Intelligent edge scheduling method, system, device and medium in distributed cloud scenario

Abstract

This application discloses an intelligent edge scheduling method, system, device, and medium in a distributed cloud scenario. The method includes: a scheduling center issuing a service probe range instruction to at least one service region. The service region includes a regional scheduling center and multiple regional service nodes. The service probe range instruction is used to bind some regional service nodes from one service region to a regional scheduling center in another service region. The bound regional service nodes periodically report their service capabilities to their respective two regional scheduling centers. Based on the service capabilities reported by each regional service node, the regional scheduling center sorts the regional service nodes to generate a priority list. The priority list includes regional service nodes in its own service region and their bound regional service nodes. In response to a received service request, the regional scheduling center matches regional service nodes and the requesting terminal according to the priority list. Through this method, this application can shorten request response time and improve service quality.

Description

Technical Field

[0001] This application relates to the field of distributed cloud services, and in particular to an intelligent edge scheduling method, system, device and medium in a distributed cloud scenario. Background Technology

[0002] Distributed cloud has a significant cost advantage over centralized cloud and is becoming increasingly popular. The biggest difference between distributed cloud and centralized cloud is that distributed cloud uses a wide variety of distributed service nodes of various types to provide services to requesters from the nearest location.

[0003] However, current distributed clouds only provide localized cloud services using distributed nodes, still heavily relying on powerful central scheduling capabilities. The number of distributed nodes in a distributed cloud is enormous, and their specifications vary greatly. All of this necessitates a powerful scheduling center to compute and respond to all service requests, making the scheduling center crucial to the quality and cost of distributed cloud services.

[0004] The scheduling center that distributed clouds rely on is relatively far from the distributed service nodes and service objects, and the latency of service requests will have a significant impact on service quality. In addition, the scheduling center needs to calculate and respond to a large number of service requests, so the performance requirements are very high, which will also generate a lot of traffic costs, thus increasing the cost of distributed clouds. Summary of the Invention

[0005] This application provides an intelligent edge scheduling method, system, device, and medium in a distributed cloud scenario to solve the problem of low service quality and high service cost caused by the high requirements of existing distributed cloud center scheduling capabilities.

[0006] To address the aforementioned technical issues, this application adopts the following technical solution: providing an intelligent edge scheduling method in a distributed cloud scenario. This method includes: a scheduling center issuing a service probe range instruction to at least one service region, each service region comprising a regional scheduling center and multiple regional service nodes. The service probe range instruction is used to bind at least some regional service nodes of one service region to a regional scheduling center of another service region. The bound regional service nodes periodically report their service capabilities to their respective two regional scheduling centers. The regional dispatch center sorts the regional service nodes based on the service capabilities reported by each regional service node to generate a priority list of each regional service node. The priority list includes regional service nodes of the service area to which the regional dispatch center belongs and regional service nodes of another service area bound to the regional dispatch center. When the regional dispatch center receives a service request within its service area, it matches the regional service node and the service request terminal to which the service request belongs according to the priority list.

[0007] In an optional embodiment of this application, before the regional scheduling center sorts the regional service nodes, the intelligent edge scheduling method further includes: The regional dispatch center selects regional service nodes that meet the preset network quality requirements to form a regional service node list, and scores the network quality of each regional service node in the regional service node list.

[0008] In one optional embodiment of this application, the preset network quality requirements include latency, packet loss rate, and network speed; the regional dispatch center selects regional service nodes that meet the preset network quality requirements to form a regional service node list, and performs network quality scoring on each regional service node in the regional service node list, including: The regional dispatch center uses a first detection method to screen regional service nodes that meet preset latency and packet loss rate requirements; The regional dispatch center uses a second detection method to select regional service nodes that meet preset network speeds from those that meet preset latency and packet loss rate requirements, and forms a list of regional service nodes. A network quality score is obtained by weighting the latency, packet loss rate, and network speed of each regional service node in the regional service node list.

[0009] In one optional embodiment of this application, the service capabilities reported by the regional service node include node load and the network quality score; The regional dispatch center sorts the regional service nodes based on the service capabilities reported by each regional service node, including: The regional dispatch center performs a weighted calculation based on the node load of each regional service node and the network quality score to sort each regional service node and generate a priority list for each regional service node.

[0010] In one optional embodiment of this application, the service area is divided by the first regional attribute of the regional service node, and the service area is further divided into multiple sub-service areas according to the second regional attribute of each regional service node. The regional dispatch center sorts the regional service nodes based on the service capabilities reported by each regional service node, and also includes: The regional dispatch center also sorts the regional service nodes of each sub-service area based on the priority list, and generates a priority sub-list of regional service nodes for each sub-service area; the regional range corresponding to the second regional attribute is smaller than the regional range corresponding to the first regional attribute.

