge management and control device 103 uses the image corresponding to the image version to upgrade the instance to be upgraded according to the upgrade policy. Further, if an edge management and control device 103 is deployed in each edge cloud node, the central management and control device 101 can send the upgrade strategy and the image corresponding to the mirror version to the edge management and control device 103 in the edge cloud node to which the instance to be upgraded belongs, and the waiting The edge management and control device 103 in the edge cloud node to which the upgraded instance belongs uses the image corresponding to the image version to upgrade the instance to be upgraded according to the upgrade strategy. For the edge management and control device 103, the upgrade method indicated by the upgrade strategy can be used. At the upgrade time indicated by the upgrade strategy, the mirror corresponding to the mirror version is provided to the resource device where the instance to be upgraded is located, and the resource device uses the mirror to upgrade the instance. Upgrade. For other descriptions of upgrading the instance to be upgraded, reference may be made to the description in the foregoing embodiment, which is not repeated here.

Optionally, during the migration of the instance to be migrated, the edge management and control device 103 may assist the central management and control device 101 to control the corresponding resource device in the second edge cloud node to reserve or allocate resources for the instance to be migrated. Among them, the central management and control device 101 may determine the scheduled resource information in the second edge cloud node according to the resource requirements of the instance to be migrated, and provide the resource information to the edge management and control device 103, and the edge management and control device 103 controls the resource according to the resource information. The corresponding resource device in the second edge cloud node reserves or allocates resources for the instance to be migrated. Further, if the edge management and control device 103 is deployed in the second edge cloud node, the central management and control device 101 can provide resource information to the edge management and control device 103 in the second edge cloud node, and the edge management and control device in the second edge cloud node 103, according to the resource information, controls the corresponding resource device in the second edge cloud node to reserve or allocate resources for the instance to be migrated.

In addition, in the process of migrating the instance to be migrated, the edge management and control device 103 may also assist the central management and control device 101 in migrating the instance to be migrated to the resources reserved or allocated by the corresponding resource device in the second edge cloud node. The central management and control device 101 may send a migration instruction to the edge management and control device 103. The migration instruction instructs the edge management and control device 103 to obtain the image or instance snapshot corresponding to the instance to be migrated and provide it to the corresponding resource device in the second edge cloud node for the second edge cloud The corresponding resource device in the node creates an instance to be migrated on the reserved or allocated resources according to the image or instance snapshot. Further, if the edge management and control device 103 is deployed in the second edge cloud node, the central management and control device 101 may send a migration instruction to the edge management and control device 103 in the second edge cloud node to instruct the edge management and control device 103 in the second edge cloud node The image or snapshot corresponding to the instance to be migrated is obtained and provided to the corresponding resource device in the second edge cloud node for the corresponding resource device in the second edge cloud node to create the instance to be migrated on the reserved or allocated resources according to the image or snapshot.

Optionally, if a snapshot is used in the instance migration process, the way in which the edge management and control device 103 obtains the snapshot will be different according to the storage mode of the snapshot. If the snapshot is stored in the first edge cloud node, it depends on whether the state of the first edge cloud node is available. If the first edge cloud node is unavailable, it is not suitable to use the snapshot for instance migration, and you need to use mirroring instead. Instance migration; if the first edge cloud node is in an available state, the edge management and control device can obtain a snapshot from the first edge cloud node. If the snapshots are scattered and stored in other edge cloud nodes, the process of obtaining the snapshot has nothing to do with the state of the first edge cloud node, and the edge management and control device can obtain the snapshot from other edge cloud nodes when the other edge cloud nodes are available. . After obtaining the snapshot, the edge management and control device provides the snapshot copy to the corresponding resource device in the second edge cloud node for the corresponding resource device to create an instance to be migrated through the snapshot. Among them, creating an instance through a snapshot can restore the data saved when the snapshot is taken.

Optionally, if a mirror is used in the instance migration process, the edge management and control device 103 may first determine whether the mirror corresponding to the instance to be migrated is stored in the second edge cloud node when acquiring the mirror. If the second edge cloud node has a corresponding image, the edge management and control device can directly provide the corresponding image in the second edge cloud node to the corresponding resource device in the second edge cloud node for the corresponding resource device to create an instance to be migrated through the image. If the second edge cloud node does not have a corresponding image, the edge management and control device can request the corresponding image from the central management and control device; the central management and control device can obtain the image from the mirror library and provide it to the edge management and control device, or instruct the edge management and control device to store the corresponding image from another The image is obtained at the edge cloud node; the edge management and control device provides the corresponding resource device in the second edge cloud node after obtaining the corresponding image, so that the corresponding resource device can create an instance to be migrated through the image. The process of the central management and control device instructing the edge management and control device to obtain images from other edge cloud nodes that store corresponding images may refer to the description in the subsequent image management and distribution related embodiments, which will not be repeated here.

In the following embodiments of the present application, various other functions implemented by the central control device or the cooperation of the central control device and the edge control device will be described.

Resource scheduling function:

The central management and control device can perform resource scheduling on at least one edge cloud, which mainly refers to determining the target edge cloud node and target edge cloud node that can be scheduled from at least one edge cloud node 102 in the network system 100 according to service demand description information Scheduled resource information; the resource information is sent to the edge management and control device 103 for the edge management and control device 103 to control the corresponding resource device in the target edge cloud node for resource allocation or reservation. Optionally, the number of target edge cloud nodes can be specified by the user, or can be independently determined by the resource center management and control device according to the service requirement description information, and it can be one or more. The service demand description information can be directly submitted by the service demander, or it can be extracted or calculated from the service-related information submitted by the service demander. The service demander can be a user, an application, a physical machine, or another service that requires a certain service.

The resource scheduling function described here mainly includes the selection of edge cloud nodes and the resource scheduling within the edge cloud nodes, but it is not limited to these two aspects. Among them, the internal resource scheduling of the edge cloud node is specifically embodied as the operation of determining the scheduled resource information in the target edge cloud node and providing resource information. The main purpose is to allocate cloud computing services to the final at the granularity of each edge cloud node. Basic resources, such as server and other resource equipment. Among them, the central control equipment can maintain the information of the resources contained in each edge cloud node as the basis for resource scheduling.

Optionally, the service requirement description information includes edge cloud node selection parameters and resource selection parameters. The edge cloud node selection parameter refers to the parameter required to select the target edge cloud node; the resource selection parameter refers to the information required to select the scheduled resource in the edge cloud node. Based on this, the central management and control equipment can parse out the edge cloud node selection parameters and resource selection parameters from the service demand description information; determine the scheduled target edge cloud node from at least one edge cloud node according to the edge cloud node selection parameters, and according to the resource The selection parameters determine the scheduled resource information in the target edge cloud node.

For example, the service requirement description information may include the scheduling domain and/or the QoS requirements of the cloud computing service, and these parameters may be used as edge cloud node selection parameters. Among them, the scheduling domain points to the area where cloud computing services need to be deployed, which determines the geographic location of edge cloud nodes that should be scheduled. The QoS requirements of cloud computing services may include the requirements of cloud computing services on network delay, load conditions, and/or bandwidth costs. Based on this, the central management and control device can select the edge cloud node that can meet the scheduling domain and/or QoS requirements as the target according to the QoS requirements of the scheduling domain and/or cloud computing service, combined with the geographic location and resource remaining amount of at least one edge cloud node Edge cloud node.

For example, the central management and control device may select the edge cloud node pointed to by the scheduling domain as the target edge cloud node in combination with the geographic location of at least one edge cloud node according to the scheduling domain. Alternatively, the central management and control device can also select the edge cloud node that meets the network delay, load and/or bandwidth cost requirements based on the QoS requirements of the cloud computing service, such as network delay, load conditions, and/or bandwidth cost requirements. The edge cloud node serves as the target edge cloud node. Of course, the central management and control equipment can also select the edge cloud node that can meet the scheduling domain and QoS requirements as the target edge cloud based on the QoS requirements of the scheduling domain and cloud computing services at the same time, combined with the geographic location and remaining amount of resources of at least one edge cloud node. node.

In addition to the information about the scheduling domain and/or the QoS requirements of the cloud computing service, the service requirement description information can also include the resource type, the number of resources, and/or the performance of the resource equipment required by the cloud computing service. These parameters can be As a resource selection parameter. Based on this, after determining the target edge cloud node, the central management and control device can determine the scheduled resource information in the target edge cloud node according to the resource selection parameters. The resource information here may include: resource type, resource quantity, and/or performance requirements for resource devices, so that the edge management and control device can control the corresponding resource device in the target edge cloud node to allocate or reserve resources accordingly. For example, resource types may include, but are not limited to: computing resources such as CPU and GPU, storage resources such as memory and hard disk, and resource types such as bandwidth resources. Taking CPU resources as an example, the number of resources can be 12 CPUs, 24 CPUs, etc., taking memory resources as an example, the number of resources can be 16G memory, 32G memory, etc.; taking bandwidth resources as an example, the number of resources can be 1M bandwidth, 10M Bandwidth etc.

Optionally, the central management and control device can also have the function of computing power orchestration. The computing power orchestration is oriented to relatively complex application scenarios. Multiple cloud computing services are bound together as the smallest resource requirement unit. In this way, in the resource scheduling In the process, multiple cloud computing services can be bound together as a whole, and one or several edge cloud nodes can be selected for them, and the same or several edge cloud nodes can provide resources for them together. Computing power scheduling improves the diversity of resource scheduling and increases the flexibility of resource scheduling, but it does not affect the overall process of resource scheduling.

Image management and distribution functions:

The image management function of the central control device mainly refers to the management of images and the provision of required images for edge cloud nodes. In this way, the edge cloud node can create an instance on the corresponding resource device according to the image, and then the created instance can provide users with required cloud computing services.

In practical applications, there are various scenarios where mirroring needs to be provided for edge cloud nodes. For example, in a case where a user (such as a service demander) submits service demand description information, the central management and control device can provide a corresponding image for the scheduled target edge cloud node. For another example, when there are instances on edge cloud nodes that provide users with cloud computing services, when users need to expand their business, they can submit their expansion requirements to the central control equipment. In order to achieve the purpose of expansion, they need to provide cloud computing services for users. The edge cloud node of the computing service provides a corresponding image, so that the edge cloud node creates a new instance based on the image, so as to achieve the purpose of capacity expansion. For ease of description and distinction, in the following description, the edge cloud node that needs to be mirrored is recorded as the third edge cloud node. The third edge cloud node can be any edge cloud node in the network system, depending on the application scenario. Depends. The following takes the central control device to provide a mirror image for the third edge cloud node as an example to describe the image management function of the central control device.

When it is necessary to provide a mirror image for the third edge cloud node, the central control device can first determine the target image that needs to be provided to the third edge cloud node; then, provide the target image for the third edge cloud node for use by the third edge cloud node The target image provides cloud computing services.

In the network system 100 of this embodiment, a mirror library is maintained, and the mirror library is used to store images in the system. Users can choose to use the mirror in the mirror library. For example, the user can be provided with a mirror configuration interface with a drop-down menu. The drop-down menu includes many mirrors that can be selected by the user, and the user can choose the mirror to use. Based on this, when it is necessary to provide a mirror image for the third edge cloud node, the central management and control device can obtain the image required by the third edge cloud node from the mirror library, and then provide the image to the third edge cloud node and use the image The permissions are open to the corresponding users. Optionally, the central management and control device may directly issue the target image to the third edge cloud node, or instruct the third edge cloud node to download the target image to a designated storage location.

