Edge cache node bandwidth resource planning method, system and device and storage medium

Abstract

The invention relates to the technical field of scheduling planning, and discloses an edge cache node bandwidth resource planning method, system and device and a storage medium, and the edge cache node bandwidth resource planning method comprises the steps: determining a resource planning strategy of an edge cache node in combination with historical data of a service domain name and resource attributes of the edge cache node; acquiring performance index data such as real-time bandwidth of the edge node; determining the actual service bandwidth of the edge cache node based on the real-time performance index data; determining a target service bandwidth of the edge cache node based on a resource planning strategy; and performing load evaluation on the edge cache node in combination with the actual service bandwidth and the target service bandwidth, and adjusting a resource planning strategy based on a load evaluation result so as to reduce the difference between the actual service bandwidth and the target service bandwidth. According to the method, the node can be reduced from being frequently scheduled to different service domain names, the father-back bandwidth is reduced, and the father-back bandwidth cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of scheduling planning, and particularly to a method, system, device, and storage medium for bandwidth resource planning of edge cache nodes. Background Art

[0002] CDN (Content Delivery Network) is a distributed network architecture designed to improve user experience and network performance by delivering content to nodes closer to users.

[0003] The main optimization goal of existing CDN scheduling systems is the load balancing of edge nodes, and less consideration is given to problems such as the reduction in hit rate and the increase in return bandwidth caused by frequent scheduling of nodes to different services. In addition, the existing bandwidth resource planning of CDN nodes mainly determines whether a node needs to increase or decrease the amount of business it undertakes based on the bandwidth redundancy of the node. This results in a large difference between the actual bandwidth of the node and the planned bandwidth, increasing the decision-making burden of the scheduling system, affecting the planning of other nodes, and affecting the load balancing of the entire CDN system. Summary of the Invention

[0004] In view of this, the present invention provides a method, system, device, and storage medium for bandwidth resource planning of edge cache nodes to solve the problem of frequent scheduling of edge cache nodes to different services and improve the hit rate.

[0005] In a first aspect, the present invention provides a method for bandwidth resource planning of edge cache nodes, which includes: determining a resource planning strategy for an edge cache node by combining historical data of a service domain name and resource attributes of the edge cache node, where the resource planning strategy includes a planned service domain and a planned service bandwidth amount for the edge cache node, the planned service domain includes a planned service domain name and a planned service domain area, and the edge cache node is an edge cache node of a content delivery network; obtaining real-time performance index data of the edge cache node; performing real-time resource scheduling on the edge cache node according to the resource planning strategy and the real-time performance index data, where the result of the real-time resource scheduling affects the real-time performance index data; determining the actual service bandwidth of the edge cache node based on the real-time performance index data; determining the target service bandwidth of the edge cache node based on the resource planning strategy; evaluating the load of the edge cache node by combining the actual service bandwidth and the target service bandwidth, and adjusting the resource planning strategy based on the load evaluation result to narrow the difference between the actual service bandwidth and the target service bandwidth.

[0006] In this implementation manner, the present application makes a preliminary plan based on the historical data of the service domain name and the resource attributes of the edge cache nodes, obtains the planned services and planned bandwidth conditions of the edge cache nodes, and analyzes the load conditions of the edge cache nodes by comparing the planned conditions with the actual scheduling conditions, and adjusts the resource planning strategy in real time to reduce the deviation between the planned conditions and the actual scheduling, reduce the frequency of the edge cache nodes being scheduled to different services, reasonably plan the bandwidth resources of the edge cache nodes, reduce the backhaul bandwidth, reduce the backhaul cost, further improve the hit rate. At the same time, by adjusting the resource planning strategy, the resource coverage plan can be optimized, the burden of the scheduling decision-making link can be reduced, and the rationality of the resource planning can be improved.

[0007] In an alternative implementation manner, the historical data of the service domain name includes historical metric data and customer demand data, and the resource attributes of the edge cache nodes include location information, network quality information, hardware information, hardware constraint information, and service constraint information. Determining the resource planning strategy of the edge cache nodes by combining the historical data of the service domain name and the resource attributes of the edge cache nodes includes: determining the service domain of the edge cache node based on the historical metric data of the service domain name and the hardware information of the edge cache node; determining the service domain area range of the edge cache node based on the location information, network quality information of the edge cache node, and the customer demand data of the service domain name; determining the available bandwidth capacity of the edge cache node based on the hardware constraint information and service constraint information of the edge cache node; and determining the resource planning strategy by combining the service domain, service domain area range, available bandwidth capacity of the edge cache node, and the current service demand.

[0008] In this implementation manner, by comprehensively considering the resource data of the edge cache nodes, the historical metric data, and the customer demand data of the service domain name, a resource coverage plan that meets various requirements is determined for the edge cache nodes, which can improve the rationality of the preliminary resource planning.

[0009] In an alternative implementation manner, the historical data of the service domain name includes service quality data, and the resource attributes of the edge cache nodes include resource cost data. Real-time resource scheduling of the edge cache nodes is performed according to the resource planning strategy and the real-time performance metric data, including: weighing the resource cost and the service quality data to determine the resource scheduling strategy; pulling the planned service bandwidth amount corresponding to the planned service domain to the corresponding edge cache node; if the real-time performance metric data of the current edge cache node does not meet the metric requirements, pulling the service bandwidth amount of the current edge cache node to the target edge cache node according to the resource scheduling strategy, where the real-time performance metric data of the target edge cache node meets the metric requirements, there is redundant service bandwidth amount in the target edge cache node, and the planned service domains of the target edge cache node and the current edge cache node are the same.

[0010] In this implementation, preliminary resource scheduling is performed according to the resource planning strategy and the resource scheduling strategy, so that the resource scheduling meets the scheduling requirements, can control the traffic volume covered by each node, and reduce the decision-making burden of the scheduling decision-making system.

