Edge device dynamic data center configuration method based on registration center

Abstract

The invention relates to an edge device dynamic data center configuration method based on a registration center, and the method comprises the following steps: 1, initializing the registration center: deploying the registration center, defining metadata, and configuring device information and data center information, 2, registering the edge device, and reporting the state: actively registering the registration center when the device is started, step 3, configuring a multi-dimensional intelligent strategy engine; step 4, dynamic routing decision and load balancing; 5, performing health examination and adaptive updating; step 6, performing automatic expansion and zero configuration expansion; according to the scheme, the transmission delay is remarkably reduced, the real-time performance is improved, and accurate geographic position matching is achieved; and the device dynamically obtains the nearest data center address through the registration center.

Description

TECHNICAL FIELD

[0001] The application relates to a configuration method, in particular to an edge device dynamic data center configuration method based on a registration center, and belongs to the technical field of artificial intelligence application. BACKGROUND

[0002] With the rapid development of Internet of Things (IoT), a large number of devices (such as sensors, intelligent terminals, industrial devices, etc.) are constantly connected to the network, and the frequency and scale of data reporting are growing exponentially. For example, an Internet of Things system in a smart city may contain millions of devices, generating PB-level data every day. However, the traditional Internet of Things architecture usually adopts a centralized design, with all device data reported to a single data center (or a small number of fixed data centers), resulting in the following problems:

[0003] Single data center processing capacity bottleneck: The computing, storage and network resources of a single data center are limited, and it cannot handle a large number of concurrent requests in real time, resulting in data backlog and delay.

[0004] Geographical location limitation: The physical distance between devices and data centers is far (such as devices located in remote areas and data centers located in centralized cloud areas), resulting in high data transmission delay and affecting real-time scenarios (such as autonomous driving and industrial control).

[0005] Insufficient scalability: When the number of devices increases rapidly, the data center needs to be expanded frequently, but the horizontal expansion of the traditional architecture relies on manual intervention and cannot dynamically adapt to business fluctuations.

[0006] The existing technical solutions are as follows:

[0007] Improve single data center capacity through load balancing,

[0008] Traditional solutions usually use load balancing technology to enhance the processing capacity of a single data center:

[0009] Local load balancing: Deploy load balancers within a single data center to distribute device requests to a backend server cluster, and disperse traffic through algorithms such as round robin and weighted round robin.

[0010] Session persistence and health check: Ensure request continuity (such as IP hashing) and exclude faulty nodes to improve availability. However, the following problems still exist:

[0011] Geographical location mismatch: Devices may be assigned to data centers far from their physical locations, resulting in high latency. For example, devices located in Southeast Asia may be incorrectly configured or have a load balancing strategy defect, reporting data to a European data center, increasing transmission time.

[0012] Data center-level failures affect the entire system: If a single data center hosts a large amount of device traffic, its failure will directly cause the corresponding regional services to become unavailable. For example, when a certain city's data center goes down, the intelligent transportation system in that region may also be paralyzed.

[0013] Policy adjustment lag: Traditional load balancing relies on pre-set configurations and cannot real-time perceive changes in device geographic location or data center load. For example, when a certain data center is overloaded due to traffic surge, the proxy server may still continuously forward requests, exacerbating the risk of collapse.

[0014] Limited scalability: When new edge nodes or data centers need to be added, the proxy configuration or device-side routing logic needs to be manually updated, resulting in high operational complexity.

[0015] 1. Resource waste caused by the lack of dynamic routing and static configuration: In existing IoT systems, edge devices usually use fixed IP addresses or static load balancing strategies to connect to data centers, and cannot dynamically adjust routing paths according to real-time environmental changes (such as device geographic location, data center load, network status, etc.). For example, when a certain data center is overloaded due to sudden traffic surge, devices continue to send requests to it, causing service response delay or even collapse. Traditional solutions rely on pre-set rules or static configurations of proxy servers, lacking the ability to perceive dynamic topology, resulting in unreasonable resource allocation and low overall system efficiency.

[0016] 2. Transmission delay caused by insufficient matching of geographic location and data center,

[0017] Inaccurate geographic location information: The IP geographic location library of the proxy server may lag behind actual network topology changes (such as device movement or data center migration), causing devices to connect to nodes far from their physical locations.

