CDN NODE LOAD CONFIGURATION METHOD, APPARATUS, ELECTRONIC DEVICE AND ENVIRONMENT

Abstract

The current application provides an electronic device and an environment with a load configuration method and apparatus for a CDN node. The method includes: obtaining the traffic characteristics of a target CP device where the target CP device serves as the server of a CDN node; taking as a constraint that the IO consumption of the target CP device does not exceed the upper limit of the IO throughput, and determining the maximum service traffic that can be reached by the target CP device based on the traffic characteristics of the target CP device; determining the reference service traffic of the target CP device based on the maximum service traffic that can be reached by the target CP device and the hardware load state of the target CP device; and allocating the CDN node's load to the target CP device according to the reference service traffic of the target CP device.

Description

1 TARIFF Load configuration method, apparatus, electronic device and for CDN node. ENVIRONMENT CROSS-REFERENCE TO RELEVANT REGULATIONS This application was filed with the Chinese Patent Office on December 16, 2022, and its entire content is referenced in Article 5. Included here is the "Load Configuration Method and Apparatus for CDN Node". Chinese patent application number 202211625759.X titled "ELECTRONIC DEVICE AND MEDIUM" He is claiming his right of first refusal. TECHNICAL FIELD The current application relates to the mobile communications technical field, specifically a CDN node, an electronic device, and 10 It relates to a load configuration method and apparatus for an environment. BACKGROUND OF THE INVENTION A content delivery network (CDN) is an advanced traffic distribution network built on the existing Internet infrastructure. and a user's access is routed to the nearest point to the user via a global load balancing technology. It can redirect to a streaming media server that is functioning normally. The streaming media server is on network 15 It responds directly to the user's request to alleviate congestion on the streaming media server. If the content the user wants to access is not available, the streaming media server will configure the original content accordingly. It automatically retrieves the relevant content from the server and provides the same content to the user. This process is called "origin-pull". It is called. Currently, CDNs are very large, serving numerous downstream Internet content providers (CPs). It adopts a tenant-based service sharing mode. For a CDN node, the CP device is a streaming medium. It is equivalent to a server. In general, a CDN node collects CPs (Content Servers) from different Internet content providers. These CP devices are configured with different hardware configurations. Therefore, How to accurately distribute the CDN node's load based on the service capabilities of the CP devices. The issue of how it will be adjusted needs to be resolved. 25 SUMMARY The aim of the current arrangement is to determine the CDN node's service capacity based on a CP device's service capability. A CDN node that can accurately adjust its load is a load for an electronic device and a medium. The configuration method involves providing the necessary equipment. 2 To achieve the above objectives, the amendments to the current regulation are as follows: It is applied. First, a payload configuration method is provided for a CDN node, which includes the following: The traffic of a target CP device where the target CP device serves as the server of the CDN node. Obtaining the feature; based on the traffic feature of the target CP device, the IO 5 of the target CP device. The target CP device takes as a constraint that its consumption does not exceed an upper limit of the IO throughput. Determining the maximum service traffic that can be achieved by; by the target CP target based on maximum accessible service traffic and the hardware load status of the target CP device. Determining the reference service traffic of the CP device; and the reference service traffic of the target CP device. According to the plan, the CDN node's load is allocated to the target CP device. 10 From a second perspective, a load configuration apparatus is provided for a CDN node, which includes the following: The traffic of a target CP device where the target CP device serves as a server for the CDN node. a feature acquisition module configured to acquire the feature; traffic of the target CP device. Based on this feature, the IO consumption of the target CP device is set as a constraint, setting an upper limit on the IO throughput. by taking a maximum service traffic that can be obtained by the target CP device, not exceeding 15 a capability estimation module structured to determine; accessible by the target CP Based on the maximum service traffic and the hardware load status of the target CP device, the target CP device's a reference identification module configured to determine the reference service traffic; and the target CP. Allocating the CDN node's load to the target CP device based on the device's reference service traffic. A load adjustment module configured accordingly. 20 From a third perspective, an electronic device is provided that includes: a processor; and a computer. a memory structured to store an instruction that can be executed by a computer, where the computer The instruction can be executed by the processor using the method described in the first aspect when it is executed. This leads to its implementation. From a fourth perspective, it is configured to store an executable instruction by a computer, 25 A computer-readable storage medium is provided; where the computer can read it. When an executable instruction is executed by a processor, it implements the method described in the first point. BRIEF DESCRIPTION OF THE FIGURES The technical solutions in the current regulation or the regulations of the previous technique are more clearly defined. 