Methods, apparatus, computer equipment, and storage media for canary deployment of target services
By constructing a configuration dependency graph and executing simulated test cases, the problem of insufficient verification for configuration updates on mini-program activity pages was solved. This enabled automated identification of potential faults and dynamic adjustment of canary deployment strategies, improving the stability of configuration releases and the business experience.
Patent Information
- Application Number
- CN202611257057.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-18
- Publication Date
- 2026-11-17
AI Technical Summary
In fintech scenarios, the configuration updates of mini-program activity pages lack an effective pre-launch verification mechanism, resulting in a mismatch between the configuration and the downstream business status, making it difficult to predict runtime anomalies. Existing technologies rely on insufficient coverage of manual testing and cannot identify potential faults in advance.
By parsing the configuration data of the target service page, a configuration dependency graph is constructed. Based on the dependency graph, verification and simulated test case execution are performed to dynamically generate canary release strategies, automatically identify problems in the configuration data, detect operational failures in advance, and dynamically adjust the canary release strategy.
It enables automated pre-validation of configuration data for mini-program activity pages, covering various business scenarios, reducing the probability of online failures, ensuring a smooth business operation experience, and improving the stability of configuration release.
Smart Images

Figure CN123114528A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of development and testing technology, and is applied to financial technology scenarios. In particular, it relates to a method, apparatus, computer equipment, and storage medium for the canary release of a target service. Background Technology
[0002] In fintech scenarios, mini-program activity pages commonly employ a configurable setup solution. Operations personnel configure banners, product cards, coupons, benefit entry points, redirect links, API parameters, and activity durations through the operations backend. The mini-program frontend retrieves this configuration information at runtime, parses and renders the page, and page content iteration can be completed simply by changing the configuration data, without requiring a new mini-program client release. Currently, the industry has related systems such as page building platforms, low-code activity page building tools, canary release platforms, and configuration centers, but these capabilities are independent and fragmented, lacking a pre-launch verification mechanism specifically for mini-program activity pages.
[0003] Traditional deployment verification relies on manual testing, which can only cover basic visuals and the main process. It is difficult to predict runtime anomalies caused by mismatches between configuration and downstream business status during the operation of the mini-program. After the configuration is updated and the program enters the canary release phase, potential faults cannot be identified in advance, which can easily lead to business anomalies in the online environment after the canary release. Summary of the Invention
[0004] This invention provides a method, apparatus, computer equipment, and medium for the canary release of a target service, aiming to solve the technical problem of the difficulty in effectively managing online risks caused by configuration changes and to realize dynamic adjustment of the canary release strategy.
[0005] Firstly, a method for canary deployment of a target service is provided, including: Obtain the configuration data of the target service page of the target service to be published, and parse the configuration data to obtain the configuration object; The configuration object is converted into a node, and directed edges are established between the nodes based on the dependencies of the configuration object in the configuration data to obtain a configuration dependency graph; The verification rules are executed according to the verification execution path indicated by the configuration dependency graph to obtain the verification result of the configuration data; Extract simulated use cases from the configuration dependency graph, execute the simulated use cases, and obtain the simulation results of the simulated use cases; Based on the verification results and the simulation results, a grayscale strategy is determined, and the configuration data of the target service page is distributed according to the grayscale strategy.
[0006] Secondly, a canary deployment device for a target service is provided, comprising: The configuration module is used to obtain the configuration data of the target service page of the target service to be published, and parse the configuration data to obtain the configuration object; The graph construction module is used to convert the configuration object into nodes and establish directed edges between the nodes according to the dependency relationship of the configuration object in the configuration data to obtain a configuration dependency graph. The verification module is used to execute verification rules according to the verification execution path indicated by the configuration dependency graph and obtain verification results; The test case module is used to extract simulated test cases from the configuration dependency graph, execute the simulated test cases, and obtain the simulation results of the simulated test cases; The grayscale execution module is used to determine the grayscale strategy based on the verification results and the simulation results, and to distribute the configuration data of the target service page according to the grayscale strategy.
[0007] Thirdly, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the canary release method for the aforementioned target service.