[0018] Inability to dynamically optimize paths: Devices cannot real-time perceive changes in their own location (such as vehicle-mounted devices moving) or fluctuations in data center service capabilities, resulting in non-optimal transmission paths, especially in cross-regional or remote scenarios, where delay is significantly increased.

[0019] 3. Rigidity of load balancing strategies and insufficient global perception ability,

[0020] Existing load balancing technologies (such as HAProxy / Nginx) are mainly implemented at the single data center or proxy layer, lacking real-time monitoring and dynamic adjustment capabilities for global data center resource status:

[0021] Local load balancing, global imbalance: Load balancing within a single data center cannot solve the problem of resource allocation across data centers. For example, when a certain data center is overloaded, devices may still be incorrectly allocated traffic while other nodes are idle.

[0022] Static weight configuration: the weighted round robin strategy of the proxy server depends on the artificial preset weight, and cannot automatically adapt to the real-time performance (such as CPU usage, network bandwidth, and storage capacity) changes of the data center, resulting in low resource utilization.

[0023] 4. Data center level failure affects the whole world: when a single data center fails, the devices in the region cannot report data, and the failure will directly cause the corresponding regional service to be unavailable.

[0024] 5. The contradiction between scalability and operation and maintenance complexity,

[0025] The prior art is difficult to meet the large-scale dynamic expansion demand:

[0026] Coupling of device and data center configuration: when adding an edge node or a data center, the proxy configuration or the routing logic on the device side needs to be manually updated, which is high in operation and maintenance cost and prone to errors.

[0027] Multi-dimensional policy configuration difficulty: if multi-dimensional factors such as device type, business priority, and quality of service (QoS) are combined to formulate a routing policy (such as preferentially allocating low-delay nodes to medical devices), the traditional scheme lacks a flexible configuration framework, resulting in low policy update efficiency. SUMMARY

[0028] The present application is directed to the technical problems existing in the prior art, and provides an edge device dynamic data center configuration method based on a registration center, which significantly reduces transmission delay and improves real-time performance.

[0029] In order to achieve the above purpose, the technical scheme of the present application is as follows: an edge device dynamic data center configuration method based on a registration center, characterized in that the method comprises the following steps:

[0030] Step 1, registration center initialization: deploying a registration center, defining metadata, configuring device information and data center information,

[0031] Step 2, edge device registration and state reporting: the device actively registers with the registration center when starting, and dynamically reports the device state,

[0032] Step 3, multi-dimensional intelligent policy engine configuration: defining policies according to distance priority, load balancing, capability matching, and QoS priority, pushing policy templates through a configuration center (such as Nacos), and supporting the addition of dimensions (such as energy consumption and cost) to dynamically adjust the data center configuration,

[0033] Step 4, dynamic routing decision and load balancing: making decisions on the defined policies through multi-factor weighted scoring, and coordinating traffic distribution through the registration center, and automatically adjusting the score when the load exceeds the threshold,

[0034] Step 5, health check and adaptive update: Through real-time health monitoring, check the health status of edge nodes and data centers, when the link is unresponsive or response timeout, requery the registration center, real-time update configuration,

[0035] Step 6, automatic expansion and zero configuration expansion: after the new edge node is started, it automatically registers itself to the registration center; after the new data center is registered, the registration center automatically includes it in the routing strategy, realizing seamless expansion; new data nodes only need to update the registration center metadata, and the device automatically senses the new node through dynamic routing strategy without modifying the device end logic.

[0036] Among them, step 1 registration center initialization is as follows:

[0037] Deploy the registration center: choose a distributed registration center (such as Nacos, Consul, Zookeeper or self-developed service), deploy it in a high-availability cluster, and be responsible for maintaining the metadata of edge nodes and data centers;

[0038] Define the metadata structure: the registration center stores the following information:

[0039] Edge device information: device ID, geographic location (GPS coordinates or IP geographic location), device type (such as real-time monitoring, batch data reporting), business priority (such as medical level, industrial level, consumer level), resource status (CPU / memory / storage), network status (bandwidth, delay);

[0040] Data center information: data center ID, geographic location, current load (CPU / memory / bandwidth usage), available resource capacity, service capability (low latency processing capability, big data analysis capability), service quality level (QoS);

[0041] Health check configuration: set the heartbeat detection mechanism (such as sending heartbeat packet every 5 seconds), if 3 consecutive heartbeats are not received, mark the node as "offline".

