CDN service method, system, device and equipment

Abstract

The invention discloses a CDN (Content Delivery Network) service method, system, device and equipment, and the method comprises the steps: receiving a content access request containing a service chain at an edge node, sequentially executing each intelligent processing task according to the service chain, and when each intelligent processing task is executed, carrying out the corresponding intelligent processing task; sending the context information and the current intelligent processing task to an execution node corresponding to the current intelligent processing task, and receiving a task execution result returned by the execution node and context information used when a next intelligent processing task is executed; and finally, generating response content based on the received task execution results and returning the response content to the user side. Through synchronization and updating of the context information among the nodes, cross-node context collaboration is realized, and the problem that an existing CDN cannot support a complex business process needing multi-node or multi-model collaboration is solved.

Description

Technical Field

[0001] This application belongs to the field of computer technology, and in particular relates to a CDN service method, system, apparatus and equipment. Background Technology

[0002] A Content Delivery Network (CDN) is a network technology that accelerates content access by deploying distributed edge node servers around the world to cache static content (such as web pages, images, and video files) closer to users.

[0003] However, with the increasing popularity of business models with high real-time attributes such as personalized recommendations, real-time interaction, and AIGC (AIGC), more and more business scenarios require CDNs not only to distribute static content, but also to perform real-time intelligent processing during the distribution process. For example, CDNs can dynamically insert personalized advertisements based on user characteristics, generate multilingual subtitles for live streams in real time, or perform content security filtering based on context.

[0004] To adapt to this trend, existing solutions deploy lightweight AI models on CDN nodes, treating the models as isolated point services deployed on edge nodes. Their processing flow is limited to one-time input and output. Moreover, since nodes in the traditional CDN architecture are stateless, meaning that nodes process requests independently and do not share user state or pass intermediate information generated during processing, they cannot support complex business processes that require collaboration among multiple nodes or models. Summary of the Invention

[0005] To address the aforementioned issues, this application provides a CDN service method, system, apparatus, and device that achieves cross-node contextual collaboration through dynamic updating and synchronization of context information between CDN nodes, supporting complex business processes requiring multi-node or multi-model collaboration.

[0006] In a first aspect, this application provides a CDN service method applied to edge nodes, the method comprising: The system receives a content access request sent by a user client. The content access request includes a content identifier and a service chain. The content identifier is used to indicate the target content of the request. The service chain is used to indicate one or more intelligent processing tasks that need to be executed sequentially on the target content. Determine the context information corresponding to the content access request, the context information including information reflecting user profile characteristics and / or information reflecting the user's access environment; Based on the content identifier, the context information, and the service chain, a cache query key is constructed, and the response content corresponding to the cache query key is queried in a preset cache library; if the response content corresponding to the cache query key is not found, performing each intelligent processing task in the service chain; wherein, when each intelligent processing task is performed, the context information and the current intelligent processing task are sent to the execution node corresponding to the current intelligent processing task, and the task execution result returned by the execution node and the context information used when the next intelligent processing task is performed are received; generating the response content based on the received task execution results and returning the response content to the user terminal.

[0007] In a possible implementation, before the performing each intelligent processing task in the service chain, the method further includes: when it is detected that the content access request includes a service chain enhancement authorization identifier, matching the context information with a predefined enhancement rule set; wherein, the enhancement rule set includes at least one enhancement rule, and each enhancement rule includes a rule trigger condition, an intelligent processing task to be inserted, and a relative position relationship between the intelligent processing task to be inserted and part or all original intelligent processing tasks in the service chain; determining a target enhancement rule matched successfully with the context information according to the rule trigger condition met by the context information; adding the intelligent processing task included in the target enhancement rule to the service chain according to the relative position relationship included in the target enhancement rule.

[0008] In a possible implementation, the method further includes: receiving a node link tracking record generated by the execution node; the node link tracking record includes at least one of the following information: a node identifier, model identifier information of a target model called by the execution node when the current intelligent processing task is executed, and input data digest, output data digest, and task processing time consumption of each target model; generating a link tracking sequence corresponding to the content access request based on the node link tracking records generated by the execution nodes corresponding to each intelligent processing task in the service chain; determining the context information received from the execution node corresponding to the last intelligent processing task in the service chain as the final version of the context information; generating a full link tracking log corresponding to the content access request based on the content access request, the final version of the context information, the service chain, and the link tracking sequence.

[0009] In a second aspect, the application provides a CDN service method applied to an execution node, and the method includes: receiving context information and an intelligent processing task to be executed from an edge node; performing the intelligent processing task based on the context information to obtain a task execution result; extracting semantic information from the task execution result according to a predefined extraction specification, and adding the semantic information to the context information to obtain context information used by the edge node when performing a next intelligent processing task; wherein the extraction specification defines types of semantic information that need to be extracted from the task execution result, and an extraction manner of each type of semantic information; returning the task execution result and the context information used when performing the next intelligent processing task to the edge node.

[0010] In a possible implementation, the method further includes: generating a node link tracking record, and returning the node link tracking record to the edge node; The node link tracking record includes at least one of the following information: a node identifier, model identifier information of a target model called by the execution node when performing the intelligent processing task, and input data digest, output data digest, and task processing time consumption of each target model.

[0011] In a third aspect, the present application provides a CDN service method, applied to a user end, the method includes: determining a content identifier corresponding to a target content requested by a user; generating a service chain according to a function option selected by the user, the service chain being used to indicate one or more intelligent processing tasks that need to be performed on the target content in sequence; generating a content access request based on the content identifier and the service chain, and sending the content access request to an edge node, so that the edge node performs each intelligent processing task according to the service chain when no response content corresponding to a cache query key is queried; wherein the cache query key is composed of the content identifier, context information corresponding to the content access request, and the service chain; the context information includes information reflecting a user portrait feature and / or information reflecting a user access environment; when performing each intelligent processing task, the edge node sends the context information and the current intelligent processing task to an execution node corresponding to the current intelligent processing task, and receives a task execution result returned by the execution node and context information used when performing a next intelligent processing task; receiving response content returned by the edge node, the response content being generated by the edge node based on the received task execution result.

[0012] In a possible implementation, the method further includes: determining whether to add a service chain enhancement authorization identifier to the content access request according to the service chain enhancement configuration information, the service chain enhancement authorization identifier being used to indicate whether to allow the edge node to add intelligent processing tasks in the service chain, and when the edge node detects that the content access request includes the service chain enhancement authorization identifier, matching the context information with a predefined enhancement rule set, and determining whether to add intelligent processing tasks in the service chain based on a matching result; The enhancement rule set includes at least one enhancement rule, and each enhancement rule includes a rule trigger condition, an intelligent processing task to be inserted, and a relative position relationship between the intelligent processing task to be inserted and part or all original intelligent processing tasks in the service chain.