[0008] Fourthly, a computer-readable storage medium is provided, which stores a computer program that, when executed by a processor, implements the steps of the canary release method for the aforementioned target service.
[0009] The solution implemented by the aforementioned canary release method, apparatus, computer equipment, and storage medium for the target service, by parsing the configuration data of the target service page to be released, constructing a configuration dependency graph, and planning the verification execution path according to the configuration dependency graph, performs pre-verification of the configuration data. This automatically identifies structural problems in the page configuration data, such as invalid resource references, abnormal jump links, and mismatched interface parameters, without relying entirely on manual item-by-item inspection, effectively compensating for the insufficient coverage of manual verification. Simultaneously, based on the configuration dependency graph, it automatically extracts simulated test cases and executes simulation tests, covering various business scenarios and identifying potential operational failures during the target service's runtime phase in advance, preventing related issues from flowing into the canary release stage. Furthermore, it comprehensively assesses the potential risks of configuration changes by integrating the verification results of the configuration data with the simulation results of the simulated test cases, dynamically generating canary release strategies. This significantly reduces the probability of online failures caused by configuration updates to the target service page, ensuring a smooth business operation experience. Attached Figure Description
[0010] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0011] Figure 1 This is a schematic diagram of an application environment for a canary release method of a target service in one embodiment of this application; Figure 2 This is a flowchart illustrating a canary release method for a target service in one embodiment of this application; Figure 3 This is a flowchart illustrating a canary release method for a target service in one embodiment of this application; Figure 4 This is a flowchart illustrating a canary release method for a target service in one embodiment of this application; Figure 5 This is a flowchart illustrating a canary release method for a target service in one embodiment of this application; Figure 6 This is a schematic diagram of the overall process of a canary release method for a target service in one embodiment of this application; Figure 7 This is a schematic diagram of a grayscale publishing device for a target service in one embodiment of this application; Figure 8 This is a schematic diagram of the structure of a computer device according to one embodiment of this application; Figure 9 This is another structural schematic diagram of a computer device in one embodiment of this application. Detailed Implementation
[0012] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.
[0013] The target service can be a webpage, an app, a mini-program, etc. Taking a mini-program activity page as an example, the relevant technologies use a configuration-based approach. The operation backend maintains configuration information such as page components, interfaces, redirection rules, and timeliness. The front-end mini-program parses the configuration to render the page, and modifying the configuration updates the page. Traditional online verification relies on manual testing, which can only cover basic visuals and the main process, failing to fully cover scenarios with multiple combinations of user identity, region, and client version. It's difficult to detect configuration compatibility issues during mini-program runtime in advance; faults often only surface after canary releases or real user access.
[0014] For example, in the business of vehicle service apps within financial scenarios, mini-program activity pages often encompass links such as car insurance promotions, car wash and maintenance benefits, points redemption, membership activities, service coupon redemption, and product guides. A single activity page may simultaneously rely on multiple backend interfaces, including product inventory, coupon batches, membership levels, vehicle information, policy status, order payment, and service fulfillment. If a product is no longer available but its configuration still references it, users will encounter a blank page or fail to place an order after clicking; if a coupon batch is not yet active, users will fail to redeem benefits; if the redirect path points to an older version of the mini-program page, some users may be unable to open it; if the interface returns fields and component versions do not match, the page may exhibit missing cards, abnormal price display, or unclickable buttons. Therefore, issuing configuration data without verification can easily lead to online functional abnormalities.