[0011] In an optional embodiment of this application, after generating the priority list of each regional service node and the priority sub-list of regional service nodes for each sub-service region, the intelligent edge scheduling method further includes: Each of the regional dispatch centers periodically traverses the list of regional service nodes to update the priority list and the priority sublist.

[0012] In one optional embodiment of this application, in response to a received service request, matching the regional service node and the service request terminal to which the service request belongs according to the priority list includes: In response to a received service request, the regional dispatch center parses the service request to obtain the second regional attribute contained in the service request; Based on the second region attribute and the service request, the service nodes of the region and the service request terminals to which the service request belongs are matched by traversing the priority sublist corresponding to the second region attribute. If no corresponding regional service node is found in the priority sublist, the regional service node and the service request terminal to which the service request belongs are matched by traversing the priority list.

[0013] To solve the above-mentioned technical problems, another technical solution adopted in this application is: to provide a distributed cloud system, including a scheduling center and multiple service areas, wherein the service areas include regional scheduling centers and multiple regional service nodes; The scheduling center is used to issue a service detection range instruction to at least one of the service areas. The service detection range instruction is used to bind at least some regional service nodes of one service area to the regional scheduling center of another service area. The bound regional service nodes periodically report their service capabilities to the two regional scheduling centers to which they belong. The regional dispatch center sorts the regional service nodes based on the service capabilities reported by each regional service node to generate a priority list for each regional service node. The priority list includes regional service nodes of the service area to which the regional dispatch center belongs and regional service nodes of another service area bound to the regional dispatch center.

[0014] To solve the above-mentioned technical problems, another technical solution adopted in this application is: to provide a computer device, including a memory, a processor and a computer program stored in the memory, characterized in that the processor executes the computer program to implement the steps of the intelligent edge scheduling method in the above-mentioned distributed cloud scenario.

[0015] To solve the above-mentioned technical problems, another technical solution adopted in this application is: to provide a computer-readable storage medium storing a computer program thereon, characterized in that the computer program, when executed by a processor, implements the steps of the intelligent edge scheduling method in the above-mentioned distributed cloud scenario.

[0016] The beneficial effects of this application are as follows: Unlike existing technologies, this application discloses an intelligent edge scheduling method, system, device, and medium in a distributed cloud scenario. This method involves a scheduling center issuing a service probe range instruction to at least one service area. Each service area includes a regional scheduling center and multiple regional service nodes. The service probe range instruction is used to bind at least some regional service nodes of one service area to the regional scheduling center of another service area. The bound regional service nodes periodically report their service capabilities to their respective two regional scheduling centers. Based on the reported service capabilities, the regional scheduling centers sort the regional service nodes to generate a priority list. Upon receiving a service request, the regional service nodes and the service requesting terminal to which the service request belongs are matched according to this priority list. This method reduces the service scope by dividing the service area, shortens the distance between the service requesting terminal and the regional service nodes, thereby shortening the request response time and improving service quality. The regional scheduling center only receives service requests within its own service area, significantly reducing the performance requirements of the scheduling nodes and thus reducing costs. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, wherein: Figure 1 This is a schematic diagram of the original distribution of the nodes in the distributed cloud provided in this application; Figure 2 This is a flowchart illustrating an embodiment of the intelligent edge scheduling method in a distributed cloud scenario provided in this application. Figure 3 This is a schematic diagram of the distributed cloud structure of Embodiment 1 of the intelligent edge scheduling method in the distributed cloud scenario provided in this application; Figure 4 This is a schematic diagram of the periodic reporting process of regional service nodes in Embodiment 1 of the intelligent edge scheduling method in the distributed cloud scenario provided in this application; Figure 5 This is a flowchart illustrating the process of generating a list of regional service nodes in Embodiment 1 of the intelligent edge scheduling method in a distributed cloud scenario provided in this application. Figure 6 This is a flowchart illustrating the intelligent detection task of Embodiment 1 of the intelligent edge scheduling method in the distributed cloud scenario provided in this application; Figure 7 This is a schematic diagram of the priority list and priority sub-list of the first embodiment of the intelligent edge scheduling method in the distributed cloud scenario provided in this application; Figure 8 This is a flowchart illustrating the periodic update of the priority list in Embodiment 1 of the intelligent edge scheduling method in a distributed cloud scenario provided in this application. Figure 9 This is a flowchart illustrating the process of allocating regional service nodes according to service requests in an embodiment of the intelligent edge scheduling method in a distributed cloud scenario provided in this application. Figure 10 This is a schematic diagram of the response service request process of Embodiment 1 of the intelligent edge scheduling method in the distributed cloud scenario provided in this application; Figure 11 This is a schematic diagram of the structure of Embodiment 2 of the distributed cloud system provided in this application. Detailed Implementation