[0042] Among them, step 2 edge device registration and state reporting is as follows,

[0043] Register when the device starts:

[0044] The device sends a registration request to the registration center, carrying its own metadata (such as geographic location, device type, business priority), the registration center verifies the device identity (through pre-shared key or certificate authentication), and after verification, writes the device information to the registration center,

[0045] Dynamic state reporting:

[0046] The device reports resource status and network status changes (such as CPU usage, bandwidth fluctuation) regularly (such as every 30 seconds).

[0047] If the device moves (such as a vehicle-mounted device), automatically update the geographic location information.

[0048] Among them, step 3, multi-dimensional intelligent strategy engine configuration, as follows,

[0049] Strategy definition and combination:

[0050] Distance priority: based on device GPS coordinates or IP geographic location, calculate the physical distance from each data center, and select the nearest data center first,

[0051] Load balancing: dynamically allocate weights according to the current load (CPU / memory / bandwidth usage) of the data center, to avoid overload,

[0052] Capability matching: according to the device business type (such as real-time monitoring requires low latency, batch reporting requires high throughput), select data centers with corresponding capabilities (such as edge nodes handle real-time tasks, central nodes handle big data analysis), QoS priority: high-priority devices (such as medical emergency devices) are preferentially allocated to low-latency, high-reliability data centers,

[0053] Dynamic adjustment of strategy:

[0054] Define strategy templates (such as by device type, business priority classification) through configuration center (such as Nacos, Consul KV);

[0055] When the strategy changes (such as adding a new strategy dimension), the configuration center pushes the update to the registry center, without the need to restart the service.

[0056] Among them, step 4, dynamic routing decision and load balancing, as follows,

[0057] Routing decision factors,

[0058] Multi-factor weighted scoring: calculate the comprehensive score for each data center (such as score = 0.4 × geographic location score + 0.3 × load score + 0.2 × delay score + 0.1 × QoS score), select the data center with the highest score, machine learning optimization: train the model (such as random forest or neural network) through historical data, predict the service response time under different routing strategies, and dynamically adjust the weights,

[0059] Global load balancing:

[0060] The registry center serves as the global decision layer, coordinating the allocation of traffic across data centers to avoid local overload,

[0061] When the load of a data center exceeds the threshold (such as CPU usage > 80%), automatically reduce its routing score and guide new traffic to other data centers.

[0062] Among them, step 5, health check and adaptive update, as follows,

[0063] Real-time health monitoring:

[0064] The registration center sends health check requests (such as HTTP probes or TCP connection tests) to all edge nodes and data centers regularly (such as every 5 seconds).

[0065] If the node does not respond or the response times out, it is marked as "abnormal" and a failover process is triggered, and event-driven updates are triggered:

[0066] Periodic updates: the device actively queries the registration center every 30 seconds to obtain the latest routing strategy,

[0067] Event-triggered updates: when events such as network interruption and load surge occur, the device immediately re-queries the registration center to update the configuration in real time,

[0068] Broadcast notification: the registration center broadcasts strategy changes through event notification mechanisms (such as WebSocket or message queue), and the device responds immediately.

[0069] Among them, step 6, automatic expansion and zero-configuration expansion, as follows,

[0070] Edge node automatic registration:

[0071] After the new edge node is started, it automatically registers its information (such as geographic location and service capability) to the registration center without human intervention,

[0072] Data center elastic expansion: monitor the overall system load, and when the global load exceeds the threshold, trigger the cloud platform API to automatically deploy a new data center instance. After the new data center is registered, the registration center automatically includes it in the routing strategy, achieving seamless expansion. Zero-configuration expansion: new nodes only need to update the registration center metadata, and the device automatically senses new nodes through dynamic routing strategy without modifying the device-side logic.

[0073] An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the program to implement the registration center-based edge device dynamic data center configuration method.

[0074] A computer-readable storage medium having computer instructions stored thereon, wherein the computer instructions are executed by a processor to implement the registration center-based edge device dynamic data center configuration method.

[0075] Compared with the prior art, the present application has the following advantages,

[0076] 1. Significantly reduce transmission delay, improve real-time performance, and accurately match geographical location: The device dynamically obtains the address of the nearest data center through the registration center. For example, a vehicle-mounted device can switch to the nearest data center node in real time during movement, reducing cross-regional transmission delay.