[0015] To address the aforementioned issues, this application provides a method for the canary release of a target service, applicable to applications such as... Figure 1 In the application environment, Figure 1This process includes the business side, server, and user side. After the business side submits configuration data to the server, the server verifies the configuration data, generates a canary release strategy, and then distributes it to the user side. Specifically, by obtaining the configuration data of the target service page of the target service to be released, parsing the configuration data to obtain configuration objects, converting the configuration objects into nodes, and establishing directed edges between nodes based on the dependencies of the configuration objects in the configuration data, a configuration dependency graph is obtained. Verification rules are executed according to the verification execution path indicated by the configuration dependency graph to obtain the verification results of the configuration data. Simulated test cases are extracted from the configuration dependency graph, executed, and the simulation results of the simulated test cases are obtained. The canary release strategy is determined based on the verification results and simulation results, and the configuration data of the target service page is distributed according to the canary release strategy. This method can automatically identify problems in the page configuration data without relying entirely on manual item-by-item inspection, effectively making up for the shortcomings of insufficient coverage of manual verification. At the same time, automatically extracting simulated test cases and executing simulation tests based on the configuration dependency graph can cover various business scenarios, detect potential operational failures in the target service's runtime phase in advance, and prevent related problems from flowing into the canary release stage. Furthermore, by comprehensively evaluating the potential risks of configuration changes based on the verification results of the integrated configuration data and the simulation results of the simulated test cases, a canary deployment strategy is dynamically generated, which significantly reduces the probability of online failures caused by configuration updates on the target service page and ensures a smooth business operation experience.
[0016] The server can be, but is not limited to, various personal computers, laptops, smartphones, tablets, and portable wearable devices. The present application will be described in detail below through specific embodiments.
[0017] Please see Figure 2 As shown, Figure 2 A flowchart illustrating the canary release method for the target service provided in this application embodiment includes the following steps: S101, Obtain the configuration data of the target service page of the target service to be published, and parse the configuration data to obtain the configuration object; S102, the configuration object is converted into a node, and directed edges are established between the nodes according to the dependency relationship of the configuration object in the configuration data to obtain a configuration dependency graph; S103, execute the verification rules according to the verification execution path indicated by the configuration dependency graph to obtain the verification result of the configuration data; S104, Extract simulated use cases from the configuration dependency graph, execute the simulated use cases, and obtain the simulation results of the simulated use cases; S105, determine the grayscale strategy based on the verification result and the simulation result, and distribute the configuration data of the target service page according to the grayscale strategy.
[0018] The canary release method for the target service provided in this application can be applied to canary releases in various application scenarios. The target service can be a software service, a web service, a mini-program service, etc. For example, in a fintech scenario, the target service can be an insurance application, and the target service page can be a member activity page.
[0019] In step S101 of some embodiments, configuration data of the target service page of the target service to be published is obtained, and the configuration data is parsed to obtain a configuration object. Specifically, the system receives the configuration data of the target service page submitted by the operations backend, and parses the basic page information, component tree, data source, interface parameters, jump path, user groups, activity time and event tracking configuration in sequence, converting the above-mentioned multiple types of information into a configuration object.
[0020] For example, the target service page is a mini-program activity page within a vehicle service app that carries out business links such as car insurance promotions, car wash and maintenance benefits, points redemption, membership activities, service coupon redemption, and product guides. By using structured parsing, the heterogeneity and inconsistent formatting of the original configuration fields are eliminated, outputting standardized configuration objects that can be recognized by subsequent processes, thus normalizing the original configuration information.
[0021] In step S102 of some embodiments, configuration objects are converted into nodes, and directed edges are established between nodes based on the dependencies of configuration objects in the configuration data to obtain a configuration dependency graph. Discrete configuration objects are abstracted into nodes in a graph structure. Further, directed edges are constructed between nodes based on business calls, display, redirection, and conditional rendering relationships, transforming the configuration data into a dependency topology graph that can express the order of business processes and resource reference relationships.
[0022] In one feasible implementation, the system maps standardized configuration objects to different types of dependency nodes. For example, node types include page nodes, component nodes, interface nodes, product nodes, coupon nodes, route nodes, user rule nodes, and event tracking nodes.
[0023] In step S103 of some embodiments, the verification rules are executed according to the verification execution path indicated by the configuration dependency graph to obtain the verification result of the configuration data. It should be noted that the configuration dependency graph is not only used to list all configuration items, but can also provide an ordered verification execution path. Business links have clear sequential dependencies; unordered verification can easily lead to undetected anomalies in upstream resources, resulting in distorted downstream verification results. In one feasible implementation, the system traverses the topological path of the configuration dependency graph, schedules the corresponding verification rules along the flow order of directed edges, and sequentially verifies resource reference validity, interface parameter matching, jump route legality, activity timeliness constraints, user circle rendering conditions, etc., summarizing the verification output information of all nodes to form the configuration data verification result.