In addition, the central control device can also maintain the correspondence between the issued image and the edge cloud node where the issued image is located. The correspondence relationship may include the identification information of the issued image and the identification information of the edge cloud node where the image has been issued. The issued image refers to the image that the central control device has provided (for example, issued) to one or some edge cloud nodes; the edge cloud node where the issued image is located refers to the edge cloud node to which the issued image is provided. The same image may be provided (for example, distributed) to one edge cloud node, or may be provided (for example, distributed) to multiple edge cloud nodes.

Based on the maintained correspondence between the issued image and the edge cloud node where the issued image is located, when the image needs to be provided for the third edge cloud node, the central management and control device can also control the third edge cloud node from the image that already has the image. Other edge cloud nodes acquire the image without directly providing the image to the third edge cloud node, which can reduce the processing burden of the central control device to a certain extent, and can also improve the efficiency of image acquisition under the condition of reasonable control.

In detail, when a mirror image needs to be provided to the third edge cloud node, the central management and control device may determine the image that needs to be provided to the third edge cloud node. For ease of description and distinction, in the embodiment of the present application, the third edge cloud node The image provided by the cloud node is recorded as the target image; according to the information of the target image, a match is made in the correspondence between the maintained issued image and the edge cloud node where the issued image is located; if the corresponding relationship is matched with the target Mirror the corresponding fourth edge cloud node, which means that the target image has been provided to the fourth edge cloud node, and the target image at the fourth edge cloud node can be provided to the third edge cloud node; among them, the fourth edge cloud The node can also be an edge cloud node in the network system, and the number can be one or more. For the third edge cloud node, under the control of the central management and control device 101, the target image at the fourth edge cloud node can be obtained.

It is explained here that in the case that the network system 100 includes the edge management and control device 103, the central management and control device may specifically send the information of the fourth edge cloud node and the target mirror to the edge management and control device; the edge management and control device 103 may communicate with the fourth edge cloud node according to the The target image information, the target image at the fourth edge cloud node is provided to the corresponding resource device in the third edge cloud node, for the corresponding resource device to create an instance that can provide cloud computing services based on the target image, and then provide it to the service demander The cloud computing service. The information of the fourth edge cloud node may be any information that can identify the fourth edge cloud node, for example, it may be information such as the ID, name, or geographic location of the fourth edge cloud node. The information of the target image can be any information that can identify the target image, such as the ID, name, or number of the target image.

Further, in the case where the edge management and control device 103 is deployed in both the third edge cloud node and the fourth edge cloud node, the central management and control device 101 may specifically send information about the fourth edge cloud node and the target image to the third edge cloud The edge management and control device in the node is used for the edge management and control device in the third edge cloud node to obtain the target image from the fourth edge cloud node through the communication channel between it and the edge management and control device in the fourth edge cloud node and provide it to the first edge cloud node. Corresponding resource equipment in the three edge cloud nodes. For the edge management and control device 103 in the third edge cloud node, it can receive the information of the fourth edge cloud node and the target mirror image sent by the central management and control device 101, and according to the information of the fourth edge cloud node and the target mirror image, through it and the first The communication channel between the edge management and control devices in the four-edge cloud node obtains the target image from the fourth edge cloud node, and provides the target image to the corresponding resource device in the third edge cloud node, so that the corresponding resource device can create an image based on the target image. Provide instances of cloud computing services, and then provide cloud computing services.

Furthermore, the edge management and control device 103 in the third edge cloud node obtains the target image from the fourth edge cloud node through the communication channel between it and the edge management and control device in the fourth edge cloud node. The process includes: third The edge management and control device 103 in the edge cloud node sends a request for acquiring the target image to the edge management and control device 103 in the fourth edge cloud node through the communication channel between it and the edge management and control device in the fourth edge cloud node. Carry the information of the target image. The edge management and control device 103 in the fourth edge cloud node receives the request, and determines whether the target image exists in the fourth edge cloud node according to the target image information carried in the request, and whether the target image exists in the fourth edge cloud node Next, through the communication channel between it and the edge management and control device 103 in the third edge cloud node, the target image is returned to the edge management and control device 103 in the third edge cloud node, or the target image is mirrored in the fourth edge cloud node The storage address of is returned to the edge management and control device 103 in the third edge cloud node. The edge management and control device 103 in the third edge cloud node receives the target image returned by the edge management and control device 103 in the fourth edge cloud node, or receives the target image returned by the edge management and control device 103 in the fourth edge cloud node in the fourth edge cloud. The storage address in the node, read or download the target image according to the storage address.

It is worth noting that the edge management and control device 103 in the third edge cloud node and the edge management and control device 103 in the fourth edge cloud node may establish a communication channel by themselves, or may establish a channel under the control of the central management and control device 101. Optionally, the central management and control device can also control the establishment of communication channels between different edge management and control devices, and is responsible for maintaining the information of the existing communication channels between the edge management and control devices, for example, which edge management and control devices have established communication channels and communication When the channel is established, the status of the communication channel, and the retention time information. Based on this, after the central management and control device determines that the target image has been provided to the fourth edge cloud node, and before providing the information of the fourth edge cloud node and the target image to the edge management and control device in the third edge cloud node, it can also According to the information of the existing communication channel between the maintained edge management and control devices, determine whether there is a communication channel between the edge management and control device in the third edge cloud node and the edge management and control device in the fourth edge cloud node; if the judgment result is No, that is, there is no communication channel between the edge management and control device in the third edge cloud node and the edge management and control device in the fourth edge cloud node, you can control the edge management and control device in the third edge cloud node and the fourth edge cloud The edge management and control device in the node establishes a communication channel, so that the edge management and control device in the third edge cloud node can obtain the target image from the fourth edge cloud node through the communication channel. In addition, after the edge management and control device in the third edge cloud node establishes a communication channel with the edge management and control device in the fourth edge cloud node, the central management and control device provides the fourth edge cloud node and target image information to the third edge cloud node Edge control equipment in China. Of course, if the judgment result is yes, that is, there is already a communication channel between the edge management and control device in the third edge cloud node and the edge management and control device in the fourth edge cloud node, you can directly mirror the fourth edge cloud node and the target The information is provided to the edge management and control device in the third edge cloud node.

It is worth noting that the central management and control device can also provide the information of the fourth edge cloud node and the target image to the edge management and control device in the third edge cloud node according to the existing communication channel between the maintained edge management and control devices. Information, determine whether there is a communication channel between the edge management and control device in the third edge cloud node and the edge management and control device in the fourth edge cloud node; if the judgment result is no, that is, the edge management and control device in the third edge cloud node and If there is no communication channel between the edge management and control devices in the fourth edge cloud node, the edge management and control device in the third edge cloud node and the edge management and control device in the fourth edge cloud node can be controlled to establish a communication channel to facilitate the third The edge management and control device in the edge cloud node can obtain the target image from the fourth edge cloud node through the communication channel.

In some optional embodiments, in order to ensure the efficiency of obtaining the target image by the third edge cloud node, the central management and control device may also provide the target image at the fourth edge cloud node to the third edge cloud node according to the fourth edge cloud node. The attribute of the cloud node to determine whether the fourth edge cloud node is suitable for providing the target image for the third edge cloud node; if the judgment result is yes, that is, the fourth edge cloud node is suitable for providing the target image for the third edge cloud node, the first The target image at the fourth edge cloud node is provided to the third edge cloud node; if the judgment result is no, the target image can be obtained from the image library and the target image is provided to the third edge cloud node.

It is worth noting that according to different application scenarios and application requirements, different attributes of the fourth edge cloud node can be combined to determine from different angles whether the fourth edge cloud node is suitable for providing a target image for the third edge cloud node. The following example illustrates:

For example, it can be combined with the operator to which the fourth edge cloud node belongs to determine whether the operator to which the fourth edge cloud node belongs is the same as the operator to which the first edge cloud node belongs; if the judgment result is yes, it means that the fourth edge cloud node is The first edge cloud node is an edge cloud node under the same operator. The two can perform data transmission, and the data transmission rate is faster than the cross-operator data transmission rate, which is suitable for providing target mirroring for the first edge cloud node.

For another example, the location attribute of the fourth edge cloud node can be combined to determine whether the distance between the fourth edge cloud node and the third edge cloud node is less than the set distance threshold; if the judgment result is yes, the fourth edge cloud node Close to the third edge cloud node, it is suitable to provide the target image for the third edge cloud node. In this way, the fourth edge cloud node that is closer to the third edge cloud node provides a mirror image for the third edge cloud node, which is convenient for the third edge cloud node. The edge cloud node quickly obtains the image to improve efficiency. The distance between the fourth edge cloud node and the third edge cloud node can be the average distance between two edge cloud nodes, or the distance between the centers of two edge cloud nodes, or two edge clouds The distance between nodes and the nearest outer edge can be flexibly defined according to requirements.

For another example, the bandwidth attribute of the fourth edge cloud node can be combined to determine whether the available bandwidth of the fourth edge cloud node is greater than the set bandwidth threshold; if the judgment result is yes, it means that the bandwidth resource of the fourth edge cloud node is relatively abundant, and it is suitable for The third edge cloud node provides the target image, so that the fourth edge cloud node with sufficient bandwidth resources provides the image for the third edge cloud node, which can ensure the transmission rate of the image, and facilitate the third edge cloud node to quickly obtain the image and improve efficiency .

For another example, it can be combined with the load attribute of the fourth edge cloud node to determine whether the load of the fourth edge cloud node is less than the set load threshold; if the judgment result is yes, it means that the load of the fourth edge cloud node is lighter, and it is suitable for The third edge cloud node provides the target image, so that the lighter-loaded fourth edge cloud node provides the image for the third edge cloud node. On the one hand, it can achieve load balancing, and on the other hand, it is also convenient for the third edge cloud node to quickly obtain the image. ,Improve efficiency.

It is worth noting that the several methods listed above can be used alternatively or combined in any combination. Regarding the combined use, this will not be described too much.

Further, in the case of multiple fourth edge cloud nodes, multiple attributes of the fourth edge cloud node can be combined to use the above several methods in combination, and then select a target image suitable for the first edge cloud node. The fourth edge cloud node. For example, if there are multiple fourth edge cloud nodes, the operators to which the multiple fourth edge cloud nodes belong can be combined to select the first edge cloud node belonging to the same operator as the first edge cloud node from the multiple fourth edge cloud nodes. Four edge cloud nodes; furthermore, if there are still multiple selected fourth edge cloud nodes, you can further select the minimum load or lower than the set load threshold according to the load of the selected fourth edge cloud node The fourth edge cloud node provides a target image for the first edge cloud node.