30 For the purpose of showing it in this way, it is used in the description of regulations or previous techniques. The required accompanying illustrations will be briefly introduced below. Apparently, in the description below... 3 The accompanying drawings are only some of those recorded in the regulations of the current arrangement. These are regulations, and those who are normally experts in the field can do this without creative effort. They can also obtain other accompanying drawings based on the drawings. Figure 1 is a schematic architectural diagram of an associated CDN service on a CP device. FIGURE 2, a load configuration for a CDN node in the current application's settings 5 It is a schematic flowchart of the method. FIGURE 3 shows a load configuration for a CDN node in the arrangement of the current application. This is a schematic modeling diagram of the traffic characteristics of the method. FIGURE 4 shows a load configuration for a CDN node in the arrangement of the current application. This is a schematic structural diagram of the apparatus. 10 FIGURE 5 shows the schematic structure of an electronic device in the regulations of the present application. It is a diagram. REFERENCE STATEMENTS S202. Target CP 15 where the target CP device serves as the server of a CDN node. obtaining the device's traffic capability S204. Based on the traffic characteristics of the target CP device, the IO consumption and IO throughput of the target CP device. by taking as a restriction that it should not exceed an upper limit, what can be reached by the target CP device. Determining maximum service traffic. S206. Maximum service traffic accessible by the target CP device and the target CP device's 20 Determining the reference service traffic of the target CP device based on hardware load status. S208. Load the CDN node based on the target CP device's reference service traffic. allocated to the device P. Content capacity of the target CP device D. CDN node cache capacity is 25 400. Load configuration device 410. Feature acquisition module 420. Capacity estimation module 4 430. Reference determination module 440. Load adjustment module DETAILED DESCRIPTION OF THE INVENTION To enable those with technical expertise to better understand the technical solutions in the specifications, 5 technical solutions in the regulations of the current regulation, the current regulation This will be explained clearly and completely below, with reference to the drawings in the regulations. Apparently, the announced regulations cover only a portion of the provisions in the specification. It outlines, but does not encompass, all regulations. The regulations in the specification... fundamentally, the other 10 obtained by persons normally skilled in the field without any inventive effort. All regulations must fall within the scope of protection of the specification. As mentioned above, a CDN currently provides service to a large number of downstream Internet CPUs. It adopts a multi-tenant service sharing mode. In general, a CDN node uses different Internet content providers configure CP devices. These CP devices have different hardware. It has various configurations, and adjusting the load on the CDN node is difficult. 15 Therefore, the current regulation accurately estimates the service capabilities of CP devices. a technique capable of adjusting the load on the CDN node according to the service capabilities of the CP devices. It aims to propose a solution. See Figure 1, Figure 1 is a schematic architecture of an associated CDN service on a CP device. This is a diagram. A CPU device basically includes: a CPU, an IO bus, an IO bridge, and a primary 20. Storage, a network interface, a disk, etc. The CP device requests a CDN service via its network interface. After receiving it, the search is for IO data, regardless of whether the data is searched in memory or on disk. It must be completed via this route. Therefore, the service capacity of the CP device is one The bottleneck is thought to lie in IO. In light of this, the current regulation estimates the maximum service traffic that the CP device can handle as 25. To do this, the CP device's IO performance constraint (the IO consumption of a single CP device, the upper limit of IO throughput) must be considered. (does not exceed the limit) uses CDN according to the maximum service traffic that the CP device can reach. It allocates the load to the node. In one sense, a modification of the current arrangement treats the CDN node as an execution subject. It proposes a load configuration method for a CDN node that can use the existing 30. Figure 2 shows the current 30 This is a schematic flowchart of the load configuration method in the arrangement. Method It includes the following steps. In S202, the traffic characteristics of a target CP device are obtained, where the target CP device is a CDN. It serves as the server for the node. In the current regulation, the target CP device is required to create a traffic log using a 5 It is configured with an execution program. The step is to determine the traffic characteristics of the target CP device. This may include analyzing the traffic log of the target CP device. In some cases, within the existing regulation, a client's access may be limited to a specific user. redirect to the nearest CDN node and the CDN node is then connected to a streaming media server. Global load balancing to enable a direct response to a user's request. 