[0024] In step S104 of some embodiments, simulated use cases are extracted from the configuration dependency graph, executed, and simulation results are obtained. It is understood that the configuration dependency graph records all business branches of the page, user filtering conditions, and page flow links. Based on this topology information, it can adapt to diverse user scenarios, simulate real user access links, and expose hidden problems that only appear during page runtime in advance. The simulation results include the success status and reasons for exceptions for each path. In step S105 of some embodiments, a canary release strategy is determined based on the verification results and simulation results, and the configuration data of the target service page is distributed according to the canary release strategy. It is understood that static configuration verification can discover structural defects in the configuration, and the execution of simulated use cases can provide feedback on potential runtime risks. Combining these two types of information to comprehensively assess risks allows for a more accurate assessment of the 4:2 release risk, which in turn determines the canary release strategy, achieving dynamic risk management of canary releases.
[0025] In step S105 of some embodiments, a gray-scale strategy is determined based on the verification results and simulation results, and the configuration data of the target service page is distributed according to the gray-scale strategy. It can be understood that static configuration verification can discover structural defects in the configuration, and simulated test case execution can provide feedback on potential runtime risks. Combining these two types of information to comprehensively assess the risk can more accurately assess the risk of the four-to-two release, and then determine the gray-scale strategy based on the risk to achieve dynamic risk management of gray-scale release.
[0026] In one feasible implementation, a release risk score is determined based on the verification results and simulation results, and the gray-scale strategy for the target service page is determined based on the release risk score. By combining the verification results obtained from the configuration dependency graph and the simulation results generated from the simulation test cases, the type of defect in the current activity page configuration and the affected user group are determined to establish the release risk score. A corresponding gray-scale strategy is generated based on the release risk score, including the initial gray-scale ratio, the rollout interval, and the stop threshold. Then, according to the determined gray-scale strategy, configuration data is distributed to the target users (group). It is understood that the release risk score can also be replaced with other forms, such as a release risk level.
[0027] In this embodiment, configuration data of the target service page of the target service to be published is obtained, configuration objects are parsed, and the configuration objects are converted into nodes. Directed edges between nodes are established based on the dependencies of the configuration objects in the configuration data to obtain a configuration dependency graph. Verification rules are executed according to the verification execution path indicated by the configuration dependency graph to obtain the verification result of the configuration data. Simulated test cases are extracted from the configuration dependency graph, executed, and their simulation results are obtained. A canary deployment strategy is determined based on the verification and simulation results, and the configuration data of the target service page is distributed according to the canary deployment strategy. This approach constructs a configuration dependency graph based on the configuration data, and then constructs an ordered verification link based on the topological relationship of the configuration dependency graph to verify the configuration data. This enables verification of business path failures. Furthermore, it combines automatically generated test cases for pre-simulation to predict page runtime anomalies in advance. Based on this, potential risks are comprehensively assessed by combining the verification and simulation results, and a canary deployment strategy is dynamically determined. This allows for flexible control of the configuration release scope based on risk conditions, effectively reducing online failures after configuration updates and improving the stability of configuration releases for mini-program activity pages.
[0028] like Figure 3 As shown, Figure 3 This is a flowchart of a canary release method for a target service provided in another embodiment of this application; the canary release method for the target service may include, but is not limited to, steps S201 to S205: S201, determine the user profile corresponding to the applicable user of the target service page; the user profile includes one profile or a combination of multiple profiles; S202, determine the business constraints corresponding to the user profile; S203, Based on the business constraints, determine the business execution path corresponding to the user profile from the configuration dependency graph; S204, Generate simulated use cases corresponding to the user profile based on the business execution path; S205, execute the simulated test case and obtain the simulation result of the simulated test case.
[0029] Existing technologies rely on manually written test cases, which makes it difficult to cover massive user scenarios formed by the intersection of multiple dimensions of conditions, and makes it difficult to detect page runtime anomalies in advance.