[0013] In a fourth aspect, the present application provides a CDN service device, which includes: a request receiving module configured to receive a content access request sent by a user end, the content access request including a content identifier and a service chain, the content identifier being used to indicate a target content requested, and the service chain being used to indicate one or more intelligent processing tasks that need to be sequentially performed on the target content; a context information determining module configured to determine context information corresponding to the content access request, the context information including information reflecting user portrait features and / or information reflecting a user access environment; a cache query module configured to construct a cache query key based on the content identifier, the context information, and the service chain, and query a response content corresponding to the cache query key in a preset cache library; a task execution module configured to perform each intelligent processing task according to the service chain if the response content corresponding to the cache query key is not queried, wherein when each intelligent processing task is performed, the context information and a current intelligent processing task are sent to an execution node corresponding to the current intelligent processing task, a task execution result returned by the execution node is received, and context information used when a next intelligent processing task is performed is received; a response module configured to generate a response content based on each received task execution result and return the response content to the user end.

[0014] In a fifth aspect, the present application provides a CDN service device, which includes: a task receiving module configured to receive context information and an intelligent processing task to be executed from an edge node; a task processing module configured to execute the intelligent processing task based on the context information to obtain a task execution result; a context updating module, configured to extract semantic information from the task execution result according to a predefined extraction specification, and add the semantic information to the context information, to obtain context information used by the edge node when performing a next intelligent processing task; a task response module, configured to return the task execution result and the context information used when performing a next intelligent processing task to the edge node.

[0015] In a sixth aspect, the present application provides a CDN service device, the device comprising: an information determining module, configured to determine a content identifier corresponding to target content requested by a user; a service chain generating module, configured to generate a service chain according to a function option selected by the user, the service chain being used to indicate one or more intelligent processing tasks that need to be performed on the target content in sequence; a request sending module, configured to generate a content access request based on the content identifier and the service chain, and send the content access request to an edge node, so that the edge node performs each of the intelligent processing tasks according to the service chain when no response content corresponding to a cache query key is queried; wherein the cache query key is composed of the content identifier, context information corresponding to the content access request and the service chain, the context information comprising information reflecting a user portrait feature and / or information reflecting a user access environment; when performing each of the intelligent processing tasks, the edge node sends the context information and a current intelligent processing task to an execution node corresponding to the current intelligent processing task, and receives a task execution result returned by the execution node and context information used when performing a next intelligent processing task; a response receiving module, configured to receive response content returned by the edge node, the response content being generated by the edge node based on the received task execution result.

[0016] In a seventh aspect, the present application provides a CDN service system, the system comprising: an edge node, configured to perform the CDN service method according to any one of the first aspect of the present application; an execution node, configured to perform the CDN service method according to any one of the second aspect of the present application; a user terminal, configured to perform the CDN service method according to any one of the third aspect of the present application.

[0017] In an eighth aspect, an apparatus is provided, including at least one processor; and a memory communicatively connected with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the CDN service method according to any one of the first aspect of the present application, or perform the CDN service method according to any one of the second aspect of the present application, or perform the CDN service method according to any one of the third aspect of the present application.

[0018] In a ninth aspect, a computer-readable storage medium is provided, which stores a computer program, and when the computer program is executed by a processor, the terminal device is enabled to perform the CDN service method according to any one of the first aspect of the present application, or perform the CDN service method according to any one of the second aspect of the present application, or perform the CDN service method according to any one of the third aspect of the present application.

[0019] The technical solutions provided by the embodiments of the present application at least have the following beneficial effects: The CDN service method, system, device and apparatus provided by the present application receive a content access request containing a service chain at an edge node, and sequentially schedule each intelligent processing task according to an ordered task sequence defined by the service chain, and send current version context information to a corresponding execution node during task execution, and the execution node updates the context information based on the task execution result to obtain new version context information and return to the edge node, and the edge node uses the new version context information for the next intelligent processing task, and through the synchronization and update of the context information between the execution nodes, cross-node context collaboration is realized, so that the subsequent tasks can be processed coherently based on the semantic results of the previous tasks, and the problem that the traditional CDN cannot support complex business processes requiring multi-node or multi-model collaboration is solved.

[0020] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art from the description, or can be learned by practice of the present application. The objects and other advantages of the present application will be realized and achieved by means of the structures particularly pointed out in the written description and claims hereof as well as the appended drawings. BRIEF DESCRIPTION OF DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiments of the present application will be briefly introduced as follows. Obviously, the drawings introduced below are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0022] Figure 1 An application scenario schematic diagram provided for an embodiment of the present application; Figure 2 A CDN service method flowchart applied to an edge node provided for an embodiment of the present application; Figure 3 A CDN service device schematic diagram provided for an embodiment of the present application; Figure 4 Another CDN service device schematic diagram provided for an embodiment of the present application; Figure 5 Another CDN service device schematic diagram provided for an embodiment of the present application; Figure 6 A CDN service device schematic diagram provided for an embodiment of the present application. DETAILED DESCRIPTION

[0023] To make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described below in detail with reference to the accompanying drawings. The described embodiments are some embodiments of the present application, but not all embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0024] Some terms in the embodiments of the present application are explained below to facilitate understanding by those of ordinary skill in the art.

[0025] (1) Content identifier, a string or code used to uniquely identify target content determined by a user terminal based on a user request for target content, for example, a video stream ID (such as "live_123"), a file URL hash value and a content resource URI, which can be used as a keyword for content retrieval to find corresponding cached content in a content delivery network (CDN) cache system.

[0026] (2) Edge node, the edge node in the present application is a node close to the user side in a CDN, which is used to receive content access requests from a user terminal. Unlike the edge node of a traditional CDN which only provides content caching and forwarding functions, the edge node in the present application also has the functions of performing context collection and management, service chain analysis, task coordination and processing of content returned by multiple execution nodes, and can coordinate multiple execution nodes to jointly complete intelligent processing tasks.

[0027] (3) an execution node, which is a computing entity capable of executing intelligent processing tasks. Specifically, each execution node in the embodiments of the present application can receive context information and intelligent processing tasks from an edge node, process the intelligent processing tasks based on the context information, update the context information based on the task execution results, and then return the processing results and the updated context information to the edge node, so that the execution node corresponding to the next intelligent processing task can execute the intelligent processing task based on the updated context information.

[0028] (4) a third-party service node, which is an external service provider independent of the CDN network, and is a dedicated AI inference model or a business service system accessed through a standard protocol such as a model context protocol (MCP). The third-party service node can be a subtitle generation service, a personalized advertising engine, and a content review platform, etc. It does not participate in CDN request routing and only responds to task calls from edge nodes on demand.

[0029] Unless otherwise defined, technical terms or scientific terms used in the present application shall have their ordinary meanings understood by those skilled in the art to which the present application belongs.