10 The technology is used. The CDN node's streaming media server provides the content that the user wants to access. If it doesn't have it, the streaming media server automatically retrieves the relevant content from the target CP device according to the configurations. It receives it and provides the same to the client. Here, the stage where the client requests data from the CDN node is considered a service stage. It is defined; and a source-pull stage is when the CDN node requests data from the target CP device. 15 It is defined as a stage. A model shown in Figure 3 includes the following traffic characteristics. I. Traffic characteristics during the service phase x: Service traffic requested by the user within a time unit T, in GB / s. a: an average demand size corresponding to service traffic x; a = Avg(a ), and x = ∑ a = ua. 20 u: In requests / second units, the number of times the user makes requests within a time interval of T. II. Traffic characteristics during the initial-withdrawal phase: v: In request / s units, the number of user start-pull times within a unit time T. z: Minimum start-down speed requested by the user within a unit time T, in Gb / s. traffic (theoretical initial-departure traffic). 25 b: an average origin-pull slice size corresponding to z; where z = ∑ b = etc. y: Total amount of content requested by the user within a time period T, in GB / s. The traffic characteristics x, a, and u above are served along with the CDN node's cache capacity D. It can reflect a hit rate at the stage and the above v, z and b traffic characteristics target CP 6 The device's content capacity, along with P, provides an accuracy rate during the source-drawing phase. It should not be forgotten that this can reflect on... Additionally, a traffic feature reflecting popularity has been added. p: Content popularity in % corresponding to x service traffic; where p = 1 − . y: Total amount of content requested by the user within a time period T, in GB / s. 5 In addition, based on the above, the following parameters have also been introduced: z': Origin-pull traffic actually generated by a CDN within a unit time T, in GB / m²; where, z = λz, λ ∈ [1, +∞], where λ=1 represents equal origin-pulling and λ>1 represents excessive origin-pulling. Q: a request hit rate in percentages corresponding to x service traffic; where Q = 1 − . q: Theoretical bandwidth hit rate in percentage terms corresponding to x service traffic; where q = 10 1 − . q': the actual bandwidth hit rate corresponding to x service traffic in % units; where q′ = 1 − . λ: an origin-pull amplification factor in percentage units; where λ = ' . The traffic characteristics of the target CP device within the unit time T obtained in this step are one of the following two sets of five: It may include one. 15 One quintet is [u, v, x, y, z]; the other quintet is [a, b, x, p, q]. In S204, based on the traffic characteristics of the target CP device, the IO consumption and IO throughput of the target CP device can be determined. The target CP device can reach a limit that is not exceeded, with this being taken as a restriction. Maximum service traffic is determined. Referring to the traffic characteristics model of the target CP device described above, the following 20 Assumptions can be made: x: Service traffic requested per unit of time. y: total amount of content requested per unit of time (content popularity: p = 1 − ). z: origin-pull traffic generated per unit time (content hit rate: q = 1 − ). λ: an origin-pull amplification factor λ ∈ [1, +∞], =1 equal origin-pull and >1 excessive origin-pull 25 It represents. 7 D: Disk number of the target CP device. R: The upper limit of the IO throughput capacity of each disk in the target CP device (unit: bps). θ: the ratio of the IO throughput consumption of a write operation to the IO throughput consumption of a read operation, write The size of the operation is the same as that of the write operation. Based on the five-digit [u, v, x, y, z], the maximum service traffic accessible by the target CP device is X = 5. The calculation formula is as follows: X = zy y + θλz − RD Based on the five-digit [a, b, x, p, q], the maximum service traffic X that can be accessed by the target CP device is... The calculation formula is as follows: X = RD θλ − pq − q Here, x = calculation formula is taken as an example. The target CP device has 24 disks. Assuming it is configured with an average service slice size of α = 512 KB and each disk having a throughput of 300 KB. When the speed is Mbps: R = 0.3; D = 24; IO = 0.3 × 24 = 7.2. Accordingly, the maximum service traffic x is derived as follows: IO + θIO = RD; 15 Assuming that disk write operations consume twice as many I / O resources as read operations. Let θ = 2; IO + 2IO = 7.2; Because IO = yq; IO = θz′ = θλz; can be obtained as: yq + θλz = RD; where, disk read Its I / O ignores the fact that the content itself is being loaded onto a RAM or SSD disk, instead viewing the content as a single file. It adopts a hit rate of x for the total content planned and cached in memory. 20 Furthermore, by substituting these into the formulas p = 1 − and q = 1 − x², the following are obtained. [u, v, x, y, z], X = . For the target CP device corresponding to the quintuple. [a, b, x, p, q],X = . For the target CP device corresponding to the quintuple. In S206, the maximum service traffic accessible by the target CP device and the target CP device's The reference service traffic of the target CP device is determined based on the hardware load status. 