[0030] In step S201 of some embodiments, it is understood that the user groups targeted by the mini-program activity page have multi-dimensional differentiated characteristics. The identity attributes, business status, and usage environment of different users will directly affect the page display effect and business interaction link. In one feasible implementation, the system extracts the applicable users of the page based on the configuration data of the target service page, and divides and generates user profiles from dimensions such as city region, membership level, vehicle status, policy status, new and old customer status, client version, login status, and network environment. A single profile cannot cover the complex online scenarios in reality. Therefore, multi-dimensional profiles can be combined into composite user profiles to cover the typical user groups of the page. For example, composite profiles with multiple features such as renewal users / first-tier cities / VIP members / holding valid policies / low iOS versions, and new customers / second-tier cities / ordinary members / no vehicle policy / high Android versions / normal network.
[0031] In step S202 of some embodiments, business constraints corresponding to user profiles are determined. Since different user profiles correspond to different page rendering rules, function permissions, and business access conditions, there are business mutual exclusions and prerequisite dependencies. Business operations for any user profile must be executed within the scope of legal constraints. In one feasible implementation, for user profiles, illegal combinations are filtered using preset business constraint rules, eliminating invalid profile combinations that are logically conflicting or unreachable. For example, business constraints could include new customers not being able to use exclusive benefits for existing customers, expired coupon batches not being supported for redemption, unlogged users not entering the coupon redemption and verification process, and users without insurance policies not entering the renewal benefit process, retaining only valid user profile combinations that conform to business logic.
[0032] In step S203 of some embodiments, the business execution path corresponding to the user profile is determined from the configuration dependency graph based on business constraints. It is understood that the configuration dependency graph fully records the topological dependencies between page components, interfaces, routes, and business functions. Combined with the business constraints of the user profile, it can accurately filter out the legal and feasible complete business flow execution path under that profile.
[0033] In one feasible implementation, based on the business constraints of various valid user profiles, the effective topological paths of the configuration dependency graph are traversed, and invalid nodes and jump branches that are blocked by business constraints or whose conditions are not met are eliminated. A corresponding business execution path is then matched for each typical user profile. For example, for the VIP renewal user profile, the path is matched as "page exposure - renewal card display - coupon redemption - price quote - payment redirection"; for the ordinary new customer user profile, the path is matched as "page loading - new customer benefits display - points redemption - product browsing," achieving precise binding between user profiles and business paths.
[0034] In step S204 of some embodiments, simulated use cases corresponding to user profiles are generated based on the business execution path. Generating simulated use cases according to the business execution path corresponding to different user profiles can ensure that the use cases are highly matched with real user access scenarios.
[0035] In one feasible implementation, a two-tiered strategy combining coverage generation and risk weighting is employed to automatically generate standardized, executable simulated test cases in batches. First, the system covers basic scenarios based on user profile dimensions such as city, membership level, vehicle status, policy status, new customer status, client version, login status, and network environment, using a pairwise traversal method, for example. Specifically, it automatically cross-matches and enumerates values between any two dimensions, ensuring full coverage of the vast majority of common online user interaction scenarios, thus avoiding the limitation of traditional manual testing which can only cover a single typical scenario.
[0036] Furthermore, based on business failure history and online risk weights, high-risk profiles (dimensions) are identified in the user profile. For these high-risk dimensions, in addition to the usual pairwise combinations, scenario use cases are constructed that incorporate triple profiles with high-risk dimensions. For example, based on the pairwise combinations of user profiles, scenario use cases are generated that incorporate superimposed high-risk profiles for client version issues, weak network conditions, and impending expiration of benefits.
[0037] In step S205 of some embodiments, simulated test cases are executed to obtain the simulation results. It is understood that manual testing is insufficient to cover a vast number of usage scenarios and cannot accurately identify faults in various scenarios. Therefore, automatically generated simulated test cases are executed to simulate multiple scenarios.