[0011] In an alternative implementation, the edge cache node provides service through multiple virtual IPs, and determines the actual service bandwidth of the edge cache node based on real-time performance metric data, including: determining the virtual IPs actually occupied by each service domain name in each service domain area based on the real-time performance metric data, and obtaining the actual service bandwidth of the virtual IPs; determining the target service bandwidth of the edge cache node based on the resource planning strategy, including: determining the virtual IPs planned to be occupied by each service domain name in each service domain area based on the resource planning strategy, and obtaining the planned service bandwidth of the virtual IPs; evaluating the load of the edge cache node by combining the actual service bandwidth and the target service bandwidth, and adjusting the resource planning strategy based on the load evaluation result, including: calculating the load data of each virtual IP based on the actual service bandwidth and the planned service bandwidth for the same service area of the same service domain name; accumulating the load data of multiple virtual IPs to calculate the load data of the edge cache node; the load data includes overloaded bandwidth, underloaded bandwidth, and overloaded bandwidth; adjusting the resource planning strategy based on the load data of the edge cache node.

[0012] In this implementation, taking the virtual IP as the analysis unit, comparing the difference between the actual service bandwidth and the planned bandwidth of the virtual IP to determine the load of the virtual IP, and further calculating the load of the edge cache node by combining the loads of multiple virtual IPs, can implement the method of adjusting the resource planning strategy according to the bandwidth difference, which helps to improve the resource utilization rate of the node. And ensuring that the services on the node are as concentrated as possible can reduce the increase in the return bandwidth caused by caching multiple service bandwidths.

[0013] In an alternative implementation, determining the planned virtual IPs planned to be occupied by each service domain name in each service area based on the resource planning strategy, and obtaining the planned service bandwidth of the planned virtual IPs, includes: determining the edge cache nodes planned to be occupied by each service domain name in each service area and the real-time service bandwidth based on the resource planning strategy; obtaining the number of virtual IPs planned to be occupied in the edge cache node; evenly distributing the planned service bandwidth of the planned virtual IPs based on the real-time service bandwidth and the number of virtual IPs.

[0014] In an alternative implementation, calculating the load data of each virtual IP based on the actual service bandwidth and the planned service bandwidth includes: when a virtual IP has an actual service bandwidth but no planned service bandwidth, the actual service bandwidth is taken as the overloaded bandwidth of the virtual IP; when a virtual IP has an actual service bandwidth and a planned service bandwidth, and the actual service bandwidth corresponding to the virtual IP is greater than the planned service bandwidth, the difference between the actual service bandwidth and the planned service bandwidth is taken as the overloaded bandwidth of the virtual IP; when a virtual IP has an actual service bandwidth and a planned service bandwidth, and the actual service bandwidth corresponding to the virtual IP is less than the planned service bandwidth, the difference between the actual service bandwidth and the planned service bandwidth is taken as the underloaded bandwidth of the virtual IP.

[0015] In this implementation, a calculation method for overloaded bandwidth, overloaded bandwidth, and underloaded bandwidth is proposed, which can quantify the load situation and is beneficial to subsequent evaluation of the accuracy of resource planning strategies.

[0016] In an alternative implementation, adjusting the resource planning strategy based on the load data of the edge cache node includes: when the overloaded bandwidth of the edge cache node is greater than the first bandwidth threshold, reducing the planned service bandwidth of the edge cache node; when the underloaded bandwidth of the edge cache node is greater than the second bandwidth threshold, increasing the planned service bandwidth of the edge cache node.

[0017] In this implementation, an adjustment method for the resource planning strategy is proposed, which can gradually reduce the deviation between the actual bandwidth and the planned bandwidth.

[0018] Second aspect, the present invention provides a bandwidth resource planning system for edge caching nodes. The bandwidth resource planning system for edge caching nodes includes multiple edge caching nodes, a data center, a resource planning module, a scheduling decision module, and a load evaluation module. The data center is connected to the multiple edge caching nodes, the resource planning module, the scheduling decision module, and the load evaluation module. The resource planning module is connected to the scheduling decision module and the load evaluation module. The edge caching nodes are edge caching nodes of a content delivery network. The data center is configured to obtain real-time performance metric data, resource attributes, and historical data of service domain names of the edge caching nodes, send the real-time performance metric data to the scheduling decision module and the load evaluation module, and send the resource attributes and historical data to the resource planning module. The resource planning module is configured to determine a resource planning strategy for the edge caching nodes by combining the historical data of the service domain names and the resource attributes of the edge caching nodes. The resource planning strategy includes a planned service domain and a planned service bandwidth amount for the edge caching nodes. The planned service domain includes a planned service domain name and a planned service domain area. The scheduling decision module is configured to perform real-time resource scheduling on the edge caching nodes according to the resource planning strategy and the real-time performance metric data. The result of the real-time resource scheduling affects the real-time performance metric data. The load evaluation module is configured to determine the actual service bandwidth of the edge caching nodes based on the real-time performance metric data, determine the target service bandwidth of the edge caching nodes based on the resource planning strategy, perform a load evaluation on the edge caching nodes by combining the actual service bandwidth and the target service bandwidth, and adjust the resource planning strategy based on the load evaluation result and send the adjusted resource planning strategy to the resource planning module to narrow the difference between the actual service bandwidth and the target service bandwidth.

[0019] Third aspect, the present invention provides a computer device, including: a memory and a processor, which are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to execute the edge caching node bandwidth resource planning method according to the first aspect or any corresponding embodiment thereof.

[0020] Fourth aspect, the present invention provides a computer-readable storage medium, on which computer instructions are stored. The computer instructions are used to cause a computer to execute the edge caching node bandwidth resource planning method according to the first aspect or any corresponding embodiment thereof.

[0021] Fifth aspect, the present invention provides a computer program product, including computer instructions, which are used to cause a computer to execute the edge caching node bandwidth resource planning method according to the first aspect or any corresponding embodiment thereof. Description of the Drawings

[0022] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0023] Figure 1 is a schematic diagram of a bandwidth resource planning system for an edge caching node according to an embodiment of the present invention;

[0024] Figure 2 is a schematic diagram of a group of edge caching servers according to an embodiment of the present invention;

[0025] Figure 3 is a flowchart of a method for planning the bandwidth resources of an edge caching node according to an embodiment of the present invention;

[0026] Figure 4 is a flowchart of another method for planning the bandwidth resources of an edge caching node according to an embodiment of the present invention;

[0027] Figure 5 is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. Specific Embodiments

[0028] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

[0029] On the premise of meeting the bandwidth requirements of customer services in each region, the CDN scheduling system will schedule CDN nodes based on various factors. Generally speaking, a CDN node can serve multiple services simultaneously, and the CDN scheduling system will, according to performance indicators such as the bandwidth usage rate, QPS, and health status of the node, as well as the constraint conditions of the service, transfer the service bandwidth amount that exceeds the node's bearable range to other nodes with bandwidth redundancy and meeting the conditions to achieve load balancing. Usually, the CDN support system adopts a multi-level architecture including an edge layer, a parent layer, and a source station. When a customer fails to hit the content from the edge node, they will obtain it from the parent layer. If the parent layer also fails to hit, they will obtain it from the source station.