[0018] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0019] The terms "first," "second," and "third" used in the embodiments of this application are for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first," "second," or "third" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0020] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0021] In this application, distributed cloud refers to a cloud model where a Cloud Service Provider (CSP) distributes public cloud services to different physical locations, with the CSP uniformly responsible for the operation, governance, updates, and evolution of the cloud services, and incorporating the geographical location of cloud service delivery into its definition. The original distribution of the scheduling center, service nodes, and service requesters (service request terminals) in traditional distributed clouds is based on... Figure 1 , Figure 1 This is a schematic diagram of the original distribution of nodes in the distributed cloud provided in this application. The traditional distributed cloud service delivery process is roughly as follows: 1. Service nodes periodically report their service capabilities to the scheduling center, including node performance metrics, service capability list, load, location, and other information; 2. The dispatch center refreshes and summarizes the information of all service nodes; 3. When a service requester sends a service request to the scheduling center, the scheduling center selects a suitable service node for the requester based on the service type, location, and other dimensions. 4. After obtaining the service node information, the service requester sends a request to that node to obtain the service.

[0022] Example 1 This application provides an intelligent edge scheduling method in a distributed cloud scenario, referring to... Figure 2 , Figure 2This is a flowchart illustrating an embodiment of the intelligent edge scheduling method in a distributed cloud scenario provided in this application. The intelligent edge scheduling method in a distributed cloud scenario includes: S10: The dispatch center issues a service probe range instruction to at least one service area. Each service area includes a regional dispatch center and multiple regional service nodes. The service probe range instruction is used to bind at least some regional service nodes of one service area to the regional dispatch center of another service area. The bound regional service nodes periodically report their service capabilities to their respective two regional dispatch centers.

[0023] Traditional distributed cloud architectures consist of a scheduling center and multiple service nodes. When the scheduling center or service nodes are too far from the service requesting terminal, service request delays occur. Therefore, this application restructures the distributed cloud architecture, referring to... Figure 3 , Figure 3 This is a schematic diagram of the distributed cloud structure of the first embodiment of the intelligent edge scheduling method in the distributed cloud scenario provided in this application. The service nodes are divided into different service areas according to the first regional attribute among the known natural attributes of the service nodes. Each service area is deployed with a regional scheduling center and multiple regional service nodes located in the service area. The regional scheduling center is responsible for receiving information reports from all regional service nodes in its service area and service requests from all service request terminals in its service area.

[0024] Among these, the known natural attributes of the service nodes can be administrative division attributes. The first regional attribute among these natural attributes can be the province attribute, meaning that the service nodes are divided into service areas based on the province they are located in, such as the Hubei Province service area. In the Hubei Province service area, a regional dispatch center located in Hubei Province and multiple regional service nodes located in Hubei Province are deployed. Alternatively, the natural attribute can also be latitude and longitude attributes, for example, dividing the service areas into multiple square-shaped areas based on latitude and longitude; no specific limitation is made here.

[0025] Unlike existing technologies, this application divides the service area according to the first regional attribute of the regional service node, and refines the originally huge distributed cloud according to natural geography or administrative region into multiple service areas containing regional scheduling nodes. This allows service requesting terminals to send requests to the nearest regional scheduling center and obtain the nearest regional service node to provide services. By narrowing the service range, the response time is reduced and the response is more timely, thereby improving the service quality.

[0026] In this application, the dispatch center issues a service detection range instruction to at least one service area. For example, if service area A issues a service detection range instruction to service area B, this service detection instruction will bind at least some of the regional service nodes of service area B to the regional dispatch center of service area A. That is, after the binding is completed, the bound regional dispatch center can receive the reported information from all regional service nodes within its own service area A, as well as the reported information from at least some of the regional service nodes of its bound service area B.

[0027] The service range detection command can be flexibly configured according to actual needs, including the selection of the service range and the selection of regional service nodes within the service area. For example, the dispatch center issues a service range detection command to the Hubei Province service area to detect the Hunan Province service area, binding some regional service nodes in the Hunan Province service area that are adjacent to Hubei Province (such as some regional service nodes at the border of the two provinces) to the regional dispatch center of the Hubei Province service area. In this way, the regional dispatch center of the Hubei Province service area can receive the reported information from some regional service nodes in the Hunan Province service area that are adjacent to Hubei Province. When a service request terminal in the Hubei Province service area that is adjacent to Hunan Province initiates a request, it can also match the nearest regional service nodes in the Hunan Province service area that are adjacent to Hubei Province to provide services to these service request terminals.

[0028] In addition, a regional dispatch center within a service area can also bind some or all regional service nodes in multiple other service areas, with no limit on the number of other service areas bound or the number of regional service nodes in those other service areas.