[0038] In one feasible implementation, all simulated test cases are batch-scheduled to simulate the complete access and operation process of the target service page for different user profiles. The simulation engine executes the simulated test cases, simulating the continuous operations of a user after entering the target service page, such as exposure, clicking, coupon redemption, benefit redirection, and order payment. The page rendering status, interface return data, and interaction execution results are collected and summarized to obtain the simulation results. For example, for user profile P1 (renewal user / first-tier city / VIP / holding a valid policy / low version of iOS), the corresponding test case C1 is generated: open the activity page → click the renewal card → redeem the renewal coupon → enter the quotation page → initiate payment redirection. At the same time, the corresponding expected results are matched, which are that the page is available, the coupon status is redeemable, the quotation fields are complete, and the payment routing parameters are valid. The system collects the execution status of each step throughout the process. If any step fails, the system automatically records the fault node, the corresponding exception code (the expected result is not met), and the affected user profile. Finally, the execution status of all test cases is summarized to form the simulation results.
[0039] In this embodiment, by determining the user profile corresponding to the applicable users of the target service page, the business constraints corresponding to the user profile are determined. Based on the business constraints, the business execution path corresponding to the user profile is determined from the configuration dependency graph. Based on the business execution path, simulated test cases corresponding to the user profile are generated, and the simulated test cases are executed to obtain the simulation results. This method can automatically generate simulated test cases based on user profiles, covering combined scenarios under different user identities, cities, vehicle states, and client versions. By simulating the business execution paths of different users after entering the target service page, the method achieves the review of multiple business paths.
[0040] like Figure 4 As shown, Figure 4 This is a flowchart of a canary release method for a target service provided in another embodiment of this application; the canary release method for the target service may include, but is not limited to, steps S301 to S307: S301, Obtain the configuration data of the target service page of the target service to be published, and parse the configuration data to obtain the configuration object; S302, the configuration object is converted into a node, and directed edges are established between the nodes according to the dependency relationship of the configuration object in the configuration data to obtain a configuration dependency graph; For specific implementation methods of steps S301 and S302, please refer to the aforementioned embodiments in the specification, which will not be repeated here.
[0041] S303, Starting from the page entry node in the configuration dependency graph, perform a reachable traversal along the directed edges between component nodes, data source nodes, interface nodes, business resource nodes, and route nodes to obtain multiple verification execution paths; S304, Execute the verification rules on the verification execution path to obtain the verification result; In steps S303 and S304 of one embodiment, for common scenarios of the target service page, the verification execution path mainly starts from the page entry node, proceeds along component nodes, data source nodes, interface nodes, business resource nodes, and finally reaches the routing node. It can be understood that the page business chain has a fixed flow order, with the page presentation layer progressively depending on the data source, backend interface, and business resources, ultimately pointing to the redirect route. Based on the reachability traversal of the dependency graph structure, the complete sequence of the business chain can be reconstructed. Specifically, according to the sequential order of nodes on the path, matching verification rules are scheduled to check the interface connectivity status, interface field contracts, validity of product and coupon resources, legality of the routing address, activity validity period, and user segment matching rules, etc. Verification in this way can accurately locate the fault location in the chain and determine the range of downstream nodes affected by the fault, resulting in a more accurate and comprehensive verification result.
[0042] S305, if the verification result indicates that there is no blocking problem in the configuration data, then extract the simulated test case from the configuration dependency graph, execute the simulated test case and obtain the simulation result of the simulated test case; In step S305 of one embodiment, the blocking problem refers to a high-severity defect that can cause the activity page to fail to launch securely or the core transaction link to fail directly. For example, blocking problems include unreachable core interfaces, broken contracts for key fields, missing order or payment routes, invalid coupon batches without a fallback strategy, and page inaccessibility for all users due to conflicts between activity time and group rules. If the above blocking problems are not identified in the verification results, scenario-based simulation use cases are extracted based on the configuration dependency graph to simulate the complete business process of different user profiles accessing the page, and simulation results are obtained by collecting runtime data.