[0030] The terms "first", "second", and the like used in the present application do not indicate any order, number, or importance, but are only used to distinguish different components. The terms "include" or "contain" and the like mean that the elements or objects before the terms encompass the elements or objects listed after the terms and their equivalents, and do not exclude other elements or objects. The term "module" refers to any known or later developed hardware, software, firmware, artificial intelligence, fuzzy logic, or a combination of hardware or / and software code capable of performing functions related to the element.

[0031] It should be noted that the brief description of the terms in the present application is only for the convenience of understanding the embodiments described next, and is not intended to limit the embodiments of the present application. Unless otherwise stated, these terms should be understood in accordance with their ordinary and general meanings.

[0032] A CDN is a network technology that accelerates content access by deploying distributed edge node servers around the world to cache static content (such as web pages, pictures, and video files, etc.) to nodes closer to users.

[0033] However, with the popularization of business models with higher real-time attributes such as personalized recommendation, real-time interaction, and artificial intelligence generated content (AIGC), more and more business scenarios require the CDN not only to distribute static content, but also to perform real-time intelligent processing during distribution, such as dynamically inserting personalized advertisements according to user characteristics, generating real-time multilingual subtitles for live streams, or filtering content based on context, etc.

[0034] To adapt to this trend, existing solutions deploy lightweight AI models at CDN nodes, and the models are deployed as functionally isolated single-point services at edge nodes. The processing flow of the models is limited to one-time input and output. Moreover, due to the stateless nature of nodes in traditional CDN architecture, i.e., nodes process requests independently, nodes neither share user states nor transfer intermediate information generated during processing, which leads to the inability to support complex business processes that require multi-node or multi-model collaboration.

[0035] In view of the above problems, the embodiments of the present application provide a CDN service method, system, device and equipment. The CDN service method comprises: receiving, at an edge node, a content access request comprising a service chain; sequentially performing each intelligent processing task according to the service chain; sending, when each intelligent processing task is performed, context information and the current intelligent processing task to an execution node corresponding to the current intelligent processing task; receiving a task execution result returned by the execution node and context information used when the next intelligent processing task is performed; and finally generating a response content based on the received task execution results and returning the response content to a user terminal. Through the synchronization and updating of context information between nodes, the embodiments of the present application realize cross-node context collaboration, which helps to solve the problem that the existing CDN cannot support complex business processes that require multi-node or multi-model collaboration.

[0036] The specific embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0037] It should be noted that the following scenarios are only used to illustrate the embodiments of the present application and are not limiting. In specific implementation, the technical solutions provided by the embodiments of the present application can be flexibly applied according to actual needs.

[0038] As shown in FIG. 1, a schematic diagram of an application scenario is provided by the embodiments of the present application. The application scenario comprises a user terminal 10, an edge node 20, an execution node 30, and a preset cache library 40. Figure 1

[0039] In one possible scenario, the user terminal 10 sends a content access request to the edge node 20, and the content access request at least comprises a content identifier and a service chain. The content identifier is used to indicate the target content requested by the user, such as a video stream, a picture, or a web resource. The service chain is used to indicate one or more intelligent processing tasks that need to be sequentially performed on the target content, such as content review, subtitle generation, and code rate adaptation, etc.​

[0040] After receiving the content access request, the edge node 20 first determines the context information corresponding to the content access request, which can include information reflecting the user portrait characteristics and / or information reflecting the user access environment. Subsequently, the edge node 20 forms a cache query key based on the content identifier, the context information, and the service chain, and queries the preset cache library 40 to determine whether there is a response content corresponding to the cache query key.

[0041] If there is, the edge node 20 directly returns the queried response content to the user terminal 10; If there is not, the edge node 20 sequentially executes each intelligent processing task according to the order defined in the service chain. For each intelligent processing task, the edge node 20 sends the current context information and the intelligent processing task to the corresponding execution node 30.

[0042] After the execution node 30 executes the task, the context information is updated based on the task execution result, and the task execution result and the updated context information are returned, wherein the updated context information can be used as the context information for executing the next intelligent processing task.

[0043] The edge node 20 generates the response content based on all the received task execution results and returns it to the user terminal 10.

[0044] It can be understood that the number of execution nodes 30 can be one or more, and different execution nodes 30 can execute different types of intelligent processing tasks. Figure 1 The following will take one intelligent processing task as an example for illustration.

[0045] In the embodiments of the present application, the execution node 30 can be the edge node 20 itself that receives the content access request, can also be other nodes (such as a central node) in the CDN, or can be an independently deployed third-party service node. Wherein, when the execution node corresponding to a certain intelligent processing task is the edge node that receives the content access request, it means that the intelligent processing task can be completed by local resources, without the need for cross-node cooperation.

[0046] To further illustrate the technical solutions provided in the embodiments of this application, the CDN service method provided by exemplary embodiments of this application is described below with reference to the accompanying drawings. It should be noted that the above application scenarios are only shown to facilitate understanding of the spirit and principles of this application, and the embodiments of this application are not limited in any way in this regard. Furthermore, although the operations of the method of this application are described in a specific order in the accompanying drawings, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step, and / or one step may be decomposed into multiple steps.

[0047] like Figure 2 The diagram shown is a flowchart of a CDN service method applied to edge nodes according to an embodiment of this application. The method includes the following steps S201-S205.

[0048] Step S201: Receive a content access request sent by the user client.

[0049] In this embodiment of the application, the content access request includes a content identifier and a service chain; the content identifier is used to indicate the target content of the request; the service chain is used to indicate one or more intelligent processing tasks that need to be executed sequentially on the target content.

[0050] In this embodiment, the service chain can be generated on the user's end based on the function options selected by the user. For example, when it is determined that the user requests a basketball game live video stream with the content identifier "live_233" and selects real-time subtitles and personalized promotion functions, the service chain generated on the user's end can be ["subtitle_generation", "ad_insertion"]. This service chain defines two intelligent processing tasks that need to be executed sequentially on the acquired live stream: subtitle generation service and personalized promotion service. This provides support for the distributed scheduling and multi-task collaborative processing of subsequent edge nodes, that is, the personalized promotion service can be executed based on the context information updated by the subtitle generation service.

[0051] Step S202: Determine the context information corresponding to the content access request. The context information includes information reflecting user profile characteristics and / or information reflecting the user's access environment.

[0052] In this application embodiment, the information reflecting user profile characteristics refers to features that characterize user personalized attributes, such as feature data constructed based on user historical behavior, preferences, or account attributes. Specifically, these may include: user age group, gender, interest tags (such as "sports enthusiast"), historical content preference information, membership subscription status, and language preferences.