[0029] It is understandable that by binding at least some regional service nodes of a service area to the regional dispatch center of another service area according to the service detection range instruction, the service range may overlap. That is, a certain service node can be dispatched by the regional dispatch center of service area A or the regional dispatch center of service area B to provide services to two service areas. This can maximize the flexible use of regional service nodes while meeting the service capacity requirements, and avoid resource waste while providing high-quality services.

[0030] In this application, reference is made to Figure 4 , Figure 4This is a schematic diagram of the periodic reporting process of regional service nodes in the first embodiment of the intelligent edge scheduling method in the distributed cloud scenario provided in this application. All regional service nodes in each service area need to periodically report their service capabilities to the regional scheduling center in that service area. The bound regional service nodes also need to periodically report their service capabilities (including the service node's service identifier, performance, load, location, and resources) to the two or more regional scheduling centers to which they belong.

[0031] S20: Based on the service capabilities reported by each regional service node, the regional dispatch center sorts the regional service nodes to generate a priority list for each regional service node. The priority list includes regional service nodes in the service area to which the regional dispatch center belongs and regional service nodes in another service area bound to the regional dispatch center.

[0032] In this application, before the regional dispatch center sorts the service nodes in each region, the intelligent edge dispatch method further includes: The regional dispatch center selects regional service nodes that meet the preset network quality requirements to form a regional service node list, and scores the network quality of each regional service node in the list.

[0033] In this application, the regional dispatch center periodically receives service capabilities reported by all regional service nodes within its service area and all regional service nodes bound to it. After receiving the service capabilities reported by each regional service node, the regional dispatch center, in addition to summarizing the service capabilities of each regional service node, can also initiate an intelligent detection task. The intelligent detection task refers to filtering regional service nodes based on their service capabilities and corresponding detection methods, selecting regional service nodes that meet preset network quality requirements to form a list of regional service nodes, i.e., selecting a list of truly usable regional service nodes to provide services to service requesting terminals.

[0034] Unlike existing technologies, the intelligent edge scheduling method in the distributed cloud scenario provided in this application selects regional service nodes that meet preset network quality requirements through a regional scheduling center to form a list of regional service nodes, thereby obtaining the actual usable service range. By scheduling regional service nodes through the regional scheduling center, the burden on the overall scheduling center is reduced, and the computational load of each regional scheduling center is significantly reduced, thereby reducing the performance requirements of the regional scheduling center and thus reducing costs. Furthermore, by flexibly selecting regional service nodes through intelligent detection, the response rate of requests is improved, and the service quality is enhanced.

[0035] In this application, reference is made to Figure 5 , Figure 5This is a flowchart illustrating the generation of a regional service node list in Embodiment 1 of the intelligent edge scheduling method for a distributed cloud scenario provided in this application. Preset network quality requirements include latency, packet loss rate, and network speed. The regional scheduling center selects regional service nodes that meet the preset network quality requirements to form a regional service node list, and performs network quality scoring on each regional service node in the list, including: S21: The regional dispatch center uses the first detection method to screen regional service nodes that meet the preset latency and packet loss rate requirements.

[0036] S22: The regional dispatch center uses a second detection method to select regional service nodes that meet the preset network speed from the regional service nodes that meet the preset latency and packet loss rate requirements, and forms a list of regional service nodes.

[0037] S23: Calculate the network quality score by weighting the latency, packet loss rate, and network speed of each regional service node in the regional service node list.

[0038] In this application, the intelligent detection task for each regional service node mainly includes two types. One is to execute step S21 above, where the regional dispatch center uses a first detection method to filter regional service nodes that meet preset latency and packet loss rate requirements, i.e., filtering based on the network distance between each regional service node. The other is to execute step S22 above, where the regional dispatch center uses a second detection method to filter regional service nodes that meet preset network speeds from those meeting the preset latency and packet loss rate requirements, i.e., filtering based on the network performance of each regional service node, and then forming a regional service node list from the filtered regional service nodes that meet the preset latency, packet loss rate, and network speed requirements. The first detection method can be a ping detection method, and the second detection method can be a download speed test detection method. In addition, the order of steps S21 and S22 can be sequential, reverse (i.e., first filter based on the network speed of the regional service nodes, then filter based on the latency and packet loss rate of the regional service nodes), or parallel. In short, regional service nodes that meet the preset requirements for latency, packet loss rate and network speed need to be selected. There is no limitation on the execution order here.

[0039] Reference Figure 6 , Figure 6This is a flowchart illustrating the intelligent detection task of Embodiment 1 of the intelligent edge scheduling method in the distributed cloud scenario provided in this application. After receiving the service capabilities reported by the regional service node, the regional scheduling center initiates the first detection method (ping detection method) to detect whether the latency and packet loss rate of the regional service node meet the preset latency and packet loss rate requirements. If one of them is not met, the regional service node is discarded. If both are met, the second detection method (download speed test detection method) is initiated to detect whether the network speed (and jitter) of the regional service node meets the preset network speed (and jitter) requirements. If one of them is not met, the regional service node is discarded. If both are met, it indicates that the regional service node is a real and usable regional service node, and the regional service node is added to the regional service node list.