In some optional embodiments, the target image may have been provided to the third edge cloud node. For example, in a business expansion scenario, it is necessary to create a new instance in the edge cloud node that is currently providing cloud computing services to the service demander. The image used is the same as the image used by the previous instance. If the edge cloud node still stores the image used by the previous instance, there is no need to repeatedly provide the image for the edge cloud node. In view of this situation, in order to save resources, the central management and control device can determine the maintained issued image and the edge cloud node where the issued image is located before providing the target image at the fourth edge cloud node to the third edge cloud node Whether the third edge cloud node is included in the corresponding relationship; if the judgment result is yes, it indicates that the target image has been provided to the third edge cloud node, and the target image is still stored in the third edge cloud node, then the target image can be The information is provided to the third edge cloud node for the third edge cloud node to read the target image stored in it, without the need to transmit the target image again, which can save network resources consumed by the transmission of the target image; if the judgment result is no, it indicates that it has not been sent to The third edge cloud node has provided the target image, or the target image no longer exists in the third edge cloud node, the target image at the fourth edge cloud node may be provided to the third edge cloud node. Among them, when the edge management and control device is deployed in the third edge cloud node, if the central management and control device determines that the corresponding relationship between the maintained issued image and the edge cloud node where the issued image contains the target image, the target image The information of the image is provided to the edge management and control device in the third edge cloud node, and the edge management and control device in the third edge cloud node can obtain the target image from the storage space of the image in the third edge cloud node according to the information of the target image. The image is provided to the corresponding resource device in the third edge cloud node, so that the corresponding resource device can create an instance that can provide cloud computing services according to the target image.

Further optionally, the same edge cloud node may provide multiple cloud computing services for the same user or different users, and may receive multiple images, and these images will be stored in the edge cloud node. Edge cloud nodes can provide a certain amount of storage space for storing images. Considering that the storage space of the image in the edge cloud node is limited, in order to have enough storage space to store the newly received image, the edge cloud node needs to eliminate the locally stored image. In this embodiment, the central management and control device is responsible for providing a mirroring elimination strategy for edge cloud nodes. The central management and control device can generate the elimination strategy of the image, deliver the elimination strategy to each edge cloud node, and each edge cloud node performs elimination processing on the stored image according to the elimination strategy. Among them, in the case that the network system includes edge management and control equipment, the central management and control equipment can issue the elimination strategy to the edge management and control equipment, and the edge management and control equipment eliminates the images stored in each edge cloud node according to the elimination strategy. Furthermore, in the case where edge management and control equipment is deployed in each edge cloud node, the central management and control equipment can issue the elimination strategy to the edge management and control equipment in each edge cloud node, and the edge management and control equipment in each edge cloud node will be The elimination strategy eliminates the image stored in the edge cloud node to which it belongs.

Optionally, the elimination strategy may be an elimination strategy with the earliest receiving time, that is, according to the receiving time of the image, the image with the earliest receiving time is preferentially eliminated. Alternatively, the elimination strategy may be the elimination strategy with the least frequency of use, that is, the image with the least frequency of use is preferentially eliminated according to the frequency of use of the image. Alternatively, the elimination strategy may be the elimination strategy with the largest resource occupation, that is, according to the size of the storage space occupied by the image, the image with the largest storage space is first eliminated.

For edge cloud nodes, the image stored in the node can be eliminated regularly according to the above elimination strategy; or, whenever a new image needs to be received or acquired, it can be judged whether there is enough storage space in the node for storage If there is not enough storage space in the current node for the new image, the image stored in the node is eliminated according to the above elimination strategy, so as to store the new image. Taking the third edge cloud node to obtain the target image from the fourth edge cloud node as an example, before the edge management and control device in the third edge cloud node obtains the target image from the fourth edge cloud node, the edge of the third edge cloud node The management and control device can determine whether there is enough storage space in the third edge cloud node to store the target image; if there is not enough storage space in the third edge cloud node, it will eliminate the image stored in the third edge cloud node according to the elimination strategy. In order to have enough storage space to store the target image. Optionally, if there is enough storage space in the third edge cloud node, the image stored in the third edge cloud node may not be eliminated temporarily.

Optionally, as shown in FIG. 1c, the network system 100 further includes: an image construction device 104. The image construction device 104 may be deployed in one or more edge cloud nodes, and is mainly responsible for the construction and verification of application images. The image construction device 104 can provide an edge cloud environment, can build an image that is compatible with the edge cloud environment, and can also verify whether the image is compatible with the edge cloud environment. The image that is not compatible with the edge cloud environment can be reconstructed or output Adapted prompt information, etc. Based on the image building device 104, the user can add a new image to the network system 100.

In an optional implementation of adding a new image, a user (such as a service demander) can submit a third request for adding a new image to the central management and control device, and the third request includes image construction information; The device sends a construction request, which includes image construction information; after receiving the construction request, the image construction device obtains the image construction information from it, constructs an image adapted to the edge cloud environment based on the image construction information, and returns the constructed image to the center Control equipment; the central control equipment receives the newly constructed mirror image returned by the mirror construction equipment and adds it to the mirror library to continuously enrich the mirror library.

In another optional implementation of newly-added mirroring, a mirroring rule and specification can be provided to users (such as service demanders), allowing users to make or generate mirrors by themselves. The mirrors generated or made by users need to conform to the edge cloud Environmental safety, regulations and other related requirements. After making or generating the image, the user can send a fourth request for adding a new image to the central control device. The fourth request includes the image to be added. The new image refers to the image made or generated by the user. This embodiment does not It does not limit the way users make or generate images. The central control device receives the fourth request, obtains the image to be added from the fourth request, and sends the image to be added to the image construction device; the image construction device adapts the image to be added to the edge cloud environment; if it is to be added The image is adapted to the edge cloud environment, and the image construction device returns a message to the central control device that the new image is adapted to the edge cloud environment; if the new image is not compatible with the edge cloud environment, the image construction device returns to the central control device A message that the new image is not compatible with the edge cloud environment.

For the central control equipment, if it receives a message from the mirror construction device that the new image to be added is adapted to the edge cloud environment, it will add the new mirror to the mirror library; if the mirror construction service is received, the mirror construction device returns The message that the new image to be added is not compatible with the edge cloud environment, or informs the user to re-submit the new image after reconstruction, or informs the user to provide the reconstruction method of the new image for the image building service image building equipment According to the reconstruction method, the newly added image is reconstructed into an image adapted to the edge cloud environment. If the user provides a reconstruction method for the newly added image, the central control device can provide the reconstruction method to the image construction device, and the image construction device reconstructs the newly added image according to the reconstruction method to make it compatible with the edge cloud environment It adapts and returns the reconstructed image to the central control device; the central control device receives the reconstructed image and adds it to the mirror library.

It is explained here that the image construction device 104 may be a logical device with functions such as image construction and verification (for example, it may be an instance that can provide image construction environment and resources, and has functions such as application deployment and image verification). These functions can be It can be implemented on one physical machine or virtual machine, or it can be distributed on multiple physical machines or virtual machines. Of course, the image construction device 104 of this embodiment may also be one or more physical devices with functions such as image construction and verification. The embodiments of this application do not limit the implementation structure of the image construction device, and any device structure with the above-mentioned functions is applicable to the embodiments of this application.

In the embodiments of the present application, not only mirrors can be added to the mirror library, but also mirrors that are not used or have not been used for a long time can be deleted to save storage space. For example, the central management and control device can count the usage frequency of each mirror in the mirror library regularly or in real time, use mirrors with a frequency less than the frequency threshold as the mirrors to be deleted, and execute the mirror deletion process to delete them. For another example, the central management and control device may also receive a mirror deletion request submitted by a user (such as a service demander), use the mirror deleted in the mirror deletion request as a mirror to be deleted, and execute the mirror deletion process to delete it. Wherein, the image deletion request may carry information of the image to be deleted, such as ID, name, or serial number.

For the central control device, any of the above methods can be used to determine the image to be deleted. After the image to be deleted is determined, the image to be deleted can be deleted from the mirror library on the one hand, and the image to be deleted can be indicated to be stored on the other hand. The edge cloud node will delete the image to be deleted. Among them, the central management and control device may match the maintained corresponding relationship between the issued image and the edge cloud node where the issued image is located according to the image to be deleted, and determine the edge cloud node storing the image to be deleted according to the matching result. If the fifth edge cloud node corresponding to the image to be deleted is matched in the corresponding relationship, it means that the image to be deleted has been issued to the fifth edge cloud node, and the image to be deleted is still stored in the fifth edge cloud node, so The fifth edge cloud node sends a deletion instruction, and the deletion instruction carries information about the image to be deleted to instruct the fifth edge cloud node to delete the image to be deleted stored therein. The fifth edge cloud node may be one or multiple.

It is explained here that when the network system 100 includes the edge management and control device 103, the central management and control device may specifically send a deletion instruction to the edge management and control device 103; the edge management and control device 103 receives the deletion instruction issued by the central management and control device, and then deletes the instruction from the Obtain the information of the image to be deleted in, and determine whether the image to be deleted is stored in the fifth edge cloud node according to the information of the image to be deleted; if the image to be deleted is stored in the storage, delete the image to be deleted in the fifth edge cloud node. Further, if the edge management and control device 103 is deployed in the fifth edge cloud node, the central management and control device 101 may specifically send a deletion instruction to the edge management and control device 103 in the fifth edge cloud node; the edge management and control device in the fifth edge cloud node 103 receives the delete instruction issued by the central management and control device, obtains the information of the image to be deleted from the delete instruction, and determines whether the image to be deleted is stored in the fifth edge cloud node according to the information of the image to be deleted; The image to be deleted stored in the fifth edge cloud node is deleted.

When the central control device deletes the image to be deleted from the image library, and the edge cloud node storing the image to be deleted also deletes the image to be deleted stored in it, the image deletion process is completed.

Regarding the resource equipment in the edge cloud node, regardless of the method, after the image is obtained, the capabilities that can be supported by hardware or software under the control of the central control device 101 or the edge control device 103 are in the form of virtualization as an example Provide resources such as computing, network, and storage, and the corresponding image will be mounted to the corresponding instance in the form of a system disk. After the instance is created, try to start the instance. After the corresponding instance is successfully started, the capabilities of these resource devices can be used to provide cloud computing services. Among them, the resource device provides computing, network, and storage resources for the instance under the control of the edge management and control device, including: the edge management and control device applies for related resources from the resources allocated or reserved in the target edge node cloud according to the resource template provided by the central management and control device The computing resources, storage resources and/or network resources of the target edge cloud node are used to create related resources by calling the calculation, storage, network and other executors in the target edge cloud node. Among them, resource creation actions include: processing storage-related resources, creating an instance system disk based on the configuration information and content of the image, creating a corresponding data disk based on the resource template; creating network resources that the instance depends on, such as IP addresses, virtual switches And so on; and combine resource templates to create computing resources.

Operation and maintenance management functions:

In the embodiment of the present application, the central management and control device may perform operation and maintenance management and control on at least one edge cloud node with the assistance of the edge management and control device. In detail, the edge management and control device may perform operation and maintenance monitoring on at least one edge cloud node and report the operation and maintenance monitoring data to the central management and control device, so that the central management and control device can manage and control at least one edge cloud node according to the operation and maintenance monitoring data. The central management and control device can perform operation and maintenance control on at least one edge cloud node according to the operation and maintenance monitoring data reported by the edge management and control device. Optionally, for the edge management and control equipment, the operation and maintenance monitoring of at least one edge cloud node can be carried out under the control of the central management and control equipment and the operation and maintenance monitoring data is reported to the central management and control equipment for the central management and control equipment according to the operation and maintenance The monitoring data controls the operation and maintenance of at least one edge cloud node. Alternatively, the edge management and control device may periodically perform operation and maintenance monitoring on at least one edge cloud node according to a timed task and report the operation and maintenance monitoring data to the central management and control device. In either implementation mode, the edge management and control equipment mainly performs functions such as monitoring, data collection, and reporting, while the operation and maintenance decisions are determined by the central management and control equipment.