[0053] In the embodiments of the present application, the information reflecting the user access environment refers to information related to the access environment when the user terminal sends this content access request, such as dynamic parameters related to devices, networks or geographic locations, which can specifically be terminal device types (such as mobile phones, smart televisions, PCs), operating system versions, browser types, current network types (such as 5G, Wi-Fi, 4G), network bandwidth estimates, round-trip time (RTT), packet loss rate, IP addresses (for geographic location resolution), time zones, screen resolutions, audio output capabilities, and the like.

[0054] In the embodiments of the present application, the process of determining the context information is the process of integrating multiple data sources by the edge node to build a snapshot of the current request state. Specifically, after receiving the content access request, the edge node can extract part of the context information from the request header, Cookie, user authentication token or internal user database based on the preset acquisition method, and comprehensively determine the context information corresponding to the content access request in combination with real-time detection data (such as RTT, packet loss rate).

[0055] In the embodiments of the present application, the data sources and determination methods of the context information include but are not limited to the following three.

[0056] Firstly, the metadata of the content access request is extracted.

[0057] The edge node parses the received HTTP / HTTPS request message to directly extract the content corresponding to the preset field from the message header (Header) or message body (Body).

[0058] Example 1: The device and operating system information such as device type (mobile terminal, PC), operating system and browser model are parsed from the User-Agent field.

[0059] Example 2: The IP address of the user terminal is obtained from the X-Forwarded-For or X-Real-IP field, and the geographic location information (such as country, city and time zone) is parsed through the IP address library.

[0060] Example 3: The user identification (such as User-ID, Device-ID) is parsed from the content access request or Cookie, which is the key to associate more context data. Thus, the pre-computed user tags and preferences can be obtained by querying the internal user portrait database based on the user identification.

[0061] Secondly, the user identification is used for association query.

[0062] After obtaining the user identifier, the edge node can initiate an association query to obtain deeper user portrait information. First, the edge node queries whether the user portrait corresponding to the user identifier (such as a historical preference vector) is stored in its local storage. If the user portrait information corresponding to the user identifier is not queried in the local storage, the edge node obtains the user portrait information from other nodes or third-party service nodes. For example, the edge node can request a user portrait service from the CDN center node to obtain information such as long-term interest labels, historical behavior summaries (such as frequently watched video categories), and user value classification of the user.

[0063] Third, real-time detection and calculation.

[0064] For information representing the user access environment, part of the data can be obtained through real-time detection. For example, the edge node can measure network environment information such as current network delay (RTT), available bandwidth estimation, and connection type (4G / 5G / Wi-Fi) based on network interaction with the user terminal in real time.

[0065] In the embodiments of the present application, after the edge node obtains the original context information from different sources based on the above-mentioned methods, it can perform standardization and structurization processing on the original context information, and organize it into a structured data object, that is, the context information corresponding to the current content access request. For example, the device type is mapped to a predefined enumeration value (such as "mobile" / "tv" / "pc"), the IP address is converted to a country / city code, and the bandwidth estimation value is normalized to a discrete level (such as "low" / "medium" / "high"), and finally a unified format context object (such as a JSON structure) is formed, which is convenient for subsequent cache key generation and task scheduling.

[0066] For example, when a user who prefers Chinese subtitles requests video content through a certain brand of mobile phone via a 4G network in the evening, the edge node can determine the context information as follows: {"device_type": "mobile", "network_type": "4G", "time_of_day": "evening", "preferred_language": "zh-CN"}. The context information will be used for subsequent cache query key construction and intelligent processing task execution.

[0067] It should be noted that, in order to protect user privacy, the data related to personal identity or sensitive behavior in the embodiments of the present application are anonymized or desensitized in accordance with relevant laws and regulations, and the user portrait features are only used in the form of aggregated labels or hash IDs, without transmitting or storing original personal data.

[0068] In step S203, a cache query key is constructed based on the content identifier, the context information and the service chain, and a response content corresponding to the cache query key is queried in a preset cache library.

[0069] In the embodiments of the present application, the edge node combines the content identifier, the context information and the service chain to generate a globally unique cache query key. The combination method can be string splicing and then taking a hash value, or directly serializing into a key value. Based on the cache query mechanism of the cache query key, the content under different users, different contexts and different service requirements can be distinguished, and fine-grained personalized caching is achieved.

[0070] In the embodiments of the present application, the above-mentioned preset cache library can be deployed in different levels of the CDN network, including but not limited to: the local cache of the current edge node, the regional cache cluster serving multiple edge nodes, or the central cache system of the CDN. After the edge node generates the cache query key, the corresponding response content is queried in the corresponding cache library according to the preset cache query strategy (such as querying locally first, and then querying the regional cluster if not hit).

[0071] If the corresponding response content is queried in any cache level, it is directly jumped to step S205, so as to reduce the delay, the back-to-source rate and the computing overhead through the multi-level cache mechanism. This design not only ensures fast response of high-frequency requests at the edge, but also realizes cache reuse across nodes through shared regional or central cache, thereby improving the overall content distribution efficiency.

[0072] In step S204, if the response content corresponding to the cache query key is not queried, each intelligent processing task is executed according to the service chain.

[0073] In the above-mentioned process, when the edge node executes each intelligent processing task, the context information and the current intelligent processing task are sent to the execution node corresponding to the current intelligent processing task, and the task execution result returned by the execution node and the context information used when the next intelligent processing task is executed are received.

[0074] It should be noted that the intelligent processing task in the embodiments of the present application refers to a task unit that relies on artificial intelligence model for dynamic calculation and reasoning, such as personalized recommendation, real-time subtitle generation, accurate advertisement placement, content security audit, rate adaptation (streaming mode) and audio noise reduction. Each intelligent processing task can be executed by the corresponding execution node based on the context information as input, and the valuable semantic information generated in the reasoning process is extracted and added to the context information for use by subsequent intelligent processing tasks.

[0075] In the embodiments of the present application, the edge node first analyzes the service chain and determines the execution position of each intelligent processing task before performing each intelligent processing task according to the service chain. Specifically, the edge node determines whether the task can be executed locally by the node or needs to be forwarded to an external execution node according to the type of each intelligent processing task, the required model resource, the delay requirement and the capability of the node, and decides the optimal execution position for each task.

[0076] In step S205, the response content is generated based on the received execution results of each task and returned to the user end.

[0077] In some embodiments, the present application implements the task scheduling and context delivery between the edge node and the execution node based on the MCP protocol (Model Context Protocol). Specifically, the edge node is deployed with an MCP client, and each execution node is deployed with an MCP server. When the edge node needs to hand over an intelligent processing task to an execution node, the MCP client sends a request containing the current context information and the intelligent processing task to the corresponding MCP server. After the execution node completes the task, the context information is updated according to the predefined extraction specification, and the task result and the updated context information are returned to the edge node through the MCP protocol. The protocol provides standardized communication support for tool invocation and context delivery between distributed nodes, and provides a structured data basis for the generation of full-link tracking logs.

[0078] In some embodiments, before step S204, the above method further includes steps S301-S303.