[0043] S306, if the verification result indicates that there is a blocking problem with the configuration data, then generate a problem list and stop the release of the configuration data; In step S306 of some embodiments, if a blocking issue flows into the canary release phase, it will directly cause a large number of users' core business operations to fail, resulting in a relatively serious online failure. In one feasible implementation, if the system identifies any blocking issue in the verification results, it immediately terminates all subsequent processes and stops the extraction of simulated test cases, simulation execution, and canary release operations. All blocking issues are summarized, and a structured issue list is generated by combining the fault nodes, affected links, and covered user groups recorded in the configuration dependency graph. This list is then pushed to the operations and development side for configuration rectification.
[0044] S307, determine the grayscale strategy based on the verification result and the simulation result, and distribute the configuration data of the target service page according to the grayscale strategy.
[0045] For details on the specific implementation of step S307, please refer to the aforementioned embodiments in the specification, which will not be repeated here.
[0046] In this embodiment, an ordered verification execution path is formed by performing reachability traversal along directed edges based on the configuration dependency graph. When an anomaly occurs, the link node corresponding to the fault can be directly located. At the same time, downstream affected nodes are identified based on the graph topology to determine the scope of the fault's impact. Furthermore, by identifying blocking issues and preventing the distribution of page configurations with blocking issues, it is possible to avoid significant defects from entering the online gray-scale environment.
[0047] like Figure 5 As shown, Figure 5 This is a flowchart of a canary release method for a target service provided in another embodiment of this application; the canary release method for the target service may include, but is not limited to, steps S401 to S406: S401, Generate a release risk score based on the verification results, the simulation results, the interface risk data, the number of critical business paths, and the estimated access volume; Understandably, traditional canary deployment strategies rely solely on manual experience to fix configuration parameters for increased deployment, failing to comprehensively assess deployment risks by considering current configuration flaws, business chain complexity, and traffic scale. In step S401 of some embodiments, a multi-dimensional scoring reference factor is used to quantify configuration release risks. Specifically, the verification results include the number of blocking issues, the number of warning issues, simulation results including the number of scenario anomalies exposed by simulated use cases, interface risk data including the dependency level and impact scope of interfaces, the number of critical business paths (the number of critical business paths carried by the page), and the estimated access volume (the estimated access traffic scale of the activity page). Optionally, the risk scoring reference factors may also include historical deployment anomaly records for this type of page.
[0048] Optionally, when generating the score, a weighted calculation method can be used to assign weighted values to high-weight factors such as blocking issues, payment-related critical path anomalies, and large estimated traffic, and to give light weight to minor warning issues. Finally, a normalized release risk score between zero and one hundred is generated. The higher the score, the higher the risk of going live in the current configuration and the worse the stability of the page operation.
[0049] S402, determine the gray-scale strategy of the target service page based on the release risk score, and distribute the configuration data of the target service page according to the gray-scale strategy; In step S402 of some embodiments, the gray-scale strategy for the target service page is determined based on the release risk score, and the configuration data for the target service page is distributed according to the gray-scale strategy. It is understood that the quantified risk score can intuitively represent the overall risk level of the configuration, and by configuring gray-scale release parameters differently based on the score gradient, adaptive release rhythm control can be achieved for high-risk slow releases and low-risk fast releases.
[0050] Optionally, in one embodiment, the grayscale strategy includes an initial grayscale ratio, observation duration, and incremental step size. Determining the grayscale strategy for the target service page based on the release risk score includes the following steps S4021-S4023: S4021 If the release risk score is greater than the first threshold, then the initial grayscale ratio is determined to be the first ratio, the observation period is the first duration, and the volume increase step size is the first step size. In one feasible implementation, a release strategy is determined based on the relationship between the release risk score and a first threshold. When the release risk score exceeds the first threshold, the current configuration is deemed a high-risk release, posing potential scenario compatibility issues and link stability risks. The first threshold, first ratio, first duration, and the first step size can be set according to actual needs. For example, the first threshold is set to 80. When the release risk score exceeds 80, a conservative rollout strategy is adopted, configuring an initial grayscale ratio of 1%, an observation period of 30 minutes, and a single rollout step size of 1%. By using smaller traffic volumes, longer observation periods, and incremental increases, the online impact of the high-risk configuration is controlled.