25 8 To more accurately estimate the service capacity of the target CP device, the existing The regulation further unifies the hardware load status of the target CP device. It should be understood that this is necessary. A hardware load rate of the target CP device may include: a CPU load rate and here an unlimited network card load rate. In some cases, the hardware idle rate of the target CP device in the actual editing scenario is 5. If required, this step involves the reference service traffic of the target CP device using the formula x = X. This may include calculations based on; where x represents the reference service traffic of the target CP device. X represents the maximum service traffic accessible by the target CP device; w represents the target CP device's hardware load rate; and α, ensuring that the IO consumption of the target CP device does not exceed the upper limit of its IO throughput. It represents the hardware idle rate of the target CP device under the constraint. 10 Alternatively, if there is no requirement for the hardware idle rate of the target CP device, this step can be omitted. α can be omitted and the reference service traffic of the target CP device can be directly calculated using the formula x = wX. It is calculated. In S208, a payload from the CDN node is sent to the target CP based on the target CP device's reference service traffic. It is allocated to your device. 15 The final determined reference service traffic of the target CP device, the T unit of the target CP device It should be understood that this can be considered as theoretical actual service capacity within a given time period. Based on the above content, in the method of the regulations of the current regulation, the CP device... The maximum service traffic that can reflect the service capacity, the CP device's IO consumption, and its IO throughput. The CP device's traffic characteristics are used as a restriction, and the estimate is based on the fact that it does not exceed the upper limit, and 20 The reference service traffic of the CP device, the reference service traffic of the CP device's actual service. The actual hardware load status of the CP device as it approaches the upper limit of its capacity. This is determined by combining the data. Therefore, the CDN node's traffic is determined based on the reference service traffic of the CP device. After the payload is allocated to the CP device, the allocation performance of the CP device is fully functional. While available, the stability of a CDN service is ensured. 25 The above is an introduction to load configuration for a single target CP device. Typically, CDN The node is configured with numerous CP devices. For the current arrangement, this corresponds to the CDN node. The reference service traffic for each incoming CP device can be determined in the same way. Then, the CDN Node load allocation is done by referring to the reference service traffic of each CP device. This is done. For example, the actual service traffic of each CP device is 30 times the reference service traffic. 9 Assuming it does not exceed a certain load, each CDN node distributes a specific load according to a load balancing policy. It is assigned to the CP device. Below is an example where a CDN node is configured with m CP devices in n consecutive time units. The load configuration of the CDN node is introduced by taking this information. It should be understood that the traffic characteristics of CP devices are variable. From a CDN perspective, CP 5 the traffic characteristics of the devices are a time series (a set of random variables ordered by time) This can be considered as the case where the CDN node is configured with mCP devices for hybrid operation. The mixed traffic characteristic of CP devices can also be represented as a random sequence using the following matrix. understandable: 𝑢, 𝑣, 𝑥, 𝑦, 𝑧 ⋯ 𝑢, 𝑣, 𝑥, 𝑦, 𝑧 ⋮ ⋱ ⋮ 𝑢, 𝑣, 𝑥, 𝑦, 𝑧 ⋯ 𝑢, 𝑣, 𝑥, 𝑦, 𝑧 The following parameters are defined here. N: the number of CP devices in a group. D: Number of disks for each CP device. R: The upper limit of the IO throughput of each disk (unit: bps). θ: the ratio of the IO throughput consumed by a write operation to that consumed by a read operation, 15 The size of the write operation is the same as that of the write operation. i, j, k: balancing factors of lvs / slb, slb / cache, cache / disk respectively, in percentage units, with values ​​of i, j, k ≥ 1. λ: CP origin-pull amplification factor in percentage units, within the range of λ ≥ 1. x: Service traffic requested by CP users. y: Total amount of content requested by CP users. 20 z: A resource-pull traffic requested by CP users. p: Popularity of content requested by CP users; p = 1 − . 