[0030] The main optimization goal of the CDN scheduling system is the load balancing of edge nodes, mainly based on the bandwidth redundancy of nodes to determine whether the amount of business undertaken by nodes needs to be increased or decreased. However, due to different business characteristics, when a node may not receive the expected amount of business, quality problems will occur, resulting in the CDN scheduling system frequently removing the node. In addition, the existing CDN node bandwidth resource planning leads to a large difference between the actual bandwidth amount and the planned bandwidth amount of the node, increasing the decision-making burden of the scheduling system, affecting the planning of other nodes, and affecting the load balancing of the entire CDN system. At the same time, the existing technology selects resource cross-service coverage in the scheduling decision link, which may affect the efficiency of the scheduling decision.

[0031] Therefore, in view of the problems such as the reduced hit rate and the increased return-to-parent bandwidth caused by less consideration of nodes being frequently scheduled to different services, the present application proposes an edge cache node bandwidth resource planning system and a corresponding edge cache node bandwidth resource planning method, which can, on the basis of the current scheduling system, reduce the frequency of nodes being scheduled to different service domains, reduce the return-to-parent bandwidth, and reduce the bandwidth cost of the return-to-parent. And reduce unnecessary removal of nodes, reduce the burden of scheduling decisions, and make the resource coverage plan more reasonable.

[0032] According to an embodiment of the present invention, there is provided an edge cache node bandwidth resource planning system. Please refer to Figure 1 , Figure 1 which is a schematic diagram of an edge cache node bandwidth resource planning system according to an embodiment of the present invention. As Figure 1 shown, the edge cache node bandwidth resource planning system of the present application includes a plurality of edge cache nodes, a data center, a resource planning module, a scheduling decision module, and a load evaluation module. Among them, the data center is connected to a plurality of edge cache nodes, a resource planning module, a scheduling decision module, and a load evaluation module. The resource planning module is connected to the scheduling decision module and the load evaluation module.

[0033] Among them, each edge cache node includes a plurality of edge cache server groups, and each edge cache server group includes a plurality of edge cache servers and a data collection module corresponding to each edge cache server. Please refer to Figure 2 , Figure 2 which is a schematic diagram of an edge cache server group according to an embodiment of the present invention.

[0034] Among them, the edge cache server is a content delivery network edge cache server, the edge cache server group is a content delivery network edge cache server group, and the edge cache node is a content delivery network edge cache node. Among them, CDN is a distributed network architecture designed to improve user experience and network performance by delivering content to nodes closer to users.

[0035] The data acquisition module respectively collects the monitoring data and network conditions of the corresponding edge cache servers, including performance metric data such as traffic bandwidth data, the number of packets received and sent per second (Packets Per Second, PPS), the number of requests processed per second (Query PerSecond, QPS), and the health status of the machine. Among them, the health status of the machine includes performance metrics such as CPU usage, memory usage, disk load, network card usage, and response time of common ports.

[0036] Integrate the performance metric data of multiple edge cache servers to obtain the performance metric data of the edge cache server group, and integrate the performance metric data of the edge cache server group to obtain the performance metric data of the edge cache node.

[0037] In one implementation, the bandwidth resource planning takes the edge cache server group as the planning unit; in another implementation, the bandwidth resource planning takes the edge cache node as the planning unit. Among them, the bandwidth resource planning is the same, and the following will explain it with the edge cache node as the planning unit.

[0038] Furthermore, the edge cache node sends the performance metric data to the data center.

[0039] Among them, the data center is responsible for collecting the scheduling data of the entire system and providing a data basis for modules such as bandwidth resource planning, scheduling decision-making, and overload assessment.

[0040] Specifically, the data center obtains the historical data of the business domain name, the real-time performance metric data and resource attributes of the edge cache node. Among them, the historical data of the business domain name includes historical metric data, customer demand data, and service quality data, and the resource attributes include location information, network quality information, hardware information, hardware constraint information, and business constraint information.

[0041] Furthermore, the data center performs preliminary cleaning on the data and processes the data into data directly available to the resource planning module, the scheduling decision module, and the load assessment module.

[0042] Furthermore, the data center sends the historical data to the resource planning module. The historical data is mainly used for the preliminary planning of resources and the analysis of business service conditions.

[0043] Furthermore, the data center sends the real-time performance metric data, resource attributes, and historical data to the scheduling decision module and the load assessment module. It can be understood that the performance metric data of the edge cache node is updated in real time, and the data center sends the real-time performance metric data in real time. The real-time data provides a reference basis for real-time scheduling decisions.

[0044] Among them, the resource planning module initially plans the services and regions covered by the edge cache node resources based on the historical data of the service domain names provided by the scheduling data center and the resource attributes of the edge cache nodes, and dynamically adjusts the resource coverage according to the real-time response of the load evaluation module.

[0045] Specifically, the resource planning module is used to determine the planned service domain and the planned service bandwidth of the edge cache node based on the historical data of the service domain name and the resource attributes of the edge cache node. The planned service domain includes the planned service domain name and the planned service domain region.

[0046] In one implementation, based on historical data such as domain name service bandwidth, number of requests, download speed, average file size, traffic hit rate, etc., the hardware information of the edge cache node such as CPU usage rate, memory usage rate, disk load, network card usage rate, common port response time, etc., and the constraint conditions measured by node stress testing, determine the service domain of the edge cache node.

[0047] Specifically, determine the service domain names that the edge cache node can undertake according to historical data and hardware data. Exemplarily, if the download speed of edge cache node a is less than the required download speed of service domain name b, then the service domain of edge cache node a does not include service domain name b.