[0079] In step S301, when it is detected that the content access request includes a service chain enhanced authorization identifier, the context information is matched with a predefined enhanced rule set. The enhanced rule set includes at least one enhanced rule, and each enhanced rule includes a rule trigger condition, an intelligent processing task to be inserted and a relative position relationship between the intelligent processing task to be inserted and part or all of the original intelligent processing tasks in the service chain.

[0080] In step S302, the target enhanced rule matched successfully with the context information is determined according to the rule trigger condition satisfied by the context information.

[0081] In step S303, the intelligent processing task included in the target enhanced rule is added to the service chain according to the relative position relationship included in the target enhanced rule.

[0082] Specifically, the rule trigger condition refers to a matching condition set based on a specific field in the context information, for example, triggering a corresponding enhanced rule when the network bandwidth is lower than a threshold, the device type is a mobile terminal, or the geographic location is in a specific area.

[0083] Exemplarily, specific fields in the context information can include: network_type: representing the type of the network the user is currently accessing, which can take values including "5G", "4G", "Wi-Fi", "3G", etc. bandwidth_level: representing the estimated level of the bandwidth available in the current network, which is usually calculated by the edge node based on real-time detection data such as RTT, throughput, etc., and can take values such as "high" (high, such as >5 Mbps), "medium" (medium, 1-5 Mbps), "low" (low, <1 Mbps); user_value_level: representing the value level or service level of the user, which is usually determined by the user's subscription status, payment behavior or operation label, and can take values such as "premium" (high-value user), "standard" (ordinary user), "trial" (trial user), etc.

[0084] Based on the above fields, the service chain can be optimized based on the enhanced rules as follows: If network_type == "4G" and bandwidth_level == "low", the rule triggering condition of low-bandwidth environment optimization is met, then a video transcoding task can be inserted at the beginning of the service chain to convert the high-definition video stream into a lower bit rate version that adapts to the current network conditions. This rule is suitable for mobile weak network scenarios, and by reducing the bit rate in advance, it can avoid playback lag; If user_value_level == "premium", the triggering condition of high-value user service enhancement is met, then a dedicated content recommendation task can be inserted. For example, based on the updated context (such as the identified video theme), exclusive promotion content (such as a star interview segment) that highly matches the preferences of high-value users is filtered and injected. This exclusive content can be inserted as an independent segment together with regular advertisements in the final video stream.

[0085] By defining relative position relationships (such as "inserted at the beginning of the service chain", "inserted after the subtitle generation task", "appended at the end of the service chain", etc.) in the enhanced rules, the application can realize adaptive enhancement of the service chain based on context information by the edge node without changing the original intention of the user. For example, in a weak network environment, a transcoding task can be automatically inserted to avoid the failure of subsequent subtitle or advertisement tasks due to lag.

[0086] In some embodiments, the execution node updates the context information based on the task execution result after completing the intelligent processing task. Specifically, the execution node extracts semantic information from the task execution result according to a predefined extraction specification, and adds the semantic information to the context information to obtain updated context information. The extraction specification defines the types of semantic information that need to be extracted from the task execution result, and the extraction method of each type of semantic information.

[0087] For example, when the intelligent processing task is content understanding and subtitle generation, the types of semantic information that need to be extracted can be video topic classification, key entity recognition, language type of generated subtitles, etc. After the extracted semantic information is added to the context information, it can be used for subsequent tasks such as advertisement recommendation, content review, etc., to ensure semantic coherence of the entire processing flow.

[0088] By updating the context information based on the task execution result, the present application can realize state sharing and semantic transmission among multiple tasks, and realize intelligent decision-making collaboration across tasks.

[0089] The following refers to the example of the service chain of the aforementioned embodiment [ "subtitle_generation", "ad_insertion" ], and introduces the process of the edge node executing each intelligent processing task according to the service chain.

[0090] First, if the edge node detects that the content access request contains the service chain enhancement authorization identifier, it triggers the service chain enhancement process. The edge node matches the context information with the enhancement rule set, finds that the user is currently in a subway environment (according to GPS information), and matches the "subway environment optimization" rule, which defines the insertion of a noise reduction task at the beginning of the service chain and the insertion of a data saving mode task after subtitle generation.

[0091] The enhanced service chain becomes: ["noise_reduction", "subtitle_generation", "data_saving", "ad_insertion"] After that, the edge node performs task decomposition and execution node judgment on the enhanced service chain. Specifically, the edge node parses the enhanced service chain and decomposes it into four intelligent processing tasks that are executed in sequence: noise reduction task, subtitle generation task, data saving mode task, and advertisement insertion task. Then, for each intelligent processing task, the edge node judges its corresponding execution node.

[0092] For the noise reduction task, based on its real-time audio processing characteristics, the edge node judges that it is directly executed by the node, performs background noise elimination on the live stream, and processes the audio stream as the input of the subsequent task.

[0093] For the subtitle generation task, if the current edge node has deployed a speech recognition model required by the subtitle generation task, the edge node directly calls the speech recognition model within the node, inputs the noise-reduced live stream and the current context information into the model, generates real-time subtitle results, extracts video theme information (such as "sports event") from the recognition results and updates the context information, and returns the real-time subtitle results and the updated context information to the edge node.

[0094] For the stream saving mode task, the edge node determines that a special video processing node is required to perform the task, and converts the video stream to a resolution and code rate suitable for a mobile network.

[0095] For the promotion insertion task, the edge node sends the context information including video theme, user preference, etc. to the center node. The center node calls the personalized recommendation model to perform advertisement matching based on the context information, outputs an advertisement content related to the sports event, for example, selects a silent sports text advertisement (to avoid conflict with the subtitle voice), and inserts it into the intermission period, and returns the task execution result to the edge node.

[0096] Finally, the edge node receives all task execution results: noise-reduced audio, generated subtitles, transcoded video stream and personalized advertisement content, combines these contents into the final response content and returns it to the user end, and stores the cache query key composed of content identification, original context information and original service chain in association with the response content in the preset cache library.

[0097] In some embodiments, the CDN service method applied to the edge node described above further includes the following steps S401-S404.

[0098] Step S401, receiving a node link tracking record generated by an execution node.

[0099] The node link tracking record includes at least one of the following information: node identification, model identification information of a target model called by the execution node when performing the current intelligent processing task, and input data summary, output data summary and task processing time consumption of each target model.

[0100] Specifically, each execution node generates a node link tracking record after completing its corresponding intelligent processing task, and returns it to the edge node that schedules the task.

[0101] Step S402, generating a link tracking sequence corresponding to the content access request based on the node link tracking records generated by the execution nodes corresponding to the intelligent processing tasks in the service chain.