𝑞 : Bandwidth hit rate requested by CP users; q = 1 − . Assuming there are a total of M CP devices for hybrid processing in an existing group of CDN nodes, The group's total service bandwidth x achievable: 25 x = ∑ x . The service traffic of a single CP device is calculated as 𝑥′: 𝑥′ = 𝑖𝑗. The total I / O (input / output) of the disks of a single CP device is calculated: IO = IO + θIO = ∑ yq + θ ∑ λ z ; = ∑ x′ (1 − p )q + θλ x′ (1 − q ); 5 = ∑ x′ [(1 − p )q + θλ (1 − q )]; = ∑ x [(1 − p )q + θλ (1 − q )]. Next, the I / O consumption of a single disk is calculated: IO′ = k; IO′ = ∑ x [(1 − p )q + θλ (1 − q )] ≤ R; 10 ∑ x [(1 − p )q + θλ (1 − q )] ≤ . For the group corresponding to the CDN node, a group service traffic pattern is obtained when there is no difference in quality. can be done: x = ∑ x ∑ x [(1 − p )q + θλ (1 − q )] ≤ NDR ijk Using the group service traffic model, m CP 15 on the CDN node within 1 to n unit times. The reference service traffic corresponding to the device can be dynamically determined and the CDN node The load is allocated to the relevant CP devices based on reference service traffic. From another perspective, an adjustment to the current arrangement also includes a payload for a CDN node. The configuration apparatus (400) is provided. Fig. 4 shows the load configuration apparatus (400) which includes the following: This is a schematic structural diagram: the target CP device serves as a server for the CDN node 20. obtaining a feature configured to obtain the traffic characteristics of a given target CP device module (410); based on the traffic characteristics of the target CP device, the IO consumption of the target CP device IO by taking as a constraint that it does not exceed an upper limit of its efficiency, the target CP device obtains a capacity estimate structured to determine a maximum service traffic that can be achieved module (420); maximum service traffic accessible by the target CP and the target CP device's 25 To determine the reference service traffic of the target CP device based on the hardware load status. 11 a structured reference determination module (430); and the reference service traffic of the target CP device a payload configured to allocate the CDN node's load to the target CP device. Adjustment module (440). The current regulatory framework includes a mechanism that can reflect the service capacity of the CP device. Maximum service traffic, constraint 5: The CP device's IO consumption must not exceed an upper limit of its IO throughput. This is taken into account and estimated according to the traffic characteristics of the CP device, and the reference service of the CP device. traffic, reference service traffic towards the upper limit of the CP device's actual service capacity The actual hardware load state of the CP device is determined by combining the approaches taken in this way. Therefore, CP After allocating the CDN node's load to the CP device based on the device's reference service traffic, While ensuring the stability of a CDN service, the allocation performance of the CP device is exactly 10%. available. Optionally, the traffic feature of the target CP device includes: within a unit time T. The number of times a user requests something, u, is the number of requests made by the user within a unit of time, T. Service traffic is x, and the total amount of content requested by the user within a unit time period T is y, T. Minimum source-pull traffic requested by the user within a unit time period is z and T units 15 Number of user resource-withdrawal times within the period v. Capability estimation module (420) by restricting the target CP device's IO consumption from exceeding the upper limit of its IO throughput. by obtaining the traffic characteristics of the target CP device, it can be accessed by the target CP device. Determining the maximum service traffic: the maximum amount of traffic that can be accessed by the target CP device. Service traffic is determined based on a formula x = , where θ is the number of write operations. It represents the ratio between the IO throughput consumption of a read operation and the IO throughput consumption of a write operation. The size of the operation is the same as the size of the write operation; D represents the number of disks on the target CP device. R represents the upper limit of IO throughput for each disk in the target CP device; and λ is the origin-pull. λ represents the amplification factor, λ ∈ [1, +∞], λ=1 represents equal origin-pull and λ>1 represents excess It represents origin-pulling. 25 Optionally, the traffic capability of the target CP device includes: a unit T by a user. Service traffic requested within a given period is x; the average service slice corresponding to service traffic x. Size a; service traffic x is the average source-withdrawal slice size b; service traffic x is the bandwidth hit rate q; and service traffic is the content popularity x. p . The capability estimation module (420) estimates the IO consumption of the target CP device to be within the top 30% of its IO throughput. Taking the restriction of not exceeding the limit as a basis, the target CP device's traffic characteristics determine the target CP. Determining the maximum service traffic that can be handled by the device: x = formula 12 Based on this, determining the maximum service traffic that can be accessed by the target CP device; Here, θ is the difference between the IO throughput consumption of a write operation and the IO throughput consumption of a read operation. It represents the ratio, and the size of the write operation is the same as the size of the write operation; D is the target CP. `R` represents the number of disks in the device; `R` is the upper limit of IO throughput for each disk in the target CP device. λ represents the origin-pull amplification factor, λ ∈ [1, +∞] , λ=1 equals origin-5 λ represents pull, and λ>1 represents extreme origin-pull. Optionally, by the reference determination module (430), by the target CP device target based on maximum accessible service traffic and the hardware load status of the target CP device. Determining the reference service traffic of the CP device involves using the formula x = X. The reference service traffic of the target CP device is determined as follows: where x is the reference service traffic of the target CP device. Represents the reference service traffic; the maximum service accessible by the target CP device X. w represents the traffic; w represents the hardware load rate of the target CP device; and α represents the target CP device's Under the constraint that IO consumption must not exceed the upper limit of IO throughput, the hardware of the target CP device... It represents the vacancy rate. Optionally, the hardware load rate of the target CP device can be a CPU load rate and a network card load rate of 15. It includes the ratio. Optionally, the target CP device can be configured with an executable program to generate a traffic log. is configured; and the traffic feature of the target CP device is determined by the feature acquisition module (410). This includes: analyzing the traffic log of the target CP device and the target CP device's Determining the traffic characteristics. 