[0048] In one implementation, based on the location information, network quality information, and customer demand data of the edge cache node, determine the service domain area range of the edge cache node.

[0049] Specifically, the edge cache node preferentially undertakes service domain names that are closer. When the network quality is higher, the service domain area range that can be undertaken is larger. When the network quality is lower, the service domain area range that can be undertaken is smaller.

[0050] Based on the hardware constraint information and service constraint information of the edge cache node, determine the available bandwidth that the edge cache node can undertake.

[0051] Specifically, the service constraint information includes service quality constraint information, service cost constraint information, adjacent area constraint, and customer requirement constraint. The available bandwidth that the edge cache node can undertake needs to meet the hardware information and service constraint information.

[0052] Combine the service domain, service domain area range, available bandwidth, and current service demand of the edge cache node to determine the resource planning strategy.

[0053] Specifically, allocate node resources according to the service priority and node priority. That is, preferentially select service domain names with higher service priority for planning. For the same service domain name, preferentially select nodes with higher node priority for planning.

[0054] Corresponding to the business domain names planned currently according to the priority, for the same business domain area of the same business domain name, select edge cache nodes in sequence according to the priority, and determine the planned service bandwidth of the edge cache node under the planned business domain according to the available bandwidth capacity of the edge cache node. Among them, the planned service bandwidth does not exceed the available bandwidth capacity. Understandably, when the required bandwidth of the business domain name is greater than the available bandwidth capacity, the available bandwidth capacity is preferentially used as the planned service bandwidth.

[0055] Use the above method to obtain the resource planning strategy corresponding to the current edge cache node. By comprehensively considering the performance data and node basic data of the edge cache node, a resource coverage plan that meets various requirements is determined for the edge cache node, which can improve the rationality of the preliminary resource planning.

[0056] Furthermore, the resource planning module sends the resource planning strategy to the scheduling decision module and the load evaluation module.

[0057] Among them, the scheduling decision module is used to perform real-time resource scheduling on the edge cache node according to the resource planning strategy and the real-time performance index data. Understandably, the result of the real-time resource scheduling can affect the real-time performance index data.

[0058] Specifically, the scheduling decision module first pulls the planned service bandwidth corresponding to the planned business domain in the resource planning strategy to the corresponding edge cache node. Further, real-time scheduling of the bandwidth traffic is performed according to the real-time data provided by the data center and the resource planning strategy.

[0059] In one implementation, a resource scheduling strategy is determined by weighing the resource cost and the service quality data. If the real-time performance index data of the current edge cache node does not meet the index requirements, the service bandwidth of the current edge cache node is pulled to the target edge cache node according to the resource scheduling strategy. Among them, the real-time performance index data of the target edge cache node meets the index requirements, there is redundant service bandwidth in the target edge cache node, and the planned business domain of the target edge cache node is the same as that of the current edge cache node.

[0060] Perform preliminary resource scheduling according to the resource planning strategy, so that the resource scheduling meets the scheduling requirements, can control the service volume covered by each node, and reduce the decision-making burden of the resource scheduling decision system.

[0061] Among them, the resource scheduling strategy includes a threshold scheduling strategy, a quality scheduling strategy, a cost scheduling strategy, etc.

[0062] In a possible implementation, when it is monitored that the business metrics of a node exceed the set threshold, the scheduling decision module will, according to the defined resource scheduling policy, divert traffic to other acceptable nodes. By reducing the node traffic, it ensures that metrics such as the node's bandwidth, PPS, QPS, and health status meet the requirements. And through the negative feedback regulation mechanism of the control algorithm, the node bandwidth is controlled near the scheduling line. If the bandwidth exceeds the scheduling line by a large amount and the overshoot is large, it will lead to a large amplitude and a large amount of diverted traffic. At this time, the node bandwidth of the part of the traffic diverted in will also have a large amplitude, resulting in an increase in the node resource cost. Therefore, it is necessary to control the business volume covered by each node in the resource planning module to reduce the decision-making burden on the scheduling decision system.

[0063] In a possible implementation, if the cache server has network quality problems such as high packet loss rate and high latency, or machine performance problems such as high CPU (Central Processing Unit) and high disk I / O wait, the traffic will be diverted from the faulty machine to a healthy machine with bandwidth redundancy.

[0064] In a possible implementation, on the premise of meeting customer requirements and service quality, with the lowest net cost and the highest benefit as the optimization goal, the cost is reduced as much as possible. Among them, the customer benefit coefficient and the estimated customer business volume can be used to estimate the benefit value, and the cost coefficient of the node resources can be used to estimate the cost value. The difference between the two is the net benefit.

[0065] Among them, the load evaluation module combines the resource planning policy provided by the resource planning module and the real-time data provided by the data center. According to the difference between the real-time service bandwidth of the node service and the planned bandwidth, combined with the business monitoring metrics on the node and the health status of the node, it evaluates the load situation of the node through the resource planning policy adjustment method.

[0066] Specifically, the load evaluation module is used to determine the actual service bandwidth of the edge cache node based on the real-time performance metric data; determine the target service bandwidth of the edge cache node based on the resource planning policy; evaluate the load of the edge cache node by combining the actual service bandwidth and the target service bandwidth, and adjust the resource planning policy based on the load evaluation result.

[0067] In one implementation, the edge cache node provides services in units of edge cache servers, and the edge cache server provides services to the business domain name in the form of a virtual IP externally.

[0068] Among them, the virtual IP (Virtual IP Address, VIP) is also called the upstream traffic IP. One virtual IP can be mapped to multiple actual IPs, and load balancing between multiple actual IPs can be achieved through VIP scheduling.

[0069] In one implementation, the present application analyzes the same business domain area for the same business domain name. The virtual IPs actually occupied by each business domain name in each business domain area are obtained from the real-time performance index data, and the actual service bandwidth of the virtual IPs is obtained.

[0070] The virtual IPs planned to be occupied by each business domain name in each business domain area and the real-time service bandwidth of each business domain name in each business domain area are obtained from the resource planning strategy. The real-time service bandwidth is evenly distributed on the virtual IPs to obtain the planned service bandwidth of the virtual IPs.