Among them, the central management and control device controls the edge management and control device to perform operation and maintenance monitoring of at least one edge cloud node, which can adopt but not limited to the following optional implementation manners:

In an optional embodiment, the central management and control device may send the first type of operation and maintenance monitoring instruction to the edge management and control device to instruct the edge management and control device to perform operation and maintenance monitoring on at least one edge cloud node from at least one operation and maintenance dimension and to The operation and maintenance monitoring data in the operation and maintenance dimension is reported to the central control equipment. The first type of operation and maintenance monitoring instruction is a monitoring instruction that instructs the edge management and control device to perform operation and maintenance monitoring of at least one edge cloud node from at least one operation and maintenance dimension and report operation and maintenance monitoring data in at least one operation and maintenance dimension. For edge management and control equipment, it can receive the first type of operation and maintenance monitoring instructions sent by the central management and control equipment, and according to the first type of operation and maintenance monitoring instructions, perform operation and maintenance monitoring of at least one edge cloud node from at least one operation and maintenance dimension, and Report the operation and maintenance monitoring data on at least one operation and maintenance dimension to the central control equipment. The central management and control device controls the operation and maintenance of at least one edge cloud node according to the operation and maintenance monitoring data in at least one operation and maintenance dimension reported by the edge management and control device. It is worth noting that at least one operation and maintenance dimension can be flexibly set according to application requirements and preset into edge control equipment and central control equipment. For examples of operation and maintenance dimensions, refer to the subsequent embodiments.

In another optional implementation manner, the central management and control device may selectively perform operation and maintenance control on at least one edge cloud node in one or some operation and maintenance dimensions. Based on this, the central management and control device can send the second type of operation and maintenance monitoring instructions to the edge management and control device. The second type of operation and maintenance monitoring instruction corresponds to the specified operation and maintenance dimension, and is used to instruct the edge management and control device to check at least one edge in the specified operation and maintenance dimension. The cloud node performs operation and maintenance monitoring and reports the operation and maintenance monitoring data on the specified operation and maintenance dimension. For edge management and control equipment, it can receive the second type of operation and maintenance monitoring instructions sent by the central management and control equipment, and perform operation and maintenance monitoring on at least one edge cloud node in the specified operation and maintenance dimension according to the second type of operation and maintenance monitoring instructions, and specify The operation and maintenance monitoring data in the operation and maintenance dimension is reported to the central management and control device, so that the central management and control device can perform operation and maintenance control on at least one edge cloud node according to the operation and maintenance monitoring data in the designated operation and maintenance dimension. The central management and control device is also used to receive the operation and maintenance monitoring data in the specified operation and maintenance dimension sent by the edge management and control device, and perform operation and maintenance control on at least one edge cloud node according to the operation and maintenance monitoring data in the specified operation and maintenance dimension.

Optionally, the edge management and control device periodically performs operation and maintenance monitoring of at least one edge cloud node according to a timed task, may periodically perform operation and maintenance monitoring of at least one edge cloud node from at least one operation and maintenance dimension according to a timed task; Further, the operation and maintenance monitoring data in at least one operation and maintenance dimension can be reported to the central control equipment. Among them, the monitoring period on different operation and maintenance dimensions may be the same or different. For example, the edge management and control device can scan the edge cloud node for security vulnerabilities every 10 minutes, or monitor the traffic of the edge cloud node every 5 minutes.

It is worth noting that there can be one or more designated O&M dimensions. In the case of multiple designated operation and maintenance dimensions, each designated operation and maintenance dimension can correspond to a second-type operation and maintenance monitoring instruction, that is, the central control device can send multiple second-type operation and maintenance monitoring instructions to the edge control device. The second type of operation and maintenance monitoring instruction corresponds to a specified operation and maintenance dimension. Or, in the case of multiple designated operation and maintenance degrees, multiple designated operation and maintenance dimensions can also correspond to the same second-type operation and maintenance monitoring instruction, that is, the central management and control device can send a second-type operation and maintenance monitoring to the edge management and control device Instruction, this second type of operation and maintenance monitoring instruction corresponds to multiple specified operation and maintenance dimensions.

The aforementioned at least one operation and maintenance dimension or specified operation and maintenance dimension may include but is not limited to the following dimensions: the object dimension in the running state, the log dimension, the security dimension, and the resource dimension. Further, the object dimension in the running state may include the operating state dimension of the object and/or the life cycle dimension of the object; the security dimension may include: the traffic attack dimension and/or the security vulnerability dimension.

Combining the several O&M dimensions listed above, the central management and control device, with the assistance of the edge management and control device, performs O&M control on at least one edge cloud node, including but not limited to at least one of the following O&M control examples:

Operation and maintenance management and control example 1: The central management and control device controls the edge management and control device to monitor the status of objects in at least one edge cloud node that are in operation. Among them, the control method includes sending a first type of operation and maintenance monitoring instruction to the edge management and control device or sending a second type of operation and maintenance monitoring instruction corresponding to the operating state dimension of the object. The edge management and control equipment is under the control of the central management and control equipment, or periodically according to timing tasks, monitors the status of the objects in the running state of at least one edge cloud node, and reports the running status of the monitored objects in the running state to Central control equipment. The central management and control equipment identifies objects with abnormal operating status from the operating status of the objects in the operating status reported by the edge management and control equipment. For ease of description and distinction, the objects with abnormal operating status are called target objects, and exception handling is performed on the target objects. . Among them, the objects in the running state in the edge cloud node include, but are not limited to: instances, images, containers, other virtual components, physical machines, CPUs, and/or hard disks. According to the different objects in the running state, the abnormal situation of the running state will be different. For example, for the instance, possible abnormal conditions include, but are not limited to: interruption, error reporting, and/or failure. For another example, for a physical machine, possible abnormal conditions include, but are not limited to: crashes, black screens, alarms, and/or crashes of applications running on the physical machine. Depending on the target object and the abnormal situation of the running state, the exception handling method will be different, for example, it can include but not limited to: alarm, stop or restart the target object, migrate, delete and rebuild the target object, etc.

Operation and maintenance control example 2: The central control device controls the edge control device to monitor the life cycle of at least one edge cloud node in the running state. Among them, the control method includes sending the first type of operation and maintenance monitoring instruction to the edge management and control device or sending the second type of operation and maintenance monitoring instruction corresponding to the life cycle dimension of the object. The edge control device is under the control of the central control device, or periodically according to a scheduled task, monitors the life cycle of at least one edge cloud node in the running state, and reports the life cycle of the monitored object in the running state Give the center control equipment. The central management and control device controls the stopping, restarting, migration or deletion of the running object after stopping, according to the life cycle of the running object reported by the edge management and control device.

Operation and maintenance control example 3: The central control device controls the edge control device to collect log data in at least one edge cloud node. Among them, the control method includes sending the first type of operation and maintenance monitoring instruction to the edge management and control device or sending the second type of operation and maintenance monitoring instruction corresponding to the log dimension. The edge management and control device collects log data in at least one edge cloud node under the control of the central management and control device or periodically according to a timed task, and reports the collected log data to the central management and control device. The central management and control device receives the log data reported by the edge management and control device, performs data analysis on the log data, and performs follow-up actions based on the data analysis results, such as billing, risk control, and/or adding or subtracting instances. Depending on the log data, follow-up actions will vary. Optionally, log data may include, but is not limited to: various performance, indicators and other data in edge cloud nodes, such as: instance bandwidth traffic, instance current running status, instance IO load, physical machine bandwidth traffic, physical machine The current operating status, the IO load of the physical machine, the operating status of the edge management and control equipment, and/or the operating status of other virtualization components, etc.

Optionally, the central control device can not only collect the log data of each edge cloud node reported by the edge control device, but also has the ability to perform data inspection. For some data, if the data stored by the central control device is inconsistent with the data in the edge cloud node, The latest data can be actively synchronized with the edge cloud node, for example, the latest version of the image can be synchronized with the edge cloud node.

Operation and maintenance control example 4: The central control device controls the edge control device to monitor the traffic of at least one edge cloud node. Among them, the control method includes sending the first type of operation and maintenance monitoring instruction to the edge management and control device or sending the second type of operation and maintenance monitoring instruction corresponding to the traffic attack dimension. The edge management and control equipment is under the control of the central management and control equipment, or periodically according to a timing task, monitors the flow of at least one edge cloud node, and reports the monitored traffic attack events to the central management and control equipment. The central management and control equipment blocks traffic attack events that occur in edge cloud nodes. Optionally, the edge management and control device may also report the monitored flow data to the central management and control device, and the central management and control device may also perform flow attack defense on at least one edge cloud node based on the flow data.

Operation and maintenance control example 5: The central control equipment controls the edge control equipment to scan for network security vulnerabilities on at least one edge cloud node. Among them, the control method includes sending the first type of operation and maintenance monitoring instruction to the edge management and control device or sending the second type of operation and maintenance monitoring instruction corresponding to the network security dimension. The edge management and control equipment is under the control of the central management and control equipment, or periodically according to timing tasks, scans for network security vulnerabilities on at least one edge cloud node, and reports the scanned network security vulnerabilities to the central management and control equipment. The central control equipment receives the network security vulnerabilities reported by the edge control equipment, and repairs the network security vulnerabilities.

Operation and maintenance control example 6: The central control device controls the edge control device to monitor the resource usage in at least one edge cloud node. Wherein, the control method includes sending the first type of operation and maintenance monitoring instruction to the edge management and control device or sending the second type of operation and maintenance monitoring instruction corresponding to the resource dimension. The edge management and control device is under the control of the central management and control device, or periodically according to a timed task, monitors the resource usage in at least one edge cloud node, and reports the monitored resource usage information to the central management and control device. The central management and control device performs resource expansion or reduction on at least one edge cloud node based on the resource usage information reported by the edge management and control device. The resources here include various resource information, such as equipment resources such as physical machines, storage resources, computing resources such as CPUs and GPUs, and network resources such as bandwidth.

Further, if edge management and control equipment is deployed in each edge cloud node, each edge management and control device can, under the control of the central management and control equipment, perform operation and maintenance monitoring on its edge cloud node and monitor the operation and maintenance of its edge cloud node. The operation and maintenance monitoring data is reported to the central control equipment. The central management and control device can receive the operation and maintenance monitoring data reported by the edge management and control device in each edge cloud node, and perform operation and maintenance management and control on each edge cloud node according to the operation and maintenance monitoring data in each edge cloud node.

The embodiments of the present application do not limit the implementation structure of the central management and control device and the edge management and control device. Optionally, a structural framework of a central management and control device includes: a resource scheduling management and control module, a mirror management and control module, and a central operation and maintenance module; the central operation and maintenance module further includes: a central monitoring unit, a central log unit, and a central security unit. Correspondingly, a structural framework of edge management and control equipment includes: a resource scheduling service module, a mirroring service module, and an edge operation and maintenance module; the edge operation and maintenance module further includes: an edge monitoring unit, an edge log unit, and an edge security unit.

Among them, the resource scheduling management and control module in the central management and control device cooperates with the resource scheduling service module in the edge management and control device to perform resource scheduling on edge cloud nodes. For the resource scheduling function, please refer to the description below. The image management and control module in the central management and control device cooperates with the image service module in the edge management and control device to perform image management and distribution for edge cloud nodes. For image management and distribution functions, please refer to the description below.