20 Optionally, the CDN node can handle multiple CP devices, including the target CP device. It is configured and each CP device serves as a server for the CDN node. Load scheduling Module 540 adjusts the CDN node's load according to the target CP device's reference service traffic. Allocating the CDN node's payload to the target CP device involves: The target CP device is 25 based on the reference service traffic and the reference service traffic of other CP devices. including the allocation of numerous CP devices. The model load configuration apparatus (400) shown in FIGURE 4 is one of the methods shown in FIGURE 2. It can serve as the subject of execution and thus the same principle as the method shown in FIGURE 2. Therefore, it can implement its steps and related functions here, which are not particularly repetitive. Figure 5 is a schematic structural diagram of an electronic device in one arrangement of the specification. 30 Referring to Figure 5, at the hardware level, the electronic device includes a processor and optionally 13 It includes an internal data bus, a network interface, and memory. The memory is high-speed random access memory. It can include internal memory such as RAM or at least some non-volatile memory such as disk storage. Of course. Electronic devices may also include hardware necessary for other services. The processor, network interface, and memory can be connected to each other via an internal data bus, and the internal data bus can connect to others. as well as an industry standard architecture (ISA) data bus, a peripheral component interconnect (PCI) data 5 The data bus can be an address, either a bus or an Extended Industry Standard Architecture (EISA) bus. A data bus can be divided into data buses, control buses, and so on. Ease of representation. In this respect, Figure 5 shows only one bidirectional arrow, but this only represents a data path or It does not indicate a data bus type. Memory is structured to store a program. In some cases, the program contains a code of 10... It may include program code that contains computer operating instructions. Memory includes internal memory and temporary memory. It may contain non-existent memory and provides instructions and data to the processor. Optionally, the processor reads a relevant computer program from non-volatile memory to internal memory. and then by running the computer program for the CDN node shown in Figure 4 above It creates the load configuration apparatus (400) at the logical level. Accordingly, the processor has 15 in memory. It executes the stored program and is specifically configured to perform the following operations. The traffic characteristics of a target CP device are obtained, where the target CP device is a CDN node. It serves as a server. Based on the traffic characteristics of the target CP device, the IO consumption of the target CP device should be a maximum of its IO throughput. The maximum limit that can be reached by the target CP device is 20, with the restriction that it should not exceed this limit. Service traffic is determined. The reference service traffic of the target CP device is the maximum service traffic that can be accessed by the target CP device. This is determined based on traffic and the hardware load status of the target CP device. The CDN node's load is allocated to the target CP device based on the target CP device's reference service traffic. is done. 25 The current arrangement allows the electronic device to reflect the service capacity of the CP device. The maximum service traffic is constrained by the condition that the CP device's IO consumption does not exceed the upper limit of its IO throughput. The traffic of the CP device is estimated based on the data obtained, and the reference service traffic of the CP device is calculated as follows: The actual hardware load status of the device is further determined by combining the reference service traffic. The CP device accurately approaches the upper limit of its actual service capacity. Therefore, CP 30 After allocating the CDN node's load to the CP device based on the device's reference service traffic, 14 While ensuring the stability of a CDN service, the allocation performance of the CP device is fully realized. available. The load configuration method described above in the regulations shown in the specification. It can be implemented by or applied to the processor. The processor has signal processing capabilities. It may be an integrated circuit chip. In a modification process, all steps of the above method are 5 with an instruction in the form of an integrated logic circuit or software in the processor of the hardware It can be completed. The processor mentioned above is a central processing unit (CPU), a network processor (NP), and so on. This could be a general-purpose processor, including a digital signal processor (DSP). application-specific integrated circuit (ASIC), field-programmable gate array (FPGA) or other Programmable logic devices, discrete gate or transistor logic devices, and discrete hardware 10 It may have components. Various methods, steps, and regulations are described in the provisions of the current regulation. Logic block diagrams can be implemented or realized. A general-purpose processor is a microprocessor. It could be, or this processor could be any conventional processor, etc. The current arrangement The steps of the method described in connection with the regulations, a hardware decoding processor can be directly embodied or decoded to be executed and completed by 15 It can be executed through a combination of hardware and software modules in its processor, and can be completed. A software module can be placed in random access memory, a flash memory, or a read-only memory. in memory, in programmable read-only memory, or in electrically erasable programmable memory It can be found in memory, in a record, and in other mature storage media in the technology. Storage media It resides in memory. The processor reads the information from memory and, along with the hardware, uses the method described above (20). He completes his steps. In addition to software applications, the specification may include electronic devices, logic devices, or software and It does not exclude other applications such as hardware combinations, i.e., one of the following process flows The executive branch is not limited to each logic unit, but can be hardware or logic devices. In addition, an amendment to the current regulation also designates one or more programs as 25 It proposes a storage medium that is readable by the computer storing it, and one or more The program includes instructions. Optionally, the above instructions can be used with a portable system that includes numerous editing programs. when operated by an electronic device, the portable electronic device is shown in Figure 2. This may cause the method to perform its steps. The steps of the method shown in 2 are 30. This can cause the target CP device to act as the server of the CDN node. Obtaining the traffic characteristics of a given target CP device; the traffic characteristics of the target CP device Based on this, the IO consumption of the target CP device is set as a constraint, setting an upper limit on the IO throughput. by taking the maximum service traffic that can be accessed by the target CP device, without exceeding it. Determining; the maximum service traffic that can be accessed by the target CP device and the target CP device's Determining the reference service traffic of the target CP device based on the hardware load status; and the target The CDN node's load is allocated to the target CP device based on the CP device's reference service traffic. to be done. Experts in this field say that the regulations in the specification concern methods, systems, or computers. It should be understood that the program can be provided as products. Therefore, the specification must include full hardware. regulations, full software regulations, or combining software and hardware aspects. It can adopt various forms of arrangement. Additionally, the specification may include existing program codes on the computer. storage media available on one or more computers (a disk drive, a CD-ROM, a (including but not limited to optical memory and similar devices) computer applications The program may adopt a form of its products. The specific regulation of the specification is described above. Other regulations are in the attached claims. is within the scope. In some cases, the actions or steps recorded in the requests are 15 years from the regulation. It can be executed in a different order and still achieve the expected results. Also, see the attached diagrams. The process shown is a specific sequence or order of steps taken to achieve the expected results. It is not required. In some configurations, multitasking and parallel processing may be possible or advantageous. it could be. The above are merely provisions of the specification, and the purpose of limiting the specification is 20 It does not include any technical expertise. For those specializing in the field, the specifications have various changes and variations. It is possible. Any change made within the spirit and principle of the specification, equivalent replacement, Improvements and similar measures will all be included within the requirements of the specification. In addition, other things that are normally obtained in technology by experts without creative effort The regulations should fall within the scope of protection of this document. 25

Claims

16 REQUESTS 1. A payload configuration method for a content delivery network (CDN) node that includes the following: Its characteristic is:  A destination content provider (CP) that serves as a server for the CDN node Obtaining the traffic feature of the device, 5  Based on the traffic characteristics of the target CP device, the target CP device has an input / output (IO) The goal is to set a constraint by ensuring that consumption does not exceed an upper limit of the IO throughput. Determining the maximum service traffic that can be accessed by the CP device;  Maximum service traffic accessible by the target CP device and the target CP device's Based on the hardware load status, the reference service traffic of the target CP device is 10. determination and  The load of the CDN node relative to the reference service traffic of the target CP device. Allocation to CP device It includes the steps of the process.