[0102] Edge nodes, following the service chain execution order, sort the node tracing records returned by each task by timestamp and assemble them into a tracing sequence. For example, the tracing sequence of content access request i can be represented by the following formula. :

[0103] In the formula: Node identifier ( ): The identity identifier of the execution node j; Execution model ( ): Records the name of the artificial intelligence model invoked when execution node j calls the artificial intelligence model to perform intelligent processing tasks; Model version ( ): Records the version number of the artificial intelligence model called by execution node j; Input data summary ( ): The hash value or structured summary calculated from the input data of the artificial intelligence model invoked on execution node j; Output data summary ( ): The hash value or structured digest calculated from the output of the artificial intelligence model invoked on execution node j; Processing time ( ): The time taken for the AI ​​model invoked by node j to go from input to output; Completion timestamp ( ): The time (e.g., Unix timestamp) when the AI ​​model calling node j completes the task.

[0104] Each item is arranged in chronological order, fully reflecting the end-to-end link of the request.

[0105] Step S403: The context information received from the execution node corresponding to the last intelligent processing task in the service chain is determined as the context information of the final version.

[0106] The context information in this final version reflects a snapshot of the state after the entire service chain has been executed. It includes the original context information and the semantic information injected into the original context information by all intelligent processing tasks. It can be used for subsequent log association, user behavior analysis, or caching strategy optimization.

[0107] Step S404: Based on the content access request, the context information of the final version, the service chain, and the link tracing sequence, generate the full link tracing log corresponding to the content access request.

[0108] Specifically, the edge node encapsulates the following elements into a structured end-to-end tracing log. :

[0109] In the formula, For content access requests, For the context information of the final version, For the service chain, for The corresponding link tracing sequence; To request a final status (such as "success", "failed", "partial_success"); , They are respectively The start and end times of the request are the time when the edge node receives the content access request and the time when the edge node returns the response content to the user. For error information (if the status is unsuccessful, record the exception type, stack summary, or error code).

[0110] Through the aforementioned end-to-end tracing mechanism, this application embodiment achieves complete observability for each intelligent CDN service request, providing detailed data support for system operation and maintenance, fault diagnosis, performance optimization, and compliance auditing. For example, operation and maintenance personnel can... Track the flow path of a single request between edge, region, or central nodes; the algorithm team can combine... and Analyze the model's performance; the product team can utilize the final context. Evaluate the benefits of personalized strategies.

[0111] Based on the same inventive concept, this application also provides a CDN service method applied to an execution node, the method including the following steps S501-S504: Step S501: Receive context information and intelligent processing tasks to be executed from the edge node; Step S502: Execute intelligent processing tasks based on context information to obtain task execution results; Step S503: Extract semantic information from the task execution result according to the predefined extraction specification, and add the semantic information to the context information to obtain the context information used by the edge node when executing the next intelligent processing task; wherein, the extraction specification defines the types of semantic information that need to be extracted from the task execution result, as well as the extraction method of each type of semantic information; Step S504: Return the task execution result and the context information used when executing the next intelligent processing task to the edge node.

[0112] In some embodiments, the CDN service method applied to the execution node further includes: generating a node link tracking record, and returning the node link tracking record to the edge node. The node link tracking record includes at least one of the following information: node identification, model identification information of a target model called by the execution node when performing an intelligent processing task, and input data digest, output data digest and task processing time consumption of each target model.

[0113] The specific implementation of the CDN service method applied to the execution node can refer to the foregoing embodiments, which will not be described here.

[0114] Based on the same inventive concept, the embodiments of the present application also provide a CDN service method applied to a user end, which includes the following steps S601-S604: Step S601: determining a content identification corresponding to a target content requested by a user; Step S602: generating a service chain according to a function option selected by the user, the service chain being used to indicate one or more intelligent processing tasks that need to be performed on the target content in sequence; Step S603: generating a content access request based on the content identification and the service chain, and sending the content access request to an edge node, so that the edge node performs each intelligent processing task according to the service chain when no response content corresponding to a cache query key is queried; wherein the cache query key is composed of the content identification, context information corresponding to the content access request and the service chain, the context information including information reflecting user portrait features and / or information reflecting user access environment; when performing each intelligent processing task, the edge node sends the context information and the current intelligent processing task to an execution node corresponding to the current intelligent processing task, and receives a task execution result returned by the execution node and context information used when performing a next intelligent processing task; Step S604: receiving a response content returned by the edge node, the response content being generated by the edge node based on the received task execution results.

[0115] In some embodiments, the CDN service method applied to the user end further includes: determining whether to add a service chain enhancement authorization identifier to the content access request according to service chain enhancement configuration information, the service chain enhancement authorization identifier being used to indicate whether the edge node is allowed to add an intelligent processing task to the service chain, so that when the edge node detects that the content access request includes the service chain enhancement authorization identifier, the edge node matches the context information with a predefined enhancement rule set, and determines whether to add an intelligent processing task to the service chain based on a matching result; The enhanced rule set includes at least one enhanced rule. Each enhanced rule includes the rule triggering condition, the smart processing task to be inserted, and the relative positional relationship between the smart processing task to be inserted and some or all of the original smart processing tasks in the service chain.

[0116] The specific implementation of the CDN service method applied to the user end described above can be referred to the foregoing embodiments, and will not be repeated here.

[0117] Based on the same inventive concept, embodiments of this application also provide a CDN service system, which includes: Edge nodes are used to execute the CDN service methods described above that are applied to edge nodes; The execution node is used to execute the CDN service methods applied to the execution node as described above; The user-side is used to execute the CDN service methods described above.

[0118] For specific implementation methods, please refer to the foregoing embodiments, which will not be repeated here.

[0119] Based on the same inventive concept, embodiments of this application also provide a CDN service device, such as... Figure 3 As shown, the device includes: The request receiving module 301 is used to receive content access requests sent by the user terminal. The content access request includes a content identifier and a service chain. The content identifier is used to indicate the target content of the request. The service chain is used to indicate one or more intelligent processing tasks that need to be executed sequentially on the target content. The context information determination module 302 is used to determine the context information corresponding to the content access request. The context information includes information reflecting user profile characteristics and / or information reflecting the user's access environment. The cache query module 303 is used to construct a cache query key based on the content identifier, context information and service chain, and query the response content corresponding to the cache query key in the preset cache library; The task execution module 304 is used to execute each intelligent processing task according to the service chain if no response content corresponding to the cache query key is found. When executing each intelligent processing task, the context information and the current intelligent processing task are sent to the execution node corresponding to the current intelligent processing task, and the task execution result returned by the execution node is received, as well as the context information used when executing the next intelligent processing task. The response module 305 is used to generate response content based on the received task execution results and return it to the user terminal.