[0040] For example, if the detection results of service node A in a certain area are 400ms latency, 8% packet loss rate, and 3Mbps network speed, and assuming that the preset latency and packet loss rate requirements are no more than 500ms latency and no more than 10% packet loss rate, and the preset network speed requirement is greater than 1Mbps, then service node A in this area meets the preset requirements in terms of latency, packet loss rate, and network speed, and is determined to be a real and usable regional service node, and is then added to the regional service node list.

[0041] In this application, the latency and packet loss rate of the regional service node are obtained through a first probing method, and the network speed of the regional service node is obtained through a second probing method. Based on the latency, packet loss rate, and network speed of the regional service node, the network quality of the regional service node can also be scored.

[0042] The scoring rules can be based on a weighted calculation of latency, packet loss rate, and network speed of regional service nodes. For example, the weighted scoring formula can be: Regional service node network quality score = (500 - latency) + (10 - packet loss rate) * 10 + network speed * 10. If the network quality assessment result of a certain regional service node B is latency of 100ms, packet loss rate of 5%, and network speed of 5Mbps, then the network quality score of regional service node B = (500 - 100) + (10 - 5) * 10 + 8 * 10 = 530.

[0043] Unlike existing technologies, the intelligent edge scheduling method for distributed cloud scenarios provided in this application uses a first detection method to filter regional service nodes that meet preset latency and packet loss rate requirements, and a second detection method to filter regional service nodes that meet preset network speeds from the regional service nodes that meet the preset latency and packet loss rate requirements, forming a list of regional service nodes. By filtering regional service nodes based on network distance and network performance, the availability of regional service nodes is improved, thereby enhancing service quality.

[0044] In this application, the service capabilities reported by the regional service nodes include node load and the network quality score; the regional dispatch center ranks the regional service nodes based on the service capabilities reported by each regional service node, including: The regional dispatch center performs weighted calculations based on the node load and network quality scores of each regional service node to sort the regional service nodes and generate a priority list for each regional service node.

[0045] In this application, to achieve rapid response to service requests, the priority of each regional service node in the regional service node list can be sorted. The priority of a regional service node can be determined based on its service capabilities, which may include indicators such as the regional service node's service identifier, node load, network quality score, and location information. The priority of a regional service node is calculated by weighting its network quality score and node load (including network load and CPU load) as an example. If a service node B in a certain region has a network quality score of 530, a network load (i.e., the proportion of network usage to effective bandwidth) of 60%, and a CPU load (i.e., CPU utilization) of 50%, the weighting of the corresponding regional service node can be flexibly adjusted according to the actual situation when calculating the priority of the corresponding regional service node based on the weighted calculation of network quality score and node load. For example, if the priority value = network quality assessment result * 0.3 + (100 - network load) * 0.5 + (100 - CPU load) * 0.2, then the priority value of the regional service node B = 530 * 0.3 + (100 - 60) * 0.5 + (100 - 50) * 0.2 = 189.

[0046] The priority values ​​of all regional service nodes in the regional service list are calculated using the above method. Then, the regional service nodes are sorted from largest to smallest according to their priority values ​​to generate a priority list for each regional service node.

[0047] Unlike existing technologies, the intelligent edge scheduling method in the distributed cloud scenario provided in this application calculates the priority of regional service nodes based on the network quality score and node load weighting of each regional service node, sorts the regional service nodes, and generates a priority list of regional service nodes. This facilitates the allocation of regional service nodes according to priority when a service request is received from a service requesting terminal, enabling rapid response to service requests. Allocating the highest priority regional service node also improves service quality.

[0048] In this application, the service area is divided by the first region attribute of the regional service node, and the service area is further divided into multiple sub-service areas according to the second region attribute of each regional service node. The regional dispatch center ranks the regional service nodes based on their reported service capabilities, and also includes: The regional dispatch center also sorts the regional service nodes of each sub-service area based on the priority list, and generates a priority sub-list of regional service nodes for each sub-service area; the regional range corresponding to the second regional attribute is smaller than the regional range corresponding to the first regional attribute.

[0049] In this application, to facilitate faster response to service requests, the existing service area can be further subdivided into smaller sub-service areas. This subdivision can be based on the second regional attribute of each service node. Since the known natural attribute of a service node can be an administrative division attribute, and the first regional attribute can be a province, the second regional attribute can be a city. For example, a service area within Hubei Province can be further subdivided into sub-service areas within Wuhan City. Furthermore, even smaller service areas can be further subdivided based on the sub-service areas corresponding to the second regional attribute. For example, within the sub-service areas corresponding to the city attribute, service areas corresponding to counties and districts can be further subdivided. Different granularities and sizes of service areas can be flexibly divided according to actual needs; no limitations are imposed here.