The central operation and maintenance module in the central management and control equipment and the edge operation and maintenance module in the edge management and control equipment cooperate with each other to perform operation and maintenance management and control on the edge cloud nodes. The above operation and maintenance control examples 1-6 can be implemented by the corresponding units in the central operation and maintenance module and the edge operation and maintenance module. Operation and maintenance control example 3 can be realized by the cooperation of the central log unit in the central operation and maintenance module and the edge log unit in the edge operation and maintenance module. In detail, the central log unit sends the first type of operation and maintenance monitoring instruction or the second type of operation and maintenance monitoring instruction corresponding to the log dimension to the edge log unit; the edge log unit collects the edge according to the first or second type of operation and maintenance monitoring instruction The log data in the cloud node is reported to the central log unit; the central log unit performs data analysis on the log data and executes follow-up actions based on the data analysis results. Operation and maintenance control examples 4 and 5 can be realized by the cooperation of the central security unit in the central operation and maintenance module and the edge security unit in the edge operation and maintenance module. In detail, the central security unit sends the first type of operation and maintenance monitoring instruction to the edge security unit or sends the second type of operation and maintenance instruction corresponding to the traffic attack or the network security dimension; the edge security unit can be based on the first or second type of operation and maintenance Instruct the edge cloud nodes to perform traffic monitoring or network security vulnerability scanning, and report the monitored traffic attack events or network vulnerability security issues to the central security unit; the central security unit blocks traffic attack events or conducts network security vulnerability issues repair. Operation and maintenance control examples 1, 2 and 6 can be implemented by the central monitoring unit in the central operation and maintenance module and the edge monitoring unit in the edge operation and maintenance module, and the detailed implementation process is not repeated.

It can be seen from the above that with the assistance of the edge management and control equipment, the central management and control equipment can understand the health, resource usage, log data and/or infrastructure conditions of each instance in the edge cloud node, and can realize remote operation and maintenance, log management, etc.

In the embodiments of the present application, in addition to the central management and control device that can perform O&M management and control on at least one edge cloud node, in the case where the central management and control device does not perform O&M management and control on the edge cloud node or cannot perform O&M management and control on the edge cloud node , The edge management and control device can autonomously perform operation and maintenance management and control on at least one edge cloud node.

For example, the edge management and control device can monitor the connection between it and the central management and control device. When the connection with the central management and control device is lost, it can be determined that the central management and control device cannot perform operation and maintenance control on the edge cloud node. One operation and maintenance dimension controls the operation and maintenance of at least one edge cloud node.

For another example, when the central management and control device sends the first type of operation and maintenance monitoring instructions to the edge management and control device to control the edge management and control device to perform operation and maintenance monitoring of at least one edge cloud node, the edge management and control device can wait to receive the central management and control device to send If it does not receive the first type of operation and maintenance monitoring instruction sent by the central control device, it can be determined that the central control device is incorrect or cannot perform the operation and maintenance control on at least one edge cloud node. At least one operation and maintenance dimension controls the operation and maintenance of at least one edge cloud node. Optionally, the edge management and control device and the central management and control device may pre-appoint the waiting time for the first type of operation and maintenance monitoring instruction. If the waiting time is exceeded and the first type of operation and maintenance monitoring instruction sent by the central management and control device is not received, then It is determined that the first type of operation and maintenance monitoring instruction sent by the central control equipment has not been received.

For another example, the central management and control device sends the second type of operation and maintenance monitoring instructions corresponding to the specified operation and maintenance dimension to the edge management and control device to control the manner in which the edge management and control device performs operation and maintenance monitoring of at least one edge cloud node from the specified operation and maintenance dimension The edge management and control device can wait to receive the second type of operation and maintenance monitoring instruction sent by the central control device. If the second type of operation and maintenance monitoring instruction sent by the central control device is not received in the specified operation and maintenance dimension, it can be determined that the central control device is in If the specified operation and maintenance dimension is incorrect or unable to perform operation and maintenance control on at least one edge cloud node, it is possible to autonomously perform operation and maintenance control on at least one edge cloud node from the specified operation and maintenance dimension.

Further optionally, if the edge management and control device autonomously controls the operation and maintenance of at least one edge cloud node from at least one operation and maintenance dimension when the connection with the central management and control device is lost, then after the connection with the central management and control device is restored, The operation and maintenance control data during the loss of connection can be synchronized to the central control equipment. It is worth noting that the operation and maintenance control data mainly includes data such as strategies, methods, and effects of operation and maintenance control, and of course, it can also include operation and maintenance monitoring data.

The above-mentioned at least one operation and maintenance dimension or the designated operation and maintenance dimension may include, but is not limited to, the following dimensions: the object dimension in the running state, the log dimension, the security dimension, and the resource dimension. Further, the object dimension in the running state may include the operating state dimension of the object and/or the life cycle dimension of the object; the security dimension may include: the traffic attack dimension and/or the security vulnerability dimension.

Combining the several O&M dimensions listed above, the edge management and control device autonomously performs O&M control on at least one edge cloud node, including but not limited to at least one of the following O&M control examples:

Operation and maintenance management and control example a: autonomously monitor the status of objects in the running state in at least one edge cloud node, and perform exception handling for the monitored target objects whose running status is abnormal. Regarding the objects in the running state and the abnormal conditions of the running state, please refer to the above description, which will not be repeated here.

Optionally, in example a, when the edge management and control device performs abnormal processing on the target object, it is specifically used to: analyze the abnormal operating state of the target object, and determine at least one candidate processing method according to the analysis result; In the candidate processing method, the target processing method is acquired, and the target object is abnormally processed according to the target processing method.

Furthermore, when the edge management and control device obtains the target processing mode, it is specifically used to: when the edge management and control device maintains a connection with the central management and control device, report at least one candidate processing method to the central management and control device for the central management and control device to use. Select the processing method; receive the processing method returned by the central control device as the target processing method; or, in the case that the edge control device loses the connection with the central control device, output at least one candidate processing method to the edge operation and maintenance control personnel for the edge The operation and maintenance personnel select the processing method; in response to the selection operation of the edge operation and maintenance management and control personnel, determine the selected processing method as the target processing method; or, in the case of loss of connection with the central control equipment, follow the set selection strategy, The target processing method is selected from at least one candidate processing method.

Operation and maintenance management and control example b. Autonomously monitor the life cycle of objects in the running state in at least one edge cloud node, and control the objects in the running state to stop, restart or delete after stopping according to the monitoring results. For containers or instances, you can control the container or instance to stop execution, restart after stopping, or delete the container or instance, etc.

Operation and maintenance control example c: autonomously collect log data in at least one edge cloud node, perform data analysis on the log data, and perform follow-up actions based on the data analysis results. Log data includes, but is not limited to, the bandwidth traffic of the instance in the edge cloud node, the current running status of the instance, the IO load of the instance, the bandwidth traffic of the physical machine, the current running status of the physical machine, the IO load of the physical machine, and the operation of edge control equipment. Status and/or operation status of other virtualization components. Optionally, subsequent actions such as billing, risk control, and resource reallocation can be performed according to the analysis result of the log data, but are not limited to this.

Operation and maintenance control example d. Autonomously monitor the traffic of at least one edge cloud node, and block the monitored traffic attack events.

Operation and maintenance control example e: Autonomously scan for network security vulnerabilities on at least one edge cloud node, and fix the scanned network security vulnerabilities.

Operation and maintenance control example f: autonomously monitor the resource usage in at least one edge cloud node, and perform resource expansion or reduction on at least one edge cloud node according to the monitoring result. The resources here include but are not limited to: equipment resources such as physical machines, storage resources such as memory and disks, computing resources such as CPU and GPU, and network resources such as bandwidth. For these resources, when the usage is high, the capacity can be expanded for these resources, and when the usage is low, the capacity can be reduced for these resources.

Further, if edge management and control equipment is deployed in each edge cloud node, each edge management and control device can autonomously belong to its edge when the central management and control device is incorrect or cannot perform operation and maintenance control on its edge cloud node. Cloud nodes perform operation and maintenance management and control.

Optionally, in the foregoing example a-example e in which the edge management and control device autonomously controls the operation and maintenance of at least one edge cloud node, the edge management and control device may periodically perform the operation, maintenance, management and control on at least one edge cloud node according to a timing task . For example, in example d, the edge management and control device can monitor the traffic of at least one edge cloud node every 10 minutes according to the scheduled task, and block the monitored traffic attack event. For another example, in Example e, the edge management and control device may scan for network security vulnerabilities on at least one edge cloud node every 5 minutes according to a scheduled task, and fix the scanned network security vulnerabilities. Of course, the edge management and control device can also autonomously control the operation and maintenance of at least one edge cloud node according to other independent strategies. For example, it can autonomously control the operation and maintenance of at least one edge cloud node at a fixed time every day. .

Combining the above examples 1-6 and example af, it can be seen that in this embodiment, the central management and control device is combined with the edge management and control device, and the central management and control device can perform operation, maintenance, management and control on at least one edge cloud node with the assistance of the edge management and control device, except In addition, the edge management and control equipment also has a certain ability of self-operation, maintenance, and control. When the central management and control equipment is not correct or the edge cloud node cannot be operated and maintained, the edge cloud node can be independently operated and maintained to achieve Two-level operation and maintenance management and control can more fully and comprehensively control the operation and maintenance of edge cloud nodes, and provide conditions for "putting cloud computing in edge cloud nodes closer to the terminal for processing", and then can use edge cloud nodes The resources to provide users with cloud computing services are conducive to reducing response delays, reducing the pressure on the central cloud or traditional cloud computing platforms, and reducing bandwidth costs.

In summary, in the network system provided by the embodiments of the present application, the resources, mirroring, instances, operation and maintenance of edge cloud nodes are uniformly controlled based on centralized management and control, and the edge cloud nodes can be managed and coordinated to the greatest extent. , It can reduce errors caused by single-point self-control or unsynchronized information of the entire network, and can use the characteristics of centralized management to achieve the optimization of resource scheduling, avoiding the waste of local resources at the edge.

In addition to the foregoing network system, the embodiments of the present application provide example management and control methods from the perspective of central management and control equipment, which are described in detail below.

Fig. 2a is a schematic flowchart of an example management and control method provided by an exemplary embodiment of this application. As shown in Figure 2a, the method includes:

21a. Determine at least one instance deployed in at least one edge cloud node in the network system, where at least one instance can provide cloud computing services for the service demander.

22a. Manage and control at least one instance, so that at least one instance provides cloud computing services for service demanders.

In this embodiment, the network system includes at least one edge cloud node, at least one instance is deployed in the at least one edge cloud node, and at least one instance can provide cloud computing services for service demanders. The central management and control device determines at least one instance in the edge cloud node, and controls the at least one instance so that the at least one instance provides cloud computing services for the service demander. The service demander here may be any device, application, system or another service that needs to use the cloud computing service provided by the instance in the edge cloud node. Taking the system as an example, the service demander can be, but not limited to: online video systems, risk management systems, customer information management systems, or data distribution systems.

Optionally, the central management and control device may perform various management and control on at least one instance, for example, it may include at least one of upgrade, migration, shutdown, restart, and release, but is not limited thereto.

As shown in Figure 2b, the process for the central control equipment to upgrade and control an instance includes the following steps:

21b. Determine the instance to be upgraded from at least one instance;

22b. Send an upgrade request to the service demander for the service demander to determine an upgrade strategy for the instance to be upgraded in combination with the business situation on the instance to be upgraded;

23b. Receive the upgrade strategy returned by the service demander, and upgrade the instance to be upgraded according to the upgrade strategy.