2. This method complies with Claim 1 and its characteristic is;  The target CP device's traffic characteristic should include: within a unit of time T, one The number of times a user requests something is u, within a unit of time T, a user The service traffic requested by x is the number of users within a unit of time T. The total amount of content requested by y is 20 within a unit of time T. Minimum origin-pull traffic requested by the user is z and a unit of time is T. v is the number of origin-withdrawal times of a user within; and  Based on the traffic characteristics of the target CP device, the IO consumption of the target CP device is determined. by taking as a constraint that it does not exceed the upper limit of its efficiency, by the target CP device Determining the maximum service traffic that can be accessed: 25  The I / O throughput of a write operation versus the I / O throughput of a read operation θ represents the ratio between them; D represents the number of disks on the target CP device; R represents an upper limit of IO throughput for each disk in the target CP device; and a λ represents the origin-pull amplification factor, λ ∈ [1, +∞] , λ =1 equal to origin- x = 30 represents the pull and λ>1 represents the extreme origin-pull. 17 Based on the formula, the maximum service that can be accessed by the target CP device. determining traffic It includes the steps of the process.

3. The method is compliant with Claim 1 and its characteristic is; 5  The target CP device's traffic feature must include: a unit T by a user. Service traffic requested within a given period is x; average corresponding to service traffic x. a service slice size 'a'; service traffic x is an average initial value corresponding to a specific starting value. pull slice size b; bandwidth hit rate q corresponding to service traffic x; and service traffic corresponds to content popularity p; and 10  Based on the traffic characteristics of the target CP device, the IO consumption of the target CP device is determined. by taking as a constraint that it does not exceed the upper limit of its efficiency, by the target CP device Determining the maximum service traffic that can be accessed:  θ is the IO throughput consumption of a write operation versus the IO throughput consumption of a read operation. the ratio between them; D the number of disks in the target CP device; R every 15 in the target CP device It represents an upper limit of IO throughput for the disk and an origin-pull amplification factor. λ, λ ∈ [1, +∞], where λ=1 represents equal origin-pulling and λ>1 represents excessive origin-pulling. Based on the formula x =, the target CP device will be able to reach it. Determining maximum service traffic. It includes the steps of the process. 20 4. This method complies with Claim 1 and its characteristic is;  Maximum service traffic accessible by the target CP device and the target CP device's Reference service traffic of the target CP device based on hardware load status. determination: 25  Reference service traffic of target CP device x; accessible by target device X maximum service traffic; w is the hardware load rate of the target CP device; and α is the target. Under the constraint that the CP device's IO consumption must not exceed the upper limit of its IO throughput. to represent the hardware idle rate of the target CP device Determining the reference service traffic based on a formula x = X 30 18 It includes the steps of the process.

5. This method complies with Claim 4, and its characteristic is that the hardware load rate of the target CPU device is a CPU. It consists of the load factor and the load factor of a network card.

6. The method must comply with any of claims 1 to 5, and its characteristic feature is:  with an execution program to create a traffic log of the target CP device structuring; and  Obtaining the traffic characteristics of the target CP device; traffic of the target CP device Analysis of the log and determination of the traffic characteristics of the target CP device 10 It includes the steps of the process.

7. The method must comply with any of claims 1 to 5, and its characteristic feature is:  Configuring the CDN node with multiple CP devices, including the target CP device. and each CP device serves as a server for the CDN node; and 15  The CDN node's load should be adjusted according to the target CP device's reference service traffic. Allocation to a CP device includes the following:  The CDN node's load should be the reference service traffic of the target CP device and other CPs. numerous devices that include the target CP device based on the reference service traffic of those devices. Allocation to CP device 20 It includes the steps of the process.

8. A load configuration apparatus (400) for a CDN node, which includes the following features:  Traffic of a target CP device serving as a server for the CDN node a feature acquisition module configured to obtain the feature (410); 25  Based on the traffic characteristics of the target CP device, the IO consumption of the target CP device can be determined. as a constraint, the target CP device should not exceed an upper limit of the IO throughput. configured to determine the maximum service traffic that can be accessed by Capacity estimation module (420); 19  Maximum service traffic accessible by the target CP and the target CP device Determine the reference service traffic of the target CP device based on the hardware load status. a reference determination module structured for (430); and  Load the CDN node's load onto the target CP device based on the target CP's reference service traffic. a load adjustment module (440) 5 configured to allocate to the device It includes.

9. An electronic device that includes the following, and whose characteristic is that it is powered by a processor and a computer. It is a memory structured to store an executable instruction; where the computer The instruction can be executed by the processor; when executed, it will trigger any of the 1st to 7th prompts. According to one person, it is the reason why they apply the method. 10 10. A computer configured to store an executable instruction. It is a readable storage medium, characterized by its ability to be executed by a computer. When the instruction is executed by a processor, it corresponds to any of prompts 1 through 7. It is the application of the method.

Images