[0120] In some embodiments, the apparatus further comprises a service chain enhancement module configured to, when the content access request comprises a service chain enhancement authorization identifier, match the context information with a predefined enhancement rule set; wherein the enhancement rule set comprises at least one enhancement rule, and each enhancement rule comprises a rule trigger condition, an intelligent processing task to be inserted, and a relative position relationship between the intelligent processing task to be inserted and some or all original intelligent processing tasks in the service chain; determine a target enhancement rule that matches the context information successfully according to a rule trigger condition satisfied by the context information; and add the intelligent processing task included in the target enhancement rule to the service chain according to the relative position relationship included in the target enhancement rule.

[0121] In some embodiments, the apparatus further comprises a link tracking module configured to receive a node link tracking record generated by an execution node; the node link tracking record comprises at least one of the following information: a node identifier, model identifier information of a target model called by the execution node when executing a current intelligent processing task, and input data digest, output data digest, and task processing time consumption of each target model; generate a link tracking sequence corresponding to the content access request based on the node link tracking records generated by the execution nodes corresponding to the intelligent processing tasks in the service chain; determine the context information received from the execution node corresponding to the last intelligent processing task in the service chain as the final version of the context information; and generate a full link tracking log corresponding to the content access request based on the content access request, the final version of the context information, the service chain, and the link tracking sequence.

[0122] Based on the same inventive concept, the embodiments of the present application also provide a CDN service apparatus, as shown in Figure 4 The apparatus comprises: a task receiving module 401 configured to receive context information and an intelligent processing task to be executed from an edge node; a task processing module 402 configured to execute the intelligent processing task based on the context information to obtain a task execution result; a context updating module 403 configured to extract semantic information from the task execution result according to a predefined extraction specification, and add the semantic information to the context information to obtain context information used by the edge node when executing a next intelligent processing task; wherein the extraction specification defines a type of semantic information that needs to be extracted from the task execution result, and an extraction manner of each type of semantic information; a task response module 404 configured to return the task execution result and the context information used when executing the next intelligent processing task to the edge node.

[0123] In some embodiments, the apparatus further comprises a node link tracking module configured to generate a node link tracking record and return the node link tracking record to the edge node; wherein the node link tracking record comprises at least one of the following information: a node identifier, model identifier information of a target model invoked by the execution node when performing the intelligent processing task, and input data digest, output data digest, and task processing time consumption of each target model.

[0124] Based on the same inventive concept, the embodiments of the present application also provide a CDN service apparatus, as shown in Figure 5 The apparatus comprises: an information determining module 501 configured to determine a content identifier corresponding to a target content requested by a user; a service chain generating module 502 configured to generate a service chain according to a function option selected by the user, the service chain being used to indicate one or more intelligent processing tasks that need to be performed on the target content in sequence; a request sending module 503 configured to generate a content access request based on the content identifier and the service chain, and send the content access request to the edge node, so that the edge node performs each intelligent processing task in the service chain when no response content corresponding to a cache query key is queried; wherein the cache query key is composed of the content identifier, context information corresponding to the content access request, and the service chain; the context information comprises information reflecting a user portrait feature and / or information reflecting a user access environment; when performing each intelligent processing task, the edge node sends the context information and the current intelligent processing task to an execution node corresponding to the current intelligent processing task, receives a task execution result returned by the execution node, and uses the context information when performing a next intelligent processing task; a response receiving module 504 configured to receive the response content returned by the edge node, the response content being generated by the edge node based on the received task execution result.

[0125] In some embodiments, the apparatus further comprises an enhanced authorization identifier configuring module configured to determine whether to add a service chain enhanced authorization identifier to the content access request according to service chain enhanced configuration information, the service chain enhanced authorization identifier being used to indicate whether the edge node is allowed to add an intelligent processing task to the service chain, so that when the edge node detects that the content access request comprises the service chain enhanced authorization identifier, the edge node matches the context information with a predefined enhancement rule set, and determines whether to add an intelligent processing task to the service chain based on a matching result; wherein the enhancement rule set comprises at least one enhancement rule, and each enhancement rule comprises a rule trigger condition, an intelligent processing task to be inserted, and a relative position relationship between the intelligent processing task to be inserted and part or all original intelligent processing tasks in the service chain.

[0126] Based on the same inventive concept, the embodiments of the present application also provide a CDN service apparatus 600, as shown inFigure 6 As shown, the device includes at least one processor 602; and a memory 601 connected with the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the CDN service method applied to the edge node, or perform the CDN service method applied to the execution node, or perform the CDN service method applied to the user end.

[0127] The memory 601 is configured to store programs. Specifically, the programs can include program codes, and the program codes include computer operation instructions. The memory 601 can be a volatile memory, such as a random-access memory (RAM); or a non-volatile memory, such as a flash memory, a hard disk drive (HDD) or a solid-state drive (SSD); or a combination of any one or more of the above-mentioned volatile memories and non-volatile memories.

[0128] The processor 602 can be a central processing unit (CPU), a network processor (NP), or a combination of the CPU and the NP. It can also be a hardware chip. The hardware chip can be an application-specific integrated circuit (ASIC), a programmable logic device (PLD) or a combination thereof. The PLD can be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL) or any combination thereof.

[0129] The embodiments of the present application also provide a computer readable storage medium, including instructions, which, when executed on a computer, cause the computer to perform the CDN service method applied to the edge node, or perform the CDN service method applied to the execution node, or perform the CDN service method applied to the user end.

[0130] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the apparatus and the module described above can refer to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0131] In several embodiments provided in the present application, it should be understood that the disclosed system, apparatus and method can be implemented in other ways. For example, the apparatus embodiments described above are merely schematic, for example, the division of the modules is only a logical function division, and actual implementation can have another division manner, for example, a plurality of modules or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed modules can be indirect coupling or communication connection through some interfaces, devices or modules, and can be electrical, mechanical or other forms.

[0132] The modules described as separate components can or can not be physically separated, and the components displayed as modules can or can not be physical modules, that is, can be located in one place, or can be distributed to a plurality of network modules. Part or all of the modules can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0133] In addition, the functional modules in each embodiment of the present application can be integrated in one processing module, or each module can exist physically, or two or more modules can be integrated in one module. The integrated module can be realized in the form of hardware or in the form of software functional module. When the integrated module is realized in the form of software functional module and sold or used as an independent product, it can be stored in a computer readable storage medium.

[0134] In the above embodiments, all or part can be realized by software, hardware, firmware or any combination thereof. When realized by software, it can be realized in the form of computer program product in whole or in part.

[0135] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a server or data center that integrates one or more available media. The available medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state disk (SSD)).