In actual applications, as business requirements change or the mirror version is updated, it is possible to upgrade the mirror or the corresponding instance. The central control device can determine the instance to be upgraded from at least one instance, and there can be one or more instances to be upgraded; send an upgrade request to the service demander, so that the service demander can determine the instance to be upgraded based on the business situation of the instance to be upgraded Upgrade strategy. The upgrade request carries the identification information of the instance to be upgraded, such as the ID and name of the instance to be upgraded. It can also be the ID and name of the service corresponding to the instance to be upgraded. It can also be the ID, name and other information of the image corresponding to the instance to be upgraded. . After receiving the upgrade request, the service demander can determine the instance to be upgraded according to the upgrade request, and combine the business conditions on the instance to be upgraded, such as the business request and response status on the instance to be upgraded, to determine whether the instance to be upgraded is suitable for upgrade. What time is suitable for upgrading, what method to use for upgrading, etc., and then generate an upgrade strategy for the instance to be upgraded and return it to the central control device. The central control equipment receives the upgrade strategy sent by the service demander, and upgrades the instance to be upgraded according to the upgrade strategy.

In an optional embodiment, the service demander can combine the business conditions on the instance to be upgraded, such as the number of business requests that have been received and not yet completed (referred to as inventory business requests), and whether there are any new business requests ( Incremental service request), etc., to determine when the instance to be upgraded can be upgraded, that is, the upgrade strategy can include the upgrade time. Based on this, the central control device can start to upgrade the instance to be upgraded at the upgrade time specified in the upgrade policy. In addition, the upgrade strategy may include an upgrade method. Based on this, the central control device may use the upgrade method specified in the upgrade strategy to upgrade the instance to be upgraded. Optionally, the upgrade strategy may include an upgrade time and an upgrade method, and the central management and control device may adopt the upgrade method specified in the upgrade strategy, and upgrade the instance to be upgraded at the upgrade time specified in the upgrade strategy. Optionally, the upgrade strategy may also include information such as whether to upgrade, and in the case of upgrade, it further includes the upgrade time and/or the upgrade method.

In an optional embodiment, upgrading the instance can be initiated by the central control device. For example, the central control device can monitor the version information of the mirror corresponding to each instance, and when a new version of the mirror is found, it can determine that the instance corresponding to the new version of the mirror needs to be upgraded; or, it can also monitor the running status of each instance, Life cycle and other information. When problems such as vulnerabilities, instability, insufficiency, excessive consumption of CPU or memory resources are found during the running of an instance, it can be determined that the instance with these problems needs to be upgraded.

In an optional embodiment, upgrading the instance may also be initiated by the service demander. For example, when an instance needs to be upgraded according to business requirements, the service demander can send upgrade description information to the central management and control device, and the upgrade description information includes instance filter conditions. In this case, step 21b includes: receiving upgrade description information sent by the service demander; and determining the instance to be upgraded from at least one instance according to the instance filter condition.

Among them, upgrading the instance to be upgraded mainly refers to: shutting down the instance to be upgraded, updating the instance to be upgraded according to the mirror image of the corresponding version (generally, the new version), and restarting the instance after the update. The image version required for upgrading the instance to be upgraded can be determined by the central management and control device. For example, the latest version of the corresponding image can be used as the image version required for the upgrade, or it can be specified by the service demander. Optionally, the service demander can carry the image version required for the upgrade in the upgrade description information and provide it to the central management and control device. For example, the upgrade description information can include "Upgrade from mirror version A to mirror version B for all or specified instances. "And other information. Based on this, upgrading the instance to be upgraded according to the upgrade strategy includes: according to the upgrade strategy, using the mirror corresponding to the mirror version to upgrade the instance to be upgraded.

Furthermore, in the case that the network system includes edge management and control equipment, according to the upgrade strategy, using the mirror corresponding to the mirror version to upgrade the instance to be upgraded can be: sending the upgrade strategy and the mirror corresponding to the mirror version to the edge management and control equipment in the network system , So that the edge management and control device uses the mirror corresponding to the mirror version to upgrade the instance to be upgraded according to the upgrade strategy.

As shown in Figure 2c, the process of the central management and control device for instance migration management and control includes the following steps:

21c. Determine the instance to be migrated from at least one instance, and the instance to be migrated belongs to the first edge cloud node;

22c. Determine that the first edge cloud node meets the intra-node migration condition; if the judgment result is yes, that is, the first edge cloud node meets the intra-node migration condition, go to step 23c; if the judgment result is no, the first edge cloud node does not meet the For intra-node migration conditions, go to step 24c.

23c. The instance to be migrated is migrated within the edge cloud node.

24c. Perform cross-edge cloud node migration for the instance to be migrated.

In some cases, the instance needs to be migrated. For example, in the case that the entire edge cloud node is faulty or unavailable, the instances in the edge cloud node need to be migrated to other edge cloud nodes. For another example, in the case of a failure or downtime of a physical machine hosting an instance, the instance on the physical machine needs to be migrated to another physical machine. For another example, it may be necessary to migrate one or some instances from one edge cloud node to other edge cloud nodes due to business needs. For another example, when resources need to be merged, one or some instances need to be migrated.

Under the control of the central control device, instances in the edge cloud node can be migrated. The central control device determines the instance to be migrated from at least one instance. There may be one or more instances to be migrated; if there are multiple instances to be migrated, the multiple instances to be migrated can be deployed in the same edge cloud node or in different edge cloud nodes.

Optionally, the central management and control device may monitor the state of at least one instance deployed in at least one edge cloud node, and obtain a failed instance and/or an instance in which a specified event occurs during operation as an instance to be migrated according to the state of at least one instance. Among them, a failed instance refers to an instance that cannot operate normally, for example, it can be an instance on a physical machine where the downtime occurs, or an instance that itself is down. The designated event mainly refers to some events that the instance can still run normally after occurrence, which can be flexibly set according to application requirements, and there is no restriction on this. For example, the specified event can be some early warning or alarm events, etc. Although some early warning or alarm events occur, the instance does not produce actual problems and can still run (that is, no failure), but there are hidden dangers of failure and can be migrated before failure. . In addition, the central control equipment maintains the information of each edge cloud node and the information of each instance deployed in each edge cloud node. Based on this, the edge cloud node to which the instance to be migrated belongs can be determined. For ease of description and distinction, the instance to be migrated The edge cloud node to which it belongs before the migration is recorded as the first edge cloud node.

Optionally, the central management and control device may determine the instance to be migrated from at least one instance according to resource merging requirements, and then migrate the instance to be migrated. Among them, resource merging is mainly the process of integrating resource fragments through instance migration. After integration, the resource fragments in edge cloud nodes will be reduced or even nonexistent, which is conducive to improving resource utilization in edge cloud nodes. It is worth noting that resource merging requirements can be system-level or node-level. System-level resource merging refers to the integration of resource fragments in the entire network system from the perspective of the entire network system through instance migration; node-level resource merging refers to the perspective of edge cloud nodes and the use of instance migration The resource fragments in the node are integrated.

Optionally, the resource consolidation requirement may be provided by the service demander. For example, when a service demander needs to deploy a new instance, if the available resources on each resource device in the edge cloud node it serves are not enough to carry the new instance, the instance in the edge cloud node can be migrated to implement resources Integration to provide sufficient resources for new instances. Or, the resource consolidation requirement can also be the regular behavior of the central control equipment. For example, the central management and control equipment regularly performs resource fragmentation checks. When the fragmentation rate reaches a certain threshold and instance migration can be performed, the resource fragmentation in each edge cloud node is integrated to improve resource utilization in the edge cloud node.

Regardless of the application scenario, after determining the instance to be migrated, the central control device can determine whether the first edge cloud node to which the instance to be migrated belongs meets the intra-node migration conditions; if the first edge cloud node meets the intra-node migration conditions, it will be treated The migration instance performs intra-edge cloud node migration; if the first edge cloud node does not meet the intra-node migration condition, the migration instance to be migrated is migrated across edge cloud nodes.

Optionally, the central management and control device may determine whether the first edge cloud node is currently available; if the first edge cloud node is currently available, determine whether the available resources of the first edge cloud node are sufficient to carry the instance to be migrated; The available resources of the edge cloud node are sufficient to carry the instance to be migrated, and it is determined that the first edge cloud node meets the migration conditions within the node; if the first edge cloud node is currently in an unavailable state, or the available resources of the first edge cloud node are insufficient to carry the instance to be migrated For example, it is determined that the first edge cloud node does not meet the intra-node migration condition. In the embodiments of the present application, the migration of instances is divided into two types: intra-node migration and cross-node migration. Among them, the available resources of the first edge cloud node mainly refer to the available resources on each resource device in the first edge cloud node; accordingly, judging whether the available resources of the first edge cloud node are sufficient to carry the instance to be migrated mainly refers to judging Whether there is a resource device with sufficient resources available in the first edge cloud node to carry the instance to be migrated.

For cross-node migration: the central control device can select a second edge cloud node from at least one edge cloud node, the second edge cloud node is different from the first edge cloud node, and the available resources in the second edge cloud node are sufficient to carry the instance to be migrated , That is, sufficient resources; migrate the instance to be migrated to the second edge cloud node, and send the attribute information of the instance to be migrated in the second edge cloud node to the service demander, so that the service demander can target The instance to be migrated performs business scheduling. Among them, the attribute information of the instance to be migrated in the second edge cloud node means that after the instance to be migrated is migrated to the second edge cloud node, an external (for example, a service demander or a third party authorized by the service demander) conducts an operation on the instance to be migrated. Information required for service scheduling may include, but is not limited to, for example, information such as the area where the second edge cloud node is located, operator information, and/or public network IP.

Method 1: According to the distance between other edge cloud nodes and the first edge cloud node, select the edge cloud node whose distance from the first edge cloud node is less than the set distance threshold, or select the closest distance to the first edge cloud node An edge cloud node, or an edge cloud node arbitrarily selected from the N edge cloud nodes closest to the first edge cloud node as the second edge cloud node. In Manner 1, the second edge cloud node is closest or relatively close to the first edge cloud node, which can save data transmission time and help improve migration efficiency.

Method 2: You can select edge cloud nodes with relatively sufficient bandwidth resources according to the bandwidth resources of other edge cloud nodes. For example, select the edge cloud node with the largest bandwidth resource, or select the bandwidth resource greater than the set bandwidth threshold, or select the bandwidth utilization rate lower The edge cloud node serves as the second edge cloud node. In method 2, the bandwidth resources of the second edge cloud node are sufficient, which can increase the data transmission rate and help improve migration efficiency.

Optionally, when migrating the instance to be migrated to the second edge cloud node, the central management and control device may reserve or allocate resources for the instance to be migrated in the second edge cloud node according to the resource requirements of the instance to be migrated; After the reservation or allocation is successful, the instance to be migrated is migrated to the resources reserved or allocated in the second edge cloud node. For example, the resource requirements of the instances to be migrated can be combined to determine the type of resources, the amount of resources and/or the performance requirements of the resource equipment required by the instances to be migrated, and resource reservation or allocation can be performed in the second edge cloud node based on this information , Which can provide resource guarantee for successful instance migration.