[0071] Specifically, the planned bandwidth of the VIP under the business domain name area = the real-time bandwidth of the business domain name area / the number of VIPs under the business domain name area.

[0072] Exemplarily, for each business domain name and business domain area, given the total real-time bandwidth of the business domain name in each business domain area (distinguishing between IPv4 and IPv6) and the planned number of VIPs, the planned bandwidth of each VIP is estimated. If a business domain name is served by n IPv4s and m IPv6s in a specified business domain area, assuming that the bandwidth of the business domain name in the business domain area can be evenly distributed to each VIP, if the business domain name has a bandwidth of a + b in the specified area (where the bandwidth of IPv4 is a and the bandwidth of IPv6 is b), then the bandwidth of each IPv4 in this area of the business domain name is a / n, and the bandwidth of each IPv6 is b / m.

[0073] In this implementation, taking the virtual IP as the analysis unit, comparing the difference between the actual service bandwidth and the planned bandwidth of the virtual IP to determine the load situation of the virtual IP, and further calculating the load situation of the edge cache node in combination with the load situations of multiple virtual IPs, a method for adjusting the resource planning strategy according to the bandwidth difference can be realized, which helps to improve the resource utilization rate of the node. And ensuring that the services on the node are as concentrated as possible can reduce the increase in the return parent bandwidth caused by caching the service bandwidths of multiple services.

[0074] Furthermore, for the same business area of the same business domain name, the load data of each virtual IP is calculated based on the actual service bandwidth and the planned service bandwidth. The load data includes overloaded bandwidth, underloaded bandwidth, and overloaded bandwidth.

[0075] In one implementation, when there is an actual service bandwidth for a virtual IP but no planned service bandwidth, the actual service bandwidth is taken as the overloaded bandwidth of the virtual IP; when there is an actual service bandwidth for the virtual IP and there is also a planned service bandwidth, and the actual service bandwidth corresponding to the virtual IP is greater than the planned service bandwidth, the difference between the actual service bandwidth and the planned service bandwidth is taken as the overloaded bandwidth of the virtual IP; when there is an actual service bandwidth for the virtual IP and there is also a planned service bandwidth, and the actual service bandwidth corresponding to the virtual IP is less than the planned service bandwidth, the difference between the actual service bandwidth and the planned service bandwidth is taken as the underloaded bandwidth of the virtual IP.

[0076] Specifically, the overloaded bandwidth refers to the bandwidth outside the plan in the specified area for the service domain name carried by the VIP. The calculation method is: overloaded bandwidth = max(real-time bandwidth - planned bandwidth, 0).

[0077] The underloaded bandwidth refers to the planned bandwidth not undertaken by the VIP in the specified area of this service domain name. The calculation method is: underloaded bandwidth = max(planned bandwidth - real-time bandwidth, 0).

[0078] The overloaded bandwidth refers to the bandwidth of the VIP that is not within the scope of the planned service domain name.

[0079] Among them, if a certain VIP has real-time bandwidth but is not within the coverage plan, the overloaded bandwidth and the underloaded bandwidth are calculated with the planned bandwidth being 0.

[0080] Furthermore, the load data of multiple virtual IPs are accumulated to calculate the load data of the edge cache node, and the resource planning strategy is adjusted based on the load data of the edge cache node.

[0081] In one implementation, when the overloaded bandwidth of the edge cache node is greater than the first bandwidth threshold, the planned service bandwidth of the edge cache node is reduced; when the underloaded bandwidth of the edge cache node is greater than the second bandwidth threshold, the planned service bandwidth of the edge cache node is increased.

[0082] Furthermore, the load evaluation module sends the adjusted resource planning strategy to the resource planning module.

[0083] Furthermore, the resource planning module further adjusts the resource coverage according to the response feedback of the load evaluation module.

[0084] Among them, the scheduling decision module schedules CDN nodes based on factors such as quality and cost, which may cause CDN edge nodes to be frequently scheduled to different services, resulting in problems such as low traffic hit rate, increased return-to-parent bandwidth, and increased bandwidth cost. The overload evaluation module combines the preliminary coverage of CDN nodes and real-time scheduling data to evaluate the load of nodes, analyze the rationality of the coverage plan, and adjust the coverage plan of nodes according to the evaluation results. After receiving the adjustment plan, the resource planning module of this application notifies the scheduling decision module according to the new plan, and the scheduling decision module schedules according to the latest planning situation. The entire process forms a negative feedback closed-loop regulation system, which can gradually reduce the deviation between the actual bandwidth and the planned bandwidth. Specifically, when the actual bandwidth is greater than the planned bandwidth, reducing the deviation between the actual bandwidth and the planned bandwidth helps to relieve the decision-making pressure of the scheduling decision module and reduce the oscillation caused by excessive transfer volume; when the actual bandwidth is less than the planned bandwidth, reducing the deviation helps to improve the resource utilization rate of nodes. And ensuring that the services on the nodes are as concentrated as possible can reduce the increase in return-to-parent bandwidth caused by caching multiple service bandwidths.

[0085] According to an embodiment of the present invention, there is provided an embodiment of a method for planning bandwidth resources of an edge cache node. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0086] In this embodiment, a method for planning bandwidth resources of an edge cache node is provided, which can be used in the above-mentioned edge cache node bandwidth resource planning system. Figure 3 It is a flowchart of a method for planning bandwidth resources of an edge cache node according to an embodiment of the present invention. It should be noted that if there are substantially the same results, this embodiment is not limited to Figure 3 the process order shown. As Figure 3 shown, the process includes the following steps:

[0087] Step S301, determine the resource planning strategy of the edge cache node by combining the historical data of the service domain name and the resource attributes of the edge cache node.

[0088] Among them, the resource planning strategy includes the planned service domain and the planned service bandwidth of the edge cache node, and the planned service domain includes the planned service domain name and the planned service domain area.

[0089] Step S302, obtain the real-time performance index data of the edge cache node.

[0090] Step S303, perform real-time resource scheduling on the edge cache node according to the resource planning strategy and the real-time performance index data.

[0091] Among them, the result of real-time resource scheduling affects the real-time performance metric data.