[0077] 2. Dynamic load balancing, optimized resource utilization, global resource awareness and allocation: The registration center monitors the load status (CPU, memory, network bandwidth) of all data centers in real time, dynamically adjusts the traffic distribution strategy, and avoids local overload or resource idling. For example, when a data center is overloaded due to sudden traffic, the system can automatically divert new requests to nodes with lower load. Adaptive weight update: without manual intervention, the system dynamically adjusts the weight based on real-time performance indicators (such as CPU usage and task queue length), ensuring that resource utilization is at a relatively high level.

[0078] 3. Enhanced high availability and fault tolerance, automatic failover: The device periodically or event-triggeredly queries the registration center to real-time perceive the data center status. If the current node fails, it can quickly switch to a backup node, reducing service interruption time to seconds.

[0079] 4. Multi-dimensional intelligent strategy supports complex scenario requirements, flexible strategy combination: Supports dynamic combination of custom strategies (such as distance priority, QoS priority, and business type matching) to meet differentiated requirements. For example, medical devices can be preferentially allocated to low-latency edge nodes. Batch data reporting tasks can be directed to data centers with large data analysis clusters. Strategy update zero configuration: When adding a new strategy or adjusting the weight, only the registration center configuration needs to be updated, without modifying the device-side logic.

[0080] 5. Elastic expansion and low-cost operation, zero-configuration expansion capability: When adding an edge node or data center, only the registration center metadata needs to be registered, and the device automatically perceives and connects, reducing human configuration errors. Reducing operation complexity: The configuration and maintenance of proxy servers and load balancing strategy adjustment in traditional solutions are simplified to centralized management of the registration center, reducing operation costs. BRIEF DESCRIPTION OF DRAWINGS

[0081] Figure 1 The figure is a schematic diagram of the overall process of the present application. DETAILED DESCRIPTION

[0082] To better understand the present application, the following detailed description of the embodiments will be made in conjunction with the accompanying drawings.

[0083] Example 1: Referring to Figure 1 A registration center-based edge device dynamic data center configuration method, the method comprising the following steps:

[0084] 1. Data center reports the geographic location, network address, network port, and list of provided functions of the data center through the registration interface provided by the registration center.

[0085] 2. The data center periodically reports the current data center load situation, such as the overall CPU load, memory usage, disk usage, and network traffic, through the state interface provided by the registration center.

[0086] 3. The policy configuration personnel can configure the allocation policy of the data center through the device policy configuration module of the registration center, including the following:

[0087] Edge device distance from data center: distance priority

[0088] Data center load situation: CPU, memory, storage, network load below what

[0089] Data center type preference provided: low-latency quasi-real-time data calculation, big data data analysis, AI computing capability, etc.

[0090] In addition to the above rules, the device can also be classified according to the reported application or other business rules, and these rules can also be freely combined to meet multiple sets of union and or sets.

[0091] 4. When deploying the edge device, configure the access address of the registration center.

[0092] 5. When the edge device starts, access the registration center through the pre-configured registration center address, provide device information, current geographic location, or latitude and longitude coordinates

[0093] 6. The registration center queries the list of data centers registered in the registration center according to the geographic location or latitude and longitude information of the device, and sorts them according to the distance.

[0094] 7. The registration center matches the data center reported by the device according to the definition in the allocation policy after obtaining the list of data centers according to the distance, and returns the configuration information of the data center to the edge device.

[0095] 8. After the edge device obtains the data center allocated by the registration center, it obtains the network address and network port of the data center from the configuration and establishes a connection with the allocated data center to report data.

[0096] 9. When the data center appears abnormal and cannot provide normal service, the registration center discovers that the data center cannot normally report the load state through the data center monitoring module, and removes the data center from the assignable list. Subsequent edge devices will not be allocated to the abnormal data center when requesting allocation of data center configuration,

[0097] 10. When the data center connected by the edge device is abnormal and cannot provide services, the edge device will repeat steps 5-8 above to reapply for allocation of a data center.

[0098] Embodiment 2: An edge device dynamic data center configuration method based on a registration center, the method comprising the following steps:

[0099] 1. Deploy the registration center, cluster deployment (at least 3 nodes), avoid single point of failure.