Optionally, if the instance to be migrated is a failed instance, that is, an instance that is not functioning normally, the central control device can also notify the service demander of the migration event, and the service demander can make appropriate response actions, such as updating the The information of the instance in the service demander, or the disaster recovery response to the downtime during the instance migration. Further, in the process of notifying the migration event, the attribute information of the instance to be migrated in the second edge cloud node may be provided to the service demander. Of course, after the instance to be migrated is successfully migrated to the second edge cloud node, the attribute information of the instance to be migrated in the second edge cloud node may be provided to the service demander.

Optionally, if the instance to be migrated is an instance in which a specified event occurs during operation, that is, an instance that can run normally despite the occurrence of a specified event, the central control device can also send a migration request to the service demander for the service demander to combine The business situation on the instance to be migrated determines the migration strategy for the instance to be migrated; the migration strategy sent by the service demander is received, and the instance to be migrated is migrated to the second edge cloud node according to the migration strategy. The migration strategy mainly includes at least one information of whether to migrate, migration time, and migration mode.

Further optionally, the central management and control device may send the attribute information of the instance to be migrated in the second edge cloud node together with the migration request to the service demander. Alternatively, after the instance to be migrated is successfully migrated to the second edge cloud node, the attribute information of the instance to be migrated in the second edge cloud node may be provided to the service demander.

Further optionally, if the instance to be migrated is an instance that has a specified event but can still run normally, during the migration process, the instance to be migrated continues to run on the first edge cloud node, so that the business during the migration process can continue to be scheduled to Ensure business continuity on the instances to be migrated in the first edge cloud node. After successfully migrating the instance to be migrated to the second edge cloud node, and the service demander ensures that all new business requests are scheduled to the migrated second edge cloud node, and the business requests in the first edge cloud node are gradually reduced In the end, there is no new service request, that is, when there are no longer any service requests on the instance to be migrated running in the first edge cloud node, the central management and control device can release the instance to be migrated in the first edge cloud node. Optionally, after determining that the service requester no longer has any service requests on the instance to be migrated running on the first edge cloud node, and there is neither an inventory service request nor an incremental service request, it may send a release notice to the central control device ; The central control device receives the release notification sent by the service demander, and releases the instance to be migrated running in the first edge cloud node according to the release notification. Further, the central management and control device may also synchronize the running state of the instance to be migrated running in the first edge cloud node to the instance to be migrated in the second edge cloud node.

Further, regardless of the instance to be migrated, migrating the instance to be migrated to the second edge cloud node is mainly to control the corresponding resource equipment in the second edge cloud node to reserve or allocate according to the mirror or snapshot corresponding to the instance to be migrated The process of creating an instance to be migrated on the resource.

Further, in the case that the edge management and control device is included in the network system, the central management and control device may determine the scheduled resource information in the second edge cloud node according to the resource requirements of the instance to be migrated, and send the resource information to the edge management and control device, and The edge management and control device controls the corresponding resource device in the second edge cloud node to reserve or allocate resources for the instance to be migrated according to the resource information. Then, the central management and control device can send a migration instruction to the edge management and control device. The migration instruction instructs the edge management and control device to obtain the image or instance snapshot corresponding to the instance to be migrated and provide it to the corresponding resource device in the second edge cloud node for the second edge cloud node According to the image or instance snapshot, the corresponding resource device creates an instance to be migrated on the reserved or allocated resources. Further, if an edge management and control device is deployed in the second edge cloud node, the central management and control device may send a migration instruction to the edge management and control device in the second edge cloud node to instruct the edge management and control device in the second edge cloud node to obtain the instance to be migrated The corresponding image or snapshot is provided to the corresponding resource device in the second edge cloud node for the corresponding resource device in the second edge cloud node to create an instance to be migrated on the reserved or allocated resources according to the image or snapshot.

In the method embodiment of the present application, under the control of the central management and control device, the instance in the edge cloud node can provide cloud computing services to the service demander, achieving the purpose of providing services to users by using the resources in the edge cloud node, so that " It has become a reality to place cloud computing in edge cloud nodes closer to the terminal, which will help reduce response delays, reduce the pressure on the central cloud or traditional cloud computing platforms corresponding to edge cloud nodes, and reduce bandwidth costs.

It should be noted that some processes described in the above embodiments and drawings include multiple operations appearing in a specific order, but it should be clearly understood that these operations may not be performed in the order in which they appear in this article. Or execute in parallel. The sequence numbers of operations, such as 21a, 22a, etc., are only used to distinguish different operations. The sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed sequentially or in parallel. It should be noted that the descriptions of "first" and "second" in this article are used to distinguish different messages, devices, modules, etc., and do not represent a sequence, nor do they limit the "first" and "second" Are different types.

FIG. 3 is a schematic structural diagram of a central management and control device provided by an exemplary embodiment of this application. As shown in FIG. 3, the central management and control device includes: a memory 31 and a processor 32.

The memory 31 is used to store computer programs, and can be configured to store various other data to support operations on the central control device. Examples of these data include instructions, messages, pictures, videos, etc. used to operate any application or method on the central control device.

The processor 32 is coupled with the memory 31 and is configured to execute the computer program in the memory 31 to determine at least one instance deployed in at least one edge cloud node in the network system, and the at least one instance can provide the cloud for the service demander Computing services: At least one instance is managed and controlled so that at least one instance provides cloud computing services for the service demander.

Optionally, the management and control of at least one instance includes: at least one of upgrade, migration, shutdown, restart, and release.

In an optional embodiment, as shown in FIG. 3, the central management and control device further includes: a communication component 33. Based on this, when the processor 32 upgrades at least one instance, it is specifically configured to: determine the instance to be upgraded from the at least one instance; send an upgrade request to the service demander through the communication component 33, so that the service demander can combine the instance to be upgraded The above business situation determines the upgrade strategy for the instance to be upgraded; the communication component 33 receives the upgrade strategy returned by the service demander, and upgrades the instance to be upgraded according to the upgrade strategy.

Further, when the processor 32 determines the instance to be upgraded from at least one instance, it is specifically configured to: receive the upgrade description information sent by the service demander through the communication component 33, where the upgrade description information includes instance filter conditions; according to the instance filter conditions, from at least In one instance, the instance to be upgraded is determined.

Further, the upgrade description information also includes: the image version required for the upgrade. Then, when the processor 32 upgrades the instance to be upgraded according to the upgrade strategy, it is specifically configured to: according to the upgrade strategy, use the mirror corresponding to the mirror version to upgrade the instance to be upgraded.

Furthermore, when the processor 32 uses the mirror corresponding to the mirror version to upgrade the instance to be upgraded according to the upgrade strategy, it is specifically used to: send the mirror corresponding to the upgrade strategy and the mirror version to the edge management and control device in the network system for the edge According to the upgrade strategy, the control equipment uses the mirror corresponding to the mirror version to upgrade the instance to be upgraded. Wherein, the upgrade strategy includes but is not limited to: at least one piece of information in whether to upgrade, upgrade time, and upgrade method.

In an optional embodiment, when the processor 32 migrates at least one instance, it is specifically configured to: determine the instance to be migrated from the at least one instance, and the edge cloud node to which the instance to be migrated belongs is recorded as the first edge cloud node; If the first edge cloud node meets the intra-node migration condition, the instance to be migrated is migrated within the edge cloud node; if the first edge cloud node does not meet the intra-node migration condition, the instance to be migrated is migrated across edge cloud nodes.

Optionally, when the processor 32 determines the instance to be migrated from the at least one instance, it is specifically configured to: according to the state of the at least one instance, use the failed instance and/or the instance in which a specified event occurs during operation as the instance to be migrated .

Optionally, when the processor 32 determines the instance to be migrated from the at least one instance, it is specifically configured to: determine the instance to be migrated from the at least one instance according to resource merging requirements.

Further, when the processor 32 determines the instance to be migrated according to the resource merging demand, it is specifically configured to: determine the first edge cloud node that needs resource merging according to the resource merging demand; combine the remaining resources on each resource device in the first edge cloud node The available resources of and the resources required by each instance in the first edge cloud node determine the instance to be migrated.

Optionally, the processor 32 is further configured to: determine whether the first edge cloud node is in an available state; if the first edge cloud node is in an available state, determine whether the available resources of the first edge cloud node are sufficient to carry the instance to be migrated; If the available resources of an edge cloud node are sufficient to carry the instance to be migrated, it is determined that the first edge cloud node meets the migration conditions within the node; if the first edge cloud node is in an unavailable state, or the available resources of the first edge cloud node are insufficient to carry the instance For the migration instance, it is determined that the first edge cloud node does not meet the intra-node migration condition.

Optionally, the processor 32 is specifically configured to select a second edge cloud node from at least one edge cloud node when the instance to be migrated is migrated across edge cloud nodes, where the second edge cloud node is different from the first edge cloud node; Migrate the instance to be migrated to the second edge cloud node, and send the attribute information of the instance to be migrated in the second edge cloud node to the service demander, so that the service demander can perform business scheduling for the instance to be migrated based on the attribute information .

Optionally, if the instance to be migrated is an instance in which a specified event occurs during operation, that is, an example in which a specified event occurs but can still run normally, then when the processor 32 migrates the instance to be migrated to the second edge cloud node, It is specifically used for: sending a migration request to the service demander through the communication component 33, so that the service demander can determine the migration strategy for the instance to be migrated in combination with the business situation on the instance to be migrated; receiving the migration strategy sent by the service demander through the communication component 33, According to the migration strategy, the instance to be migrated is migrated to the second edge cloud node.

Optionally, when the processor 32 migrates the instance to be migrated to the second edge cloud node, it is specifically configured to: according to the resource requirements of the instance to be migrated, control the corresponding resource device in the second edge cloud node to perform resources for the instance to be migrated Reservation or allocation: After the resource reservation or allocation is successful, the instance to be migrated is migrated to the resource reserved or allocated by the corresponding resource device in the second edge cloud node.

Optionally, when the processor 32 migrates the instance to be migrated to the resource reserved or allocated by the corresponding resource device in the second edge cloud node, it is specifically configured to: control the corresponding resource device in the second edge cloud node according to the instance to be migrated The corresponding image or snapshot creates an instance to be migrated on the reserved or allocated resources.

Further optionally, the processor 32 is specifically configured to send a migration instruction to the edge management and control device in the network system, and the migration instruction instructs the edge management and control device to obtain the image or instance snapshot corresponding to the instance to be migrated and provide it to the second The corresponding resource device in the edge cloud node allows the corresponding resource device to create the instance to be migrated on the reserved or allocated resources.

Optionally, the processor 32 is further configured to: receive a release notification sent by the service demander through the communication component 33, and release the instance to be migrated running in the first edge cloud node according to the release notification; wherein, during the migration process , The instance to be migrated continues to run in the first edge cloud node; wherein the release notification is sent after the service demander determines that there is no longer any service request on the instance to be migrated running in the first edge cloud node.

Furthermore, as shown in FIG. 3, the central management and control device further includes: a display 34, a power supply component 35, an audio component 36 and other components. Only some of the components are schematically shown in FIG. 3, which does not mean that the central control equipment only includes the components shown in FIG. In addition, the components in the dashed box in Figure 3 are optional components, which may be determined by the implementation of the central control equipment. If the central management and control device is a server-shaped device, it may optionally not include the display 34 and the audio component 36; if the central management and control device is a terminal device-type device, it may optionally include the display 34 and the audio component 36.