[0050] After dividing the service area into multiple sub-service areas according to the second-region attributes of each regional service node, refer to Figure 7 , Figure 7 This is a schematic diagram of the priority list and priority sub-list of the intelligent edge scheduling method in the distributed cloud scenario provided in this application. The regional scheduling center will also sort the priority of each regional service node in each sub-service area according to the division of sub-service areas, and generate a priority sub-list of regional service nodes in each sub-service area.

[0051] Since the dispatch center of a service area may be bound to some regional service nodes of other service areas, when generating the priority list of regional service nodes for that service area, these regional service nodes contained in other service areas are also sorted in that priority list. When generating the priority sub-lists of regional service nodes for each sub-service area, these regional service nodes contained in other service areas can be added to the corresponding priority sub-lists according to their own second region attributes. For example, if a service area within Hubei Province is bound to some regional service nodes located in Yueyang City within a service area within Hunan Province, when generating the priority list for the service area within Hubei Province, these regional service nodes located in Yueyang City are also sorted in that priority list according to their priority values; when generating each priority sub-list for the service area within Hubei Province, these regional service nodes located in Yueyang City generate the corresponding priority sub-list for Yueyang City. Alternatively, regional service nodes contained in other service areas can be matched with their neighboring sub-service areas according to their own second region attributes, and these regional service nodes contained in other service areas can be added to the priority sub-lists of their neighboring sub-service areas. This can be flexibly adjusted according to actual needs and is not limited here.

[0052] Unlike existing technologies, the intelligent edge scheduling method for distributed cloud scenarios provided in this application divides multiple sub-service regions according to the second region attributes of each regional service node, and then sorts the regional service nodes of each sub-service region to generate a priority sub-list of regional service nodes for each sub-service region. By generating priority lists of different granularities, the service scope is divided more finely. When allocating regional service nodes, the distance between regional service nodes and service request terminals is closer, thereby reducing request latency and improving service efficiency and service quality.

[0053] In this application, after generating the priority list of service nodes in each region and the priority sub-list of service nodes in each sub-service region, the intelligent edge scheduling method in a distributed cloud scenario further includes: Each regional dispatch center periodically traverses the list of regional service nodes to update the priority list and priority sublist.

[0054] In this application, reference is made to Figure 8 , Figure 8This is a flowchart illustrating the periodic update of the priority list in Embodiment 1 of the intelligent edge scheduling method for distributed cloud scenarios provided in this application. The regional scheduling center periodically receives service capabilities reported by each regional service node. Simultaneously, it periodically traverses the list of regional service nodes, determining if any regional service nodes have failed to report within the time limit. These time-limited regional service nodes are discarded, and all remaining regional service nodes that report on time form a new list of regional service nodes. A network quality score for each regional service node in the new list is obtained through an intelligent detection task, thus acquiring the node load and network quality score for each regional service node. Priority values ​​are recalculated based on the node load and network quality scores of each regional service node, and the priority list and priority sub-list are updated according to these priority values.

[0055] Unlike existing technologies, the intelligent edge scheduling method in the distributed cloud scenario provided in this application can flexibly allocate the regional service nodes with the strongest service capabilities to service request terminals by periodically updating the priority list and priority sub-list based on the service capabilities reported by the regional service nodes. This avoids the regional service nodes from competing for resources and reducing service quality due to too many service objects.

[0056] S30: When the regional dispatch center receives a service request within its service area, it matches the regional service node and the service request terminal to which the service request belongs according to the priority list.

[0057] In this application, reference is made to Figure 9 , Figure 9 This is a flowchart illustrating the process of matching regional service nodes based on service requests in an embodiment of the intelligent edge scheduling method for distributed cloud scenarios provided in this application. In response to a received service request, the method matches regional service nodes and the service request terminal to which the service request belongs according to a priority list, including: S31: In response to the received service request, the regional dispatch center parses the service request and obtains the second regional attribute contained in the service request.

[0058] S32: Based on the second region attribute and service request, traverse the matching region service node and the service request terminal to which the service request belongs according to the priority sublist corresponding to the second region attribute.

[0059] S33: If no corresponding regional service node is found in the priority sublist, the regional service node and the service request terminal to which the service request belongs are matched by traversing the priority list.

[0060] In this application, when a service requesting terminal needs to obtain a service, it no longer sends a request to the central dispatch center, but instead sends a service request to the regional dispatch center of its service area, which has a faster response time than sending a request to the central dispatch center.