[0136] The technical solutions provided in this application have been described in detail above. Specific examples have been used in this application to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

[0137] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0138] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1apparatus for performing the functions specified in a flow or flows and / or blocks. Figure 1

[0139] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the flow Figure 1 apparatus for performing the functions specified in a flow or flows and / or blocks. Figure 1

[0140] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the flow Figure 1 apparatus for performing the functions specified in a flow or flows and / or blocks. Figure 1

[0141] Obviously, numerous modifications and variations of the present application are possible in light of the above teachings. It is therefore to be understood that within the scope of the apparatus for performing the functions specified in a flow or flows and / or blocks. Figure 1

[0139] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the flow Figure 1 apparatus for performing the functions specified in a flow or flows and / or blocks. Figure 1

[0140] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the flow Figure 1 apparatus for performing the functions specified in a flow or flows and / or blocks. Figure 1

[0141] Obviously, numerous modifications and variations of the present application are possible in light of the above teachings. It is therefore to be understood that within the scope of the apparatus for performing the functions specified in a flow or flows and / or blocks.

Claims

1. A CDN service method, applied to an edge node, characterized in that, The method comprises the following steps: receiving a content access request sent by a user terminal, the content access request comprising a content identifier and a service chain; the content identifier is used to indicate a target content requested; the service chain is used to indicate one or more intelligent processing tasks that need to be sequentially performed on the target content; determining context information corresponding to the content access request, the context information comprising information reflecting user portrait characteristics and / or information reflecting user access environment; forming a cache query key based on the content identifier, the context information and the service chain, and querying a response content corresponding to the cache query key in a preset cache library; if the response content corresponding to the cache query key is not queried, performing each intelligent processing task in the service chain; wherein when each intelligent processing task is performed, the context information and the current intelligent processing task are sent to an execution node corresponding to the current intelligent processing task, and the execution result returned by the execution node and the context information used when the next intelligent processing task is performed are received; generating a response content based on the received execution result of each task and returning the response content to the user terminal.

2. The method of claim 1, wherein, Before the method performs each intelligent processing task in the service chain, the method further comprises the following steps: when it is detected that the content access request comprises a service chain enhanced authorization identifier, matching the context information with a predefined enhanced rule set; wherein the enhanced rule set comprises at least one enhanced rule, and each enhanced rule comprises a rule trigger condition, an intelligent processing task to be inserted, and a relative position relationship between the intelligent processing task to be inserted and part or all original intelligent processing tasks in the service chain; determining a target enhanced rule matched successfully with the context information according to the rule trigger condition satisfied by the context information; adding the intelligent processing task included in the target enhanced rule to the service chain according to the relative position relationship included in the target enhanced rule.

3. The method of claim 1, wherein, The method further comprises the following steps: receiving a node link tracking record generated by the execution node; the node link tracking record comprises at least one of the following information: a node identifier, model identifier information of a target model called by the execution node when the execution node executes the current intelligent processing task, and input data summary, output data summary and task processing time consumption of each target model; generating a link tracking sequence corresponding to the content access request based on the node link tracking record generated by the execution node corresponding to each intelligent processing task in the service chain; determining the context information received from the execution node corresponding to the last intelligent processing task in the service chain as the final version of the context information; generating a full link tracking log corresponding to the content access request based on the content access request, the final version of the context information, the service chain and the link tracking sequence. 4.A CDN service method, applied to an execution node, and characterized in that, The method comprises the following steps: receiving context information and an intelligent processing task to be executed from an edge node; performing the intelligent processing task based on the context information to obtain a task execution result; extract semantic information from the task execution result according to a predefined extraction specification, and add the semantic information to the context information, to obtain context information used by the edge node when performing a next intelligent processing task; wherein the extraction specification defines types of semantic information that need to be extracted from the task execution result, and an extraction manner of each type of semantic information; return the task execution result and the context information used when performing a next intelligent processing task to the edge node.

5. The method of claim 4, wherein, The method further comprises: generating a node link tracking record, and returning the node link tracking record to the edge node; wherein the node link tracking record comprises at least one of the following information: node identification, model identification information of a target model called by the execution node when performing the intelligent processing task, and input data summary, output data summary, and task processing time consumption of each target model. 6.A CDN service method, applied to a user end, and characterized in that, comprises: determining a content identification corresponding to target content requested by a user; generating a service chain according to a function option selected by the user, the service chain being used to indicate one or more intelligent processing tasks that need to be sequentially performed on the target content; generating a content access request based on the content identification and the service chain, and sending the content access request to an edge node, so that the edge node performs each intelligent processing task according to the service chain when no response content corresponding to a cache query key is queried; wherein the cache query key is composed of the content identification, context information corresponding to the content access request, and the service chain, the context information comprising information reflecting a user portrait feature and / or information reflecting a user access environment; when performing each intelligent processing task, the edge node sends the context information and the current intelligent processing task to an execution node corresponding to the current intelligent processing task, and receives a task execution result returned by the execution node and context information used when performing a next intelligent processing task; receiving response content returned by the edge node, the response content being generated by the edge node based on the received task execution result.

7. The method of claim 6, wherein, The method further comprises: determining whether to add a service chain enhancement authorization identifier to the content access request according to service chain enhancement configuration information, the service chain enhancement authorization identifier being used to indicate whether the edge node is allowed to add an intelligent processing task to the service chain, so that when the edge node detects that the content access request comprises a service chain enhancement authorization identifier, the edge node matches the context information with a predefined enhancement rule set, and determines whether to add an intelligent processing task to the service chain based on a matching result; wherein the enhancement rule set comprises at least one enhancement rule, and each enhancement rule comprises a rule trigger condition, a to-be-inserted intelligent processing task, and a relative position relationship between the to-be-inserted intelligent processing task and part or all original intelligent processing tasks in the service chain.

8. A CDN service apparatus characterized by comprising: comprises: a request receiving module, configured to receive a content access request sent by a user end, the content access request comprising a content identification and a service chain; The content identifier is configured to indicate the target content requested by the user; The service chain is configured to indicate one or more intelligent processing tasks that need to be sequentially performed on the target content; A context information determination module is configured to determine context information corresponding to the content access request, the context information including information reflecting user portrait characteristics and / or information reflecting a user access environment; A cache query module is configured to form a cache query key based on the content identifier, the context information, and the service chain, and query a response content corresponding to the cache query key in a preset cache library; A task execution module is configured to perform each intelligent processing task according to the service chain if the response content corresponding to the cache query key is not queried; wherein, when each intelligent processing task is executed, the context information and the current intelligent processing task are sent to an execution node corresponding to the current intelligent processing task, a task execution result returned by the execution node is received, and context information used when a next intelligent processing task is executed is received; A response module is configured to generate a response content based on each received task execution result and return the response content to the user terminal.

9. A CDN service system, characterized by, The edge node is configured to execute the method according to any one of claims 1-3; The execution node is configured to execute the method according to any one of claims 4-5; The user terminal is configured to execute the method according to any one of claims 6-7. The at least one processor and the memory connected with the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method according to any one of claims 1-3, or execute the method according to any one of claims 4-5, or execute the method according to any one of claims 6-7.

10. A CDN service device, characterized by ​

Images