[0092] Step S304: Determine the actual service bandwidth of the edge cache node based on the real-time performance metric data; determine the target service bandwidth of the edge cache node based on the resource planning strategy; evaluate the load of the edge cache node by combining the actual service bandwidth and the target service bandwidth, and adjust the resource planning strategy based on the load evaluation result.

[0093] The method for planning the bandwidth resources of the edge cache node provided in this embodiment makes a preliminary plan based on historical data and resource attributes to obtain the planned services and planned bandwidth of the edge cache node, and compares and analyzes the load of the edge cache node according to the planned situation and the actual scheduling situation, and adjusts the resource planning strategy in real time to reduce the deviation between the planned situation and the actual scheduling, reduce the frequency of the edge cache node being scheduled to different services, reasonably plan the bandwidth resources of the edge cache node, reduce the backhaul bandwidth, reduce the backhaul cost, further improve the hit rate. At the same time, by adjusting the resource planning strategy, it is possible to optimize the resource coverage plan, reduce the burden on the scheduling decision-making link, and improve the rationality of resource planning.

[0094] In this embodiment, a method for planning the bandwidth resources of an edge cache node is provided, which can be used in the above-mentioned edge cache node bandwidth resource planning system. Figure 4 It is a flowchart of another method for planning the bandwidth resources of an edge cache node according to an embodiment of the present invention. It should be noted that if there are substantially the same results, this embodiment does not Figure 4 be limited to the shown process sequence. As Figure 4 shown, the process includes the following steps:

[0095] Step S401: Determine the resource planning strategy of the edge cache node by combining the historical data of the service domain name and the resource attributes of the edge cache node.

[0096] The historical data of the service domain name includes historical metric data, customer demand data, and service quality data, and the resource attributes include location information, network quality information, hardware information, hardware constraint information, service constraint information, and resource cost data.

[0097] Specifically, the above step S401 includes:

[0098] Step S4011: Determine the serviceable service domains of the edge cache node based on the historical metric data of the service domain name and the hardware information of the edge cache node.

[0099] Step S4012: Determine the range of serviceable business domain areas for the edge caching node based on the location information of the edge caching node, network quality information, and customer demand data of the business domain name.

[0100] Step S4013: Determine the available bandwidth capacity that the edge caching node can undertake based on the hardware constraint information and business constraint information of the edge caching node.

[0101] Step S4014: Determine the resource planning strategy by combining the business domain of the edge caching node, the range of serviceable business domain areas, the available bandwidth capacity, and the current business requirements.

[0102] Step S402: Obtain the real-time performance metric data of the edge caching node.

[0103] Step S403: Perform real-time resource scheduling on the edge caching node according to the resource planning strategy and real-time performance metric data.

[0104] Among them, the result of real-time resource scheduling affects the real-time performance metric data.

[0105] Specifically, the above Step S403 includes:

[0106] Step S4031: Weigh the resource cost and service quality data to determine the resource scheduling strategy.

[0107] Step S4032: Pull the planned service bandwidth corresponding to the planned business domain to the corresponding edge caching node.

[0108] Step S4033: If the real-time performance metric data of the current edge caching node does not meet the metric requirements, pull the service bandwidth of the current edge caching node to the target edge caching node according to the resource scheduling strategy.

[0109] Among them, the real-time performance metric data of the target edge caching node meets the metric requirements, there is redundant service bandwidth in the target edge caching node, and the planned business domains of the target edge caching node and the current edge caching node are the same.

[0110] Step S404: Determine the actual service bandwidth of the edge caching node based on the real-time performance metric data; determine the target service bandwidth of the edge caching node based on the resource planning strategy; conduct a load assessment on the edge caching node by combining the actual service bandwidth and the target service bandwidth, and adjust the resource planning strategy based on the load assessment result.

[0111] Specifically, the above Step S404 includes:

[0112] Step S4041: Determine the virtual IPs actually occupied by each business domain name in each business domain area based on the real-time performance metric data, and obtain the actual service bandwidth of the virtual IPs.

[0113] Step S4042: Determine the virtual IPs planned to be occupied by each service domain name within each service domain area based on the resource planning strategy, and obtain the planned service bandwidth of the virtual IPs.

[0114] Specifically, the above step S4042 includes:

[0115] Step a1: Determine the edge cache nodes and real-time service bandwidth planned to be occupied by each service domain name within each service area based on the resource planning strategy.

[0116] Step a2: Obtain the number of virtual IPs planned to be occupied in the edge cache nodes.

[0117] Step a3: Evenly distribute the planned service bandwidth of the planned virtual IPs based on the real-time service bandwidth and the number of virtual IPs.

[0118] Step S4043: Calculate the load data of each virtual IP for the same service area of the same service domain name based on the actual service bandwidth and the planned service bandwidth.

[0119] Among them, the load data includes overloaded bandwidth, underloaded bandwidth, and overcapacity bandwidth.

[0120] Specifically, the above step S4043 includes:

[0121] Step b1: When a virtual IP has actual service bandwidth but no planned service bandwidth, use the actual service bandwidth as the overcapacity bandwidth of the virtual IP.

[0122] Step b2: When a virtual IP has actual service bandwidth and planned service bandwidth, and the actual service bandwidth corresponding to the virtual IP is greater than the planned service bandwidth, use the difference between the actual service bandwidth and the planned service bandwidth as the overloaded bandwidth of the virtual IP.

[0123] Step b3: When a virtual IP has actual service bandwidth and planned service bandwidth, and the actual service bandwidth corresponding to the virtual IP is less than the planned service bandwidth, use the difference between the actual service bandwidth and the planned service bandwidth as the underloaded bandwidth of the virtual IP.

[0124] Step S4044: Accumulate the load data of multiple virtual IPs to calculate the load data of the edge cache node.

[0125] Specifically, the above step S4044 includes:

[0126] Step c1: When the overloaded bandwidth of the edge cache node is greater than the first bandwidth threshold, reduce the planned service bandwidth of the edge cache node;

[0127] Step c2, when the underloaded bandwidth of the edge caching node is greater than the second bandwidth threshold, increase the planned service bandwidth of the edge caching node.

[0128] Step S4045, adjust the resource planning strategy based on the load data of the edge caching node.

[0129] In a specific implementation, the method of step S404 is specifically as follows:

[0130] Step d1, obtain the information of the primary resource coverage planning from the primary resource planning module.