[0061] In this application, reference is made to Figure 10 , Figure 10 This is a flowchart illustrating the response to a service request in an embodiment of the intelligent edge scheduling method in a distributed cloud scenario provided in this application. When the regional scheduling center receives a service request sent by a service request terminal within its service area, it parses the service request and obtains the service information requested by the service request terminal and its second regional attribute (which may be a city attribute).

[0062] After parsing the service request to obtain its corresponding second region attributes and service information, the priority sublist of the regional service nodes of the corresponding sub-service region is first matched according to the second region attributes. Then, the regional service nodes and service request terminals are matched according to the service information of the service request in the order of the priority sublist. If a regional service node that can provide the corresponding service is matched according to the service information, the regional service node is matched with the service request terminal to which the service request belongs. The service request terminal communicates and connects with the matched regional service node to obtain the service.

[0063] If, in step S32, traversing the priority sublist (city level) still fails to find a regional service node that can provide services to the service requesting terminal, then step S33 is executed. The priority list (provincial level) of regional service nodes in the service area to which the service requesting terminal belongs is then traversed. Regional service nodes and service requesting terminals are matched according to the service information of the service request, in the order listed in the priority list. If a regional service node capable of providing the corresponding service is found based on the service information, then that regional service node is matched with the service requesting terminal to which the service request belongs. If no match is found, then the first regional service node in the priority list is matched with the service requesting terminal to which the service request belongs, and the service requesting terminal then communicates and connects with the matched regional service node to obtain the service.

[0064] In addition, if the priority sublist (city level) still does not match a regional service node that provides services to the terminal requesting the service, the priority sublists of the neighboring cities can be traversed for matching. If the corresponding regional service node is still not matched in the priority list, the regional dispatch center can continue to send the service request to the central dispatch center, which will then obtain the priority list in other service areas for matching. This is not limited here.

[0065] Unlike existing technologies, the intelligent edge scheduling method in the distributed cloud scenario provided in this application first matches the regional service node and the service request terminal to which the service request belongs in the priority sublist according to the second regional attribute contained in the service request after receiving the service request. If no match is found, the method continues to match the regional service node and the service request terminal to which the service request belongs in the priority list. This ensures that the regional service node matched with the service request terminal is the closest regional service node with the strongest service capability, thereby improving service efficiency and service quality.

[0066] Example 2 This application provides a distributed cloud system, referring to... Figure 11 , Figure 11 This is a schematic diagram of the structure of a second embodiment of the distributed cloud system provided in this application. The distributed cloud system 100 includes a scheduling center 10 and multiple service areas 20. Each service area 20 includes a regional scheduling center 21 and multiple regional service nodes 22. The dispatch center 10 is used to issue a service detection range instruction to at least one service area 20. The service detection range instruction is used to bind at least some of the regional service nodes 22 of one service area 20 to the regional dispatch center 21 of another service area 20. The bound regional service nodes 22 periodically report their service capabilities to the two regional dispatch centers 21 to which they belong. Based on the service capabilities reported by each regional service node 22, the regional dispatch center 21 sorts the regional service nodes 22 to generate a priority list of each regional service node 22. The priority list includes the regional service nodes 22 of the service area 20 to which the regional dispatch center 21 belongs and the regional service nodes 22 of another service area 20 bound to the regional dispatch center 21.

[0067] In the distributed cloud system of this embodiment two, the specific process of the interaction between the scheduling center 10, service area 20, regional scheduling center 21 and regional service node 22 to realize regional service node scheduling and priority sorting can be referred to the detailed description of the intelligent edge scheduling method in the distributed cloud scenario of embodiment one above. The repeated parts will not be repeated here.

[0068] Example 3 Based on the same inventive concept, this application also provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the intelligent edge scheduling method in a distributed cloud scenario as described in Embodiment 1 above.

[0069] Example 4 Based on the same inventive concept, this application also provides a computer-readable storage medium storing a computer program / instruction thereon, which, when executed by a processor, implements the intelligent edge scheduling method in a distributed cloud scenario as described in Embodiment 1 above. Unlike existing technologies, this application discloses an intelligent edge scheduling method, system, device, and medium in a distributed cloud scenario. The method involves a scheduling center issuing a service probe range instruction to at least one service region. Each service region includes a regional scheduling center and multiple regional service nodes. The service probe range instruction binds at least some regional service nodes from one service region to the regional scheduling center of another service region. The bound regional service nodes periodically report their service capabilities to their respective two regional scheduling centers. Based on the reported service capabilities, the regional scheduling centers sort the regional service nodes to generate a priority list. Upon receiving a service request, the system matches the regional service nodes and the service requesting terminal to which the service request belongs according to this priority list. This method reduces the service scope by dividing the service region, shortens the distance between the service requesting terminal and the regional service nodes, thereby reducing request response time and improving service quality. The regional scheduling center only receives service requests within its own service range, significantly reducing the performance requirements of the scheduling nodes and thus lowering costs.