[0100] 2. Define metadata using JSON format, edge device information: {"device_id":"edge_device_001","location":{"lat":39.9042,"lng":116.4074},"device_type":"realtime_monitoring","priority_level":"high","resources":{"cpu":2.4,"memory":"4GB","storage":"50GB"},"network":{"bandwidth":"100Mbps","latency":"50ms"}}; data center information: {"dc_id":"dc_beijing","location":{"lat":39.9042,"lng":116.4074},"load":{"cpu_usage":40%,"mem_usage":35%,"bandwidth_usage":50%},"capacity":{"max_cpu":100,"max_mem":"512GB"},"capabilities":["low_latency","big_data"],"qos_level":"premium"}

[0101] 3. Configure health checks, such as: devices send heartbeat packets (HTTP / UDP) every 5 seconds, registration center records the last heartbeat time, if 3 consecutive heartbeats are not received (15 seconds timeout), mark the node as "offline".

[0102] 4. When the device starts, authenticate through pre-shared key or certificate (pre-installed in the device), sign the request content, and then send an HTTPS POST request to the registration center, for example: Body: {"device_id":"edge_device_001","metadata":{...}}

[0103] 5. The registration center verifies the request through shared key or certificate authentication, and when the verification is passed, records the device uploading metadata, and returns the registration result, such as the assigned data center (assigned_dc).

[0104] 6. The device reports resource status (CPU / memory / bandwidth) and network status (delay, bandwidth fluctuation) every 30 seconds, such as: Body: {"device_id":"edge_device_001","resources":{...},"network":{...}}

[0105] 7. The device updates the status update, and if the device detects a change in geographical location (such as GPS coordinate update), it immediately triggers the status report.

[0106] 8. Policy template configuration, define template configuration through yaml script, such as

[0107]

[0108] 9. Machine learning optimization, by collecting historical routing decision data (delay, success rate, load), using random forest to predict data center response time, dynamically adjusting the weight.

[0109] from sklearn.ensemble import RandomForestRegressor

[0110] model = RandomForestRegressor()

[0111] model.fit(X_train, y_train) # X_train: features (distance, load, etc.), y_train: response time;

[0112] 10. Decide the data center routing of the device through multi-factor weighted score: calculate the comprehensive score of each data center: "score = 0.4 * distance_score + 0.3 * load_score + 0.2 * latency_score + 0.1 * qos_score", example score logic: distance score "100-(actual distance / max distance)*100"; load score "100-load percentage".

[0113] 11. The registration center dynamically adjusts the weight according to the data center load (such as weighted round robin), when the load of a data center >80%, set its weight to 0, suspend the flow distribution.

[0114] 12. The registration center sends HTTP probes to nodes every 5 seconds, like: "curl -s --connect-timeout 2 http: / / <node_ip>:health_check", if the node does not respond or the response is timeout or returns an exception, mark the node as "offline", trigger the route strategy recalculation, and the device switches to the standby data center.

[0115] 13. The registration center uses WebSocket or message queue (such as Kafka) to push strategy change events, such as: {"event":"strategy_update","new_strategy":{...}}, and the device side listens to the event and immediately re-queries the registration center to get the latest route configuration after receiving the update.

[0116] It should be noted that the above embodiments are not intended to limit the scope of protection of the present application, and any equivalent transformations or substitutions made on the basis of the above technical solutions fall within the scope of protection of the claims of the present application.

Claims

1. A method for configuring a dynamic data center for edge devices based on a registration center, characterized in that: The method comprises the following steps: Step 1. Registration center initialization: deploy the registration center, define metadata, configure device information and data center information, Step 2: Edge device registration and status reporting: When the device starts, it actively registers with the registration center and dynamically reports the device status. Step 3. Multi-dimensional intelligent policy engine configuration: define policies based on distance priority, load balancing, capability matching, and QoS priority, push policy templates through the configuration center, and support new dimensions to dynamically adjust data center configuration. Step 4: Dynamic routing decision and load balancing: Make decisions on defined policies through multi-factor weighted scoring, coordinate traffic distribution through the registration center, and automatically adjust the score when the load exceeds the threshold. Step 5: Health check and adaptive update: Check the health status of edge nodes and data centers through real-time health monitoring. When the link is unresponsive or the response times out, re-query the registration center and update the configuration in real time. Step 6: Automatic expansion and zero-configuration expansion: After the new edge node is started, it automatically registers its own information with the registration center. After the new data center is registered, the registration center automatically incorporates it into the routing policy to achieve seamless expansion. Adding a new data node only requires updating the registration center metadata. The device automatically perceives the new node through the dynamic routing policy, without modifying the device-side logic.