[0131] Step d2, obtain the real-time bandwidth at the service domain name-region-VIP granularity from the data center, and calculate the planned bandwidth of each VIP.

[0132] The planned bandwidth of the VIPs under the service domain name region = the real-time bandwidth of the service domain name region / the number of VIPs under the service domain name region.

[0133] Step d3, count the overloaded bandwidth, underloaded bandwidth, and overloaded bandwidth of each VIP.

[0134] Step d4, accumulate the data of the VIPs to converge the overloaded bandwidth and underloaded bandwidth at the granularities of the CDN edge servers, CDN edge server groups, CDN nodes, etc. where the VIPs are located.

[0135] Step d5, for the CDN edge server group, if the overloaded bandwidth is large, it means that the server group cannot undertake so much traffic volume, and the level of the primary resource planning needs to be reduced. If the underloaded bandwidth is large, it means that the redundancy of the server group is relatively sufficient, and the level of the primary resource planning can be appropriately increased. If the overloaded bandwidth appears in the server group, it means that the server group is frequently scheduled to multiple unplanned service domain names, which will lead to a decrease in the traffic cache hit rate and an increase in the backhaul bandwidth.

[0136] Step d6, adjust the resource coverage plan according to the overloading evaluation in the previous step, output the adjustment plan, and send the adjustment plan to the primary resource planning module.

[0137] After receiving the adjustment plan, the resource planning module notifies the scheduling system according to the new coverage plan, and the scheduling system performs scheduling according to the latest coverage planning situation. The entire process forms a negative feedback closed-loop regulation system, which can gradually reduce the deviation between the actual bandwidth and the planned bandwidth.

[0138] This application analyzes the deviation between the bandwidth of the actual service domain names borne by the CDN edge nodes and the planned bandwidth, analyzes the rationality of resource coverage planning, reduces the frequent transfer volume of the entire CDN system, reduces the backhaul bandwidth volume, and reduces the backhaul bandwidth cost. Based on the existing scheduling decision scheme, improvements are made. By comparing the effective results after scheduling decisions with the original plan, a negative feedback mechanism for reducing deviations is adopted to optimize resource coverage planning, reduce the burden on the scheduling decision-making process, and improve the rationality of resource planning.

[0139] An embodiment of the present invention further provides a computer device having the above-mentioned Figure 1 edge cache node bandwidth resource planning system shown.

[0140] Please refer to Figure 5 , Figure 5 which is a schematic structural diagram of a computer device provided by an optional embodiment of the present invention. As Figure 5 shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including a high-speed interface and a low-speed interface. Each component communicates with each other using different buses and can be installed on a common motherboard or installed in other ways as needed. The processor can process instructions executed within the computer device, including instructions stored in the memory or on the memory to display graphical information of the GUI on an external input / output device (such as a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (such as a server array, a set of blade servers, or a multi-processor system). Figure 5 In

[0141]

[0142]

[0143]

[0143] The memory 20 may include a program storage area and a data storage area. The program storage area may store an operating system and application programs required for at least one function. The data storage area may store data created according to the use of the computer device and the like. In addition, the memory 20 may include a high-speed random access memory and may also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some alternative embodiments, the memory 20 may optionally include a memory remotely provided with respect to the processor 10, and these remote memories may be connected to the computer device through a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0144] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk, or a solid-state drive; the memory 20 may further include a combination of the above types of memories.

[0145] The computer device further includes an input device 30 and an output device 40. The processor 10, the memory 20, the input device 30, and the output device 40 may be connected through a bus or other means. Figure 5 Taking connection through a bus as an example.

[0146] The input device 30 may receive input digital or character information and generate key signal inputs related to the user settings and function controls of the computer device, such as a touch screen, a keypad, a mouse, a trackpad, a touchpad, a pointing stick, one or more mouse buttons, a trackball, a joystick, etc. The output device 40 may include a display device, an auxiliary lighting device (e.g., an LED), and a haptic feedback device (e.g., a vibration motor), etc. The above-mentioned display device includes but is not limited to a liquid crystal display, a light-emitting diode, a display, and a plasma display. In some alternative embodiments, the display device may be a touch screen.

[0147] Embodiments of the present invention also provide a computer-readable storage medium. The method according to the embodiments of the present invention can be implemented in hardware, firmware, or be implemented as computer code that can be recorded on a storage medium, or be implemented as computer code that is originally stored in a remote storage medium or a non-transitory machine-readable storage medium and downloaded through a network and will be stored in a local storage medium, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk, or a solid-state drive, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code, and when the software or computer code is accessed and executed by the computer, the processor, or the hardware, the method shown in the above embodiments is implemented.

[0148] A part of the present invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. Those skilled in the art should understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Herein, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to the computer.

[0149] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for planning bandwidth resources of edge caching nodes, characterized in that, The method includes: Determining a resource planning strategy for the edge cache node by combining historical data of the service domain name and resource attributes of the edge cache node, where the resource planning strategy includes a planned service domain and a planned service bandwidth for the edge cache node, the planned service domain includes a planned service domain name and a planned service domain area, and the edge cache node is an edge cache node of a content delivery network; Obtaining real-time performance metric data of the edge cache node; Performing real-time resource scheduling on the edge cache node according to the resource planning strategy and the real-time performance metric data, where the result of the real-time resource scheduling affects the real-time performance metric data; Determining the actual service bandwidth of the edge cache node based on the real-time performance metric data; determining the target service bandwidth of the edge cache node based on the resource planning strategy; evaluating the load of the edge cache node by combining the actual service bandwidth and the target service bandwidth, and adjusting the resource planning strategy based on the load evaluation result to narrow the difference between the actual service bandwidth and the target service bandwidth.

2. The method for planning the bandwidth resources of the edge caching node according to claim 1, wherein The historical data of the service domain name includes historical metric data and customer demand data, and the resource attributes of the edge cache node include location information, network quality information, hardware information, hardware constraint information, and service constraint information. The determining of the resource planning strategy for the edge cache node by combining the historical data of the service domain name and the resource attributes of the edge cache node includes: Determining the service domains that the edge cache node can serve based on the historical metric data of the service domain name and the hardware information of the edge cache node; Determining the range of the service domain area that the edge cache node can serve based on the location information, the network quality information of the edge cache node, and the customer demand data of the service domain name; Determining the available bandwidth capacity of the edge cache node based on the hardware constraint information and the service constraint information of the edge cache node; Combining the service domain, the service domain area range, the available bandwidth capacity of the edge cache node, and the current service demand to determine the resource planning strategy.