Correspondingly, an embodiment of the present application also provides a computer-readable storage medium storing a computer program. When the computer program is executed by one or more processors, the one or more processors can implement the above method in the above-mentioned method embodiments. Steps or operations performed by the equipment.

The communication component in FIG. 3 is configured to facilitate wired or wireless communication between the device where the communication component is located and other devices. The device where the communication component is located can access a wireless network based on communication standards, such as WiFi, 2G or 3G, or a combination of them. In an exemplary embodiment, the communication component receives a broadcast signal or broadcast related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component may further include a near field communication (NFC) module, radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, and Bluetooth (BT) technology Wait.

The display in FIG. 3 described above includes a screen, and the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touch, sliding, and gestures on the touch panel. The touch sensor may not only sense the boundary of a touch or slide action, but also detect the duration and pressure related to the touch or slide operation.

The power supply components in Figure 3 above provide power for various components of the equipment where the power supply components are located. The power supply component may include a power management system, one or more power supplies, and other components associated with the generation, management, and distribution of power for the equipment where the power supply component is located.

The audio component in FIG. 3 may be configured to output and/or input audio signals. For example, the audio component includes a microphone (MIC). When the device where the audio component is located is in an operating mode, such as call mode, recording mode, and voice recognition mode, the microphone is configured to receive external audio signals. The received audio signal can be further stored in a memory or sent via a communication component. In some embodiments, the audio component further includes a speaker for outputting audio signals.

Those skilled in the art should understand that the embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

The present invention is described with reference to flowcharts and/or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and/or block in the flowchart and/or block diagram, and the combination of processes and/or blocks in the flowchart and/or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing equipment to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing equipment are generated It is a device that realizes the functions specified in one process or multiple processes in the flowchart and/or one block or multiple blocks in the block diagram.

These computer program instructions can also be stored in a computer-readable memory that can guide a computer or other programmable data processing equipment to work in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture including the instruction device. The device implements the functions specified in one process or multiple processes in the flowchart and/or one block or multiple blocks in the block diagram.

These computer program instructions can also be loaded on a computer or other programmable data processing equipment, so that a series of operation steps are executed on the computer or other programmable equipment to produce computer-implemented processing, so as to execute on the computer or other programmable equipment. The instructions provide steps for implementing functions specified in a flow or multiple flows in the flowchart and/or a block or multiple blocks in the block diagram.

In a typical configuration, the computing device includes one or more processors (CPU), input/output interfaces, network interfaces, and memory.

The memory may include non-permanent memory in computer readable media, random access memory (RAM) and/or non-volatile memory, such as read-only memory (ROM) or flash memory (flash RAM). Memory is an example of computer readable media.

Computer-readable media include permanent and non-permanent, removable and non-removable media, and information storage can be realized by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, CD-ROM, digital versatile disc (DVD) or other optical storage, Magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media can be used to store information that can be accessed by computing devices. According to the definition in this article, computer-readable media does not include transitory media, such as modulated data signals and carrier waves.

It should also be noted that the terms "include", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, product or equipment including a series of elements not only includes those elements, but also includes Other elements that are not explicitly listed, or include elements inherent to this process, method, commodity, or equipment. If there are no more restrictions, the element defined by the sentence "including a..." does not exclude the existence of other identical elements in the process, method, commodity, or equipment that includes the element.

The above descriptions are only examples of this application and are not used to limit this application. For those skilled in the art, this application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this application shall be included in the scope of the claims of this application.

Claims

An example management and control method, characterized in that it includes:

Determine at least one instance deployed in at least one edge cloud node in the network system, where the at least one instance can provide cloud computing services for the service demander;

The at least one instance is managed and controlled, so that the at least one instance provides cloud computing services for the service demander.
The method according to claim 1, wherein the management and control of the at least one instance includes at least one of upgrade, migration, shutdown, restart, and release.
The method of claim 2, wherein upgrading the at least one instance comprises:

Determine the instance to be upgraded from the at least one instance;

Sending an upgrade request to the service demander, so that the service demander determines an upgrade strategy for the instance to be upgraded in combination with the business situation on the instance to be upgraded;

The upgrade strategy returned by the service demander is received, and the instance to be upgraded is upgraded according to the upgrade strategy.
The method according to claim 3, wherein determining the instance to be upgraded from the at least one instance comprises:

Receiving upgrade description information sent by the service demander, where the upgrade description information includes instance filter conditions;

According to the instance filter condition, the instance to be upgraded is determined from the at least one instance.
The method according to claim 4, wherein the upgrade description information further includes: an image version required for the upgrade; then, upgrading the instance to be upgraded according to the upgrade strategy includes:

According to the upgrade strategy, the instance to be upgraded is upgraded using the image corresponding to the image version.
The method according to claim 5, wherein, according to the upgrade strategy, upgrading the instance to be upgraded using the image corresponding to the image version comprises:

The upgrade strategy and the mirror corresponding to the mirror version are provided to the edge management and control device in the network system, so that the edge management and control device uses the mirror corresponding to the mirror version to perform the processing on the waiting device according to the upgrade strategy. Upgrade the instance to upgrade.
The method according to any one of claims 3-6, wherein the upgrade strategy includes at least one of information of whether to upgrade, upgrade time, and upgrade method.
The method according to claim 2, wherein migrating the at least one instance comprises:

Determine an instance to be migrated from the at least one instance, where the instance to be migrated belongs to the first edge cloud node;

If the first edge cloud node meets the intra-node migration condition, perform intra-edge cloud node migration on the instance to be migrated;

If the first edge cloud node does not meet the intra-node migration condition, perform cross-edge cloud node migration on the instance to be migrated.
The method according to claim 8, wherein determining the instance to be migrated from the at least one instance comprises:

According to the state of the at least one instance, use the failed instance and/or the instance in which a specified event occurs during operation as the instance to be migrated; or

According to resource merging requirements, an instance to be migrated is determined from at least one instance.
The method according to claim 9, characterized in that, determining the instance to be migrated from at least one instance according to resource merging requirements comprises:

According to resource merging requirements, determine the first edge cloud node that needs resource merging;

Combining the remaining available resources on each resource device in the first edge cloud node and the resources required by each instance in the first edge cloud node, determine the instance to be migrated.
The method according to claim 8, further comprising:

Determine whether the first edge cloud node is in an available state;

If the first edge cloud node is in an available state, determining whether the available resources of the first edge cloud node are sufficient to carry the instance to be migrated;

If the available resources of the first edge cloud node are sufficient to carry the instance to be migrated, determining that the first edge cloud node meets the intra-node migration condition;

If the first edge cloud node is in an unavailable state, or the available resources of the first edge cloud node are insufficient to carry the instance to be migrated, it is determined that the first edge cloud node does not meet the intra-node migration condition.
The method according to any one of claims 8-11, wherein the migration of the instance to be migrated across edge cloud nodes comprises:

Selecting a second edge cloud node from the at least one edge cloud node, where the second edge cloud node is different from the first edge cloud node;

Migrate the instance to be migrated to the second edge cloud node, and send the attribute information of the instance to be migrated in the second edge cloud node to the service demander for the service demand The party performs service scheduling for the instance to be migrated based on the attribute information.
The method according to claim 12, wherein if the instance to be migrated is an instance in which a specified event occurs during operation, migrating the instance to be migrated to the second edge cloud node comprises:

Sending a migration request to the service demander, so that the service demander determines a migration strategy for the instance to be migrated in combination with the business situation on the instance to be migrated;

Receive the migration strategy sent by the service demander, and migrate the instance to be migrated to the second edge cloud node according to the migration strategy.
The method according to claim 12, wherein migrating the instance to be migrated to the second edge cloud node comprises:

According to the resource requirements of the instance to be migrated, controlling the corresponding resource device in the second edge cloud node to reserve or allocate resources for the instance to be migrated;

After the resource reservation or allocation is successful, the instance to be migrated is migrated to the resource reserved or allocated by the corresponding resource device in the second edge cloud node.
The method according to claim 14, wherein the migrating the instance to be migrated to the resource reserved or allocated by the corresponding resource device in the second edge cloud node comprises:

Control the corresponding resource device in the second edge cloud node to create the instance to be migrated on the reserved or allocated resources according to the image or instance snapshot corresponding to the instance to be migrated.
The method according to claim 15, wherein the corresponding resource device in the second edge cloud node is controlled to create the to-be-migrated resource on the reserved or allocated resource according to the mirror image or instance snapshot corresponding to the to-be-migrated instance. Examples include:

Send a migration instruction to the edge management and control device in the network system, where the migration instruction instructs the edge management and control device to obtain the image or instance snapshot corresponding to the instance to be migrated and provide it to the corresponding resource device in the second edge cloud node , So that the corresponding resource device can create the instance to be migrated on the reserved or allocated resource.
The method according to claim 12, further comprising:

Receive a release notification sent by the service demander, and release the instance to be migrated running in the first edge cloud node according to the release notification; wherein, during the migration process, the instance to be migrated continues Running in the first edge cloud node;

Wherein, the release notification is sent by the service demander after determining that there is no longer any service request on the instance to be migrated running in the first edge cloud node.
A network system, characterized by comprising: a central management and control device, and at least one edge cloud node;

At least one instance is deployed in the at least one edge cloud node, and the at least one instance can provide cloud computing services for service demanders;

The central management and control device is configured to manage and control the at least one instance, so that the at least one instance provides cloud computing services for the service demander.
The network system according to claim 18, wherein the management and control of the at least one instance by the central management and control device includes at least one of upgrade, migration, shutdown, restart, and release.
The network system according to claim 19, wherein, when the central management and control device upgrades the at least one instance, it is specifically configured to:

Determine the instance to be upgraded from the at least one instance;

Sending an upgrade request to the service demander, so that the service demander determines an upgrade strategy for the instance to be upgraded in combination with the business situation on the instance to be upgraded;

The upgrade strategy sent by the service demander is received, and the instance to be upgraded is upgraded according to the upgrade strategy.
The network system according to claim 19, wherein, when the central management and control device migrates the at least one instance, it is specifically configured to:

Determine the instance to be migrated from the at least one instance, and determine the first edge cloud node to which the instance to be migrated belongs;

If the first edge cloud node meets the intra-node migration condition, perform intra-edge cloud node migration on the instance to be migrated;

If the first edge cloud node does not meet the intra-node migration condition, perform cross-edge cloud node migration on the instance to be migrated.
The network system according to claim 21, wherein the central management and control device is specifically configured to: when migrating the instance to be migrated across edge cloud nodes:

Selecting a second edge cloud node from the at least one edge cloud node, where the second edge cloud node is different from the first edge cloud node;

Migrate the instance to be migrated to the second edge cloud node, and send the attribute information of the instance to be migrated in the second edge cloud node to the service demander for the service demand The party performs service scheduling for the instance to be migrated based on the attribute information.
The network system according to any one of claims 18-22, further comprising: edge management and control equipment;

The edge management and control device is used to cooperate with the central management and control device to manage and control the at least one instance.
A central management and control device, which is characterized by comprising: a memory and a processor;

The memory is configured to store a computer program; when the computer program is executed by the processor, the processor is caused to implement the steps in the method of any one of claims 1-17.
A computer-readable storage medium storing a computer program, wherein when the computer program is executed by one or more processors, the one or more processors are caused to implement any one of claims 1-17 The steps in the method.