[0070] The above description is merely an embodiment of this application and does not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. An intelligent edge scheduling method in a distributed cloud scenario, characterized in that, include: The dispatch center issues a service probe range instruction to at least one service area. Each service area includes a regional dispatch center and multiple regional service nodes. The service probe range instruction is used to bind at least some regional service nodes of one service area to the regional dispatch center of another service area. The bound regional service nodes periodically report their service capabilities to the two regional dispatch centers to which they belong. The regional dispatch center uses a first detection method to screen regional service nodes that meet preset latency and packet loss rate requirements; The regional dispatch center uses a second detection method to select regional service nodes that meet preset network speeds from those that meet preset latency and packet loss rate requirements, and forms a list of regional service nodes. A network quality score is obtained by weighting the latency, packet loss rate, and network speed of each regional service node in the regional service node list; wherein, the first detection method is the ping detection method, and the second detection method is the download speed test detection method. The regional dispatch center sorts the regional service nodes based on the service capabilities reported by each regional service node to generate a priority list for each regional service node. The priority list includes regional service nodes in the service area to which the regional dispatch center belongs and regional service nodes in another service area bound to the regional dispatch center. The service capabilities reported by each regional service node include node load and network quality score. When the regional dispatch center receives a service request within its service area, it matches the regional service node and the service request terminal to which the service request belongs according to the priority list. 2.The intelligent edge scheduling method in the distributed cloud scenario according to claim 1, characterized in that, The regional dispatch center sorts the regional service nodes based on the service capabilities reported by each regional service node, including: The regional dispatch center performs a weighted calculation based on the node load of each regional service node and the network quality score to sort each regional service node and generate a priority list for each regional service node.

3. The intelligent edge scheduling method in a distributed cloud scenario according to claim 2, characterized in that, The service area is divided by the first region attribute of the regional service node, and the service area is further divided into multiple sub-service areas according to the second region attribute of each regional service node. The regional dispatch center sorts the regional service nodes based on the service capabilities reported by each regional service node, and also includes: The regional dispatch center also sorts the regional service nodes of each sub-service area based on the priority list, and generates a priority sub-list of regional service nodes for each sub-service area; the regional range corresponding to the second regional attribute is smaller than the regional range corresponding to the first regional attribute. 4.The intelligent edge scheduling method in a distributed cloud scenario according to claim 3, characterized in that, After generating the priority list of each regional service node and the priority sub-list of regional service nodes for each sub-service region, the intelligent edge scheduling method further includes: Each of the regional dispatch centers periodically traverses the list of regional service nodes to update the priority list and the priority sublist. 5.The intelligent edge scheduling method in distributed cloud scenario according to claim 4, characterized in that, In response to a received service request, the service node in the region and the service request terminal to which the service request belongs are matched according to the priority list, including: In response to a received service request, the regional dispatch center parses the service request to obtain the second regional attribute contained in the service request; Based on the second region attribute and the service request, the service nodes of the region and the service request terminals to which the service request belongs are matched by traversing the priority sublist corresponding to the second region attribute. If no corresponding regional service node is found in the priority sublist, the regional service node and the service request terminal to which the service request belongs are matched by traversing the priority list.

6. A distributed cloud system, characterized by, It includes a dispatch center and multiple service areas, wherein the service areas include regional dispatch centers and multiple regional service nodes; The scheduling center is used to issue a service detection range instruction to at least one of the service areas. The service detection range instruction is used to bind at least some regional service nodes of one service area to the regional scheduling center of another service area. The bound regional service nodes periodically report their service capabilities to the two regional scheduling centers to which they belong. The regional dispatch center uses a first detection method to screen regional service nodes that meet preset latency and packet loss rate requirements; The regional dispatch center uses a second detection method to select regional service nodes that meet preset network speeds from those that meet preset latency and packet loss rate requirements, and forms a list of regional service nodes. A network quality score is obtained by weighting the latency, packet loss rate, and network speed of each regional service node in the regional service node list; wherein, the first detection method is the ping detection method, and the second detection method is the download speed test detection method. The regional dispatch center sorts the regional service nodes based on the service capabilities reported by each regional service node to generate a priority list for each regional service node. The priority list includes regional service nodes in the service area to which the regional dispatch center belongs and regional service nodes in another service area bound to the regional dispatch center. The service capabilities reported by each regional service node include node load and network quality score.

7. A computer device, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the intelligent edge scheduling method in a distributed cloud scenario as described in any one of claims 1-5.

8. A storage medium having stored thereon a computer program, characterized in that When executed by a processor, the computer program implements the steps of the intelligent edge scheduling method in a distributed cloud scenario as described in any one of claims 1-5.

Images