3. The method for planning the bandwidth resources of the edge caching node according to claim 1, wherein The historical data of the service domain name includes service quality data, and the resource attributes of the edge cache node include resource cost data. The performing of real-time resource scheduling on the edge cache node according to the resource planning strategy and the real-time performance metric data includes: Weighing the resource cost and the service quality data to determine a resource scheduling strategy; Trafficking the planned service bandwidth corresponding to the planned service domain to the corresponding edge cache node; When the real-time performance metric data of the current edge cache node does not meet the metric requirements, trafficking the service bandwidth of the current edge cache node to a target edge cache node according to the resource scheduling strategy, where the real-time performance metric data of the target edge cache node meets the metric requirements, there is redundant service bandwidth in the target edge cache node, and the planned service domains of the target edge cache node and the current edge cache node are the same.

4. The method for planning the bandwidth resources of an edge caching node according to claim 1, wherein The edge caching node provides service through multiple virtual IPs. Determining the actual service bandwidth of the edge caching node based on the real-time performance metric data includes: Determining the virtual IPs actually occupied by each service domain name within each service domain area based on the real-time performance metric data, and obtaining the actual service bandwidth amount of the virtual IPs; Determining the target service bandwidth of the edge caching node based on the resource planning strategy includes: Determining the virtual IPs planned to be occupied by each service domain name within each service area based on the resource planning strategy, and obtaining the planned service bandwidth amount of the virtual IPs; Combining the actual service bandwidth and the target service bandwidth to evaluate the load of the edge caching node, and adjusting the resource planning strategy based on the load evaluation result includes: Calculating the load data of each virtual IP based on the actual service bandwidth amount and the planned service bandwidth amount for the same service area of the same service domain name; Accumulating the load data of multiple virtual IPs to calculate the load data of the edge caching node; the load data includes overloaded bandwidth amount, underloaded bandwidth amount, and overcapacity bandwidth amount; Adjusting the resource planning strategy based on the load data of the edge caching node.

5. The method for planning the bandwidth resources of an edge caching node according to claim 4, wherein Determining the planned virtual IPs planned to be occupied by each service domain name within each service area based on the resource planning strategy, and obtaining the planned service bandwidth amount of the planned virtual IPs includes: Determining the edge caching nodes planned to be occupied by each service domain name within each service area and the real-time service bandwidth amount based on the resource planning strategy; Obtaining the number of virtual IPs planned to be occupied in the edge caching node; Evenly distributing the planned service bandwidth amount of the planned virtual IPs based on the real-time service bandwidth amount and the number of virtual IPs.

6. The method for planning the bandwidth resources of an edge caching node according to claim 4, wherein Calculating the load data of each virtual IP based on the actual service bandwidth amount and the planned service bandwidth amount includes: When the virtual IP has the actual service bandwidth amount but no planned service bandwidth amount, then taking the actual service bandwidth amount as the overcapacity bandwidth amount of the virtual IP; When the virtual IP has the actual service bandwidth amount and also has the planned service bandwidth amount, and the actual service bandwidth amount corresponding to the virtual IP is greater than the planned service bandwidth amount, then taking the difference between the actual service bandwidth amount and the planned service bandwidth amount as the overloaded bandwidth amount of the virtual IP; When the virtual IP has the actual service bandwidth amount and also has the planned service bandwidth amount, and the actual service bandwidth amount corresponding to the virtual IP is less than the planned service bandwidth amount, then taking the difference between the actual service bandwidth amount and the planned service bandwidth amount as the underloaded bandwidth amount of the virtual IP.

7. The method for planning bandwidth resources of an edge caching node according to claim 6, wherein Adjusting the resource planning strategy based on the load data of the edge caching node includes: When the overloaded bandwidth of the edge caching node is greater than the first bandwidth threshold, reducing the planned service bandwidth amount of the edge caching node; When the underloaded bandwidth of the edge caching node is greater than the second bandwidth threshold, increasing the planned service bandwidth amount of the edge caching node.

8. An edge caching node bandwidth resource planning system, characterized in that, The system includes multiple edge caching nodes, a data center, a resource planning module, a scheduling decision module, and a load evaluation module; the data center is connected to the multiple edge caching nodes, the resource planning module, the scheduling decision module, and the load evaluation module, and the resource planning module is connected to the scheduling decision module and the load evaluation module; the edge caching nodes are content delivery network edge caching nodes; The data center is configured to obtain the real-time performance metric data, resource attributes, and historical data of the business domain names of the edge caching nodes, and send the real-time performance metric data, the resource attributes, and the historical data to the scheduling decision module and the load evaluation module; The resource planning module is configured to determine the resource planning strategy of the edge caching nodes by combining the historical data and the resource attributes, where the resource planning strategy includes the planned business domain and the planned business bandwidth of the edge caching nodes, and the planned business domain includes the planned business domain name and the planned business domain area; The scheduling decision module is configured to perform real-time resource scheduling on the edge caching nodes according to the resource planning strategy and the real-time performance metric data, where the result of the real-time resource scheduling affects the real-time performance metric data; The load evaluation module is configured to determine the actual business bandwidth of the edge caching nodes based on the real-time performance metric data; determine the target business bandwidth of the edge caching nodes based on the resource planning strategy; perform load evaluation on the edge caching nodes by combining the actual business bandwidth and the target business bandwidth, and adjust the resource planning strategy based on the load evaluation result, and send the adjusted resource planning strategy to the resource planning module to narrow the difference between the actual business bandwidth and the target business bandwidth.

9. A computer device, characterized in that, Comprising: A memory and a processor, the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to execute the edge caching node bandwidth resource planning method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, Computer instructions are stored on the computer-readable storage medium, and the computer instructions are used to cause a computer to execute the edge caching node bandwidth resource planning method according to any one of claims 1 to 7.

Images