2. The method for configuring a dynamic data center for edge devices based on a registration center according to claim 1, characterized in that: Step 1: Initialize the registration center, as follows: Deploy a registry: Choose a distributed registry, deploy it in a high-availability cluster, and maintain metadata for edge nodes and data centers. Define metadata structure: The registry stores the following information: Edge device information: device ID, geographic location, device type, service priority, resource status, and network status; Data center information: data center ID, geographic location, current load, available resource capacity, service capabilities, and quality of service (QoS); Health check configuration: Set up a heartbeat detection mechanism. If no heartbeat is received for three consecutive times, the node will be marked as "offline".

3. The method for configuring a dynamic data center for edge devices based on a registration center according to claim 1, characterized in that: Step 2: Edge device registration and status reporting, as follows: Register when the device starts: The device sends a registration request to the registration center, carrying its own metadata. The registration center verifies the device identity and writes the device information to the registration center after verification. Dynamic status reporting: Devices regularly report resource status and network status changes; If the device moves, the geographic location information is automatically updated.

4. The method for configuring a dynamic data center for edge devices based on a registration center according to claim 1, wherein: Step 3: Multi-dimensional intelligent strategy engine configuration, as follows: Strategy definition and combination: Distance priority: Based on the device's GPS coordinates or IP location, the physical distance to each data center is calculated, and the nearest data center is given priority. Load balancing: Dynamically assign weights based on the current load of the data center to avoid overload. Capability matching: Select a data center with corresponding capabilities based on the equipment business type. QoS priority: Prioritize high-priority devices to low-latency, high-reliability data centers. Dynamic adjustment of strategies: Define policy templates through the configuration center; When the policy changes, the configuration center pushes the update to the registration center without restarting the service.

5. The method for configuring a dynamic data center for edge devices based on a registration center according to claim 1, characterized in that: Step 4: Dynamic routing decision and load balancing, as follows: Routing decision factor, Multi-factor weighted scoring: Calculate a comprehensive score for each data center and select the data center with the highest score. Machine learning optimization: Use historical data to train the model, predict the service response time under different routing strategies, and dynamically adjust the weights. Global load balancing: The registration center acts as the global decision-making layer, coordinating traffic distribution across data centers to avoid local overloads. When the load of a data center exceeds the threshold, its routing score is automatically lowered and new traffic is directed to other data centers.

6. The method for configuring a dynamic data center for edge devices based on a registration center according to claim 1, characterized in that: Step 5: Health check and adaptive update, as follows: Real-time health monitoring: The registration center regularly sends health check requests to all edge nodes and data centers; if a node does not respond or the response times out, it is marked as "abnormal" and the failover process is triggered. Event-driven updates: Periodic update: The device actively queries the registration center every 30 seconds to obtain the latest routing strategy. Event-triggered update: When network interruption, load surge, or other events occur, the device immediately re-queries the registration center and updates the configuration in real time. Broadcast notification: The registration center broadcasts policy changes through the event notification mechanism, and the device responds immediately.

7. The method for configuring a dynamic data center for edge devices based on a registration center according to claim 1, characterized in that: Step 6: Automatic expansion and zero-configuration expansion, as follows: Automatic registration of edge nodes: After the new edge node is started, it will automatically register its own information with the registration center without manual intervention. Elastic expansion of data centers: Monitor the overall system load. When the global load exceeds the threshold, the cloud platform API is triggered to automatically deploy a new data center instance. After the new data center is registered, the registration center automatically incorporates it into the routing strategy to achieve seamless expansion. Zero-configuration expansion: Adding new nodes only requires updating the registration center metadata. The device automatically detects the new node through dynamic routing strategies, without modifying the device-side logic.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, it implements a registration center-based edge device dynamic data center configuration method as described in any one of claims 1 to 7 above.

9. A computer-readable storage medium having computer instructions stored thereon, characterized in that: When the computer instruction is executed by the processor, a registration center-based dynamic data center configuration method for edge devices is implemented as described in any one of claims 1-7.

Images