[0051] S4022, if the release risk score is less than the first threshold and greater than or equal to the second threshold, then the initial grayscale ratio is determined to be the second ratio, the observation period is determined to be the second duration, and the volume increase step size is determined to be the second step size; In one feasible implementation, when the risk release score is less than a first threshold, the current configuration is determined to be a medium-risk release, without fatal blocking issues but with potential compatibility risks in some scenarios. The initial grayscale ratio is then determined to be a second ratio, the observation period to be a second duration, and the scaling-up step size to be a second step size. Specifically, the second ratio is greater than the first ratio, the second duration is less than the first duration, and the second step size is greater than the first step size.
[0052] For example, a second threshold of 60 is set. When the risk score is between 60 and 80, a moderate rollout strategy is adopted, with an initial grayscale ratio of 5%, an observation period of 20 minutes, and a single rollout step size of 5%. Compared to high-risk scenarios, this strategy appropriately increases the initial user coverage, shortens the observation window, and accelerates the rollout pace, improving release efficiency while ensuring security.
[0053] S4023, if the release risk score is less than the second threshold, then the initial grayscale ratio is determined to be the third ratio, the observation period is the third duration, and the volume increase step size is the third step size.
[0054] In one feasible implementation, when the release risk score is less than a second threshold, the current configuration is determined to be a low-risk release, indicating good pre-verification and scenario simulation results, and reliable overall stability. The initial grayscale ratio is then set to a third ratio, the observation period to a third duration, and the ramp-up step size to a third step size. Specifically, the third ratio is greater than the second ratio, the third duration is less than the second duration, and the third step size is greater than the second step size. For example, when the release risk score is less than 60, an efficient ramp-up strategy is adopted, configuring an initial grayscale ratio of 10%, an observation period of ten minutes, and a single ramp-up step size of 10%. This strategy adapts to stable and reliable configuration versions, completing the full release with a larger initial traffic, a shorter observation period, and a faster ramp-up pace, balancing release security and iteration efficiency.
[0055] S403, Collect the operational business metrics of the target service page; In step S403 of some embodiments, operational business metrics of the target service page are collected. It is understood that during the gray-scale rollout process, there are real-time operational risks that cannot be covered by pre-verification, necessitating continuous monitoring of the page's operational status using real online metrics to achieve dynamic risk control during the release process. In one feasible implementation, the system collects all operational metrics of the mini-program activity page in real time after its launch, covering both technical and business metrics. Specifically, these include page rendering success rate, component loading anomaly rate, interface request success rate, interface response time, coupon redemption success rate, price quotation success rate, payment redirection success rate, user churn rate, and the number of abnormal errors, providing data support for subsequent anomaly detection and strategy adjustments.
[0056] S404, determine whether the operational business indicators exceed the abnormal threshold; In step S404 of some embodiments, by setting a standardized anomaly threshold, normal fluctuations and fault anomalies can be accurately distinguished, avoiding misjudgment that leads to suspension of releases or blind expansion. In one feasible implementation, real-time collected operational business indicators are compared with preset anomaly thresholds to determine whether each indicator exceeds the safe fluctuation range.
[0057] S405, if the abnormal threshold is exceeded, the gray-scale release of the target service page is suspended and the corresponding handling strategy is executed according to the abnormal type of the operational business indicator. In step S405 of some embodiments, if the abnormal threshold is exceeded, it indicates that there is an online operation failure in the current configuration, and the risk spread must be blocked immediately and targeted measures must be taken.
[0058] In one feasible implementation, when any operational business metric exceeds a preset abnormal threshold, the current gray-scale rollout process is immediately paused, and the scope of configuration distribution is stopped. Simultaneously, corresponding handling strategies are matched based on the specific type of the abnormal metric. For example, when page open success rate, interface error rate, or payment redirection success rate is abnormal, the rollout is automatically paused and reverted to the previous stable configuration, which can shorten fault recovery time and reduce losses in product exposure, add-to-cart, and payment conversion.
[0059] The exception types include both indicator-level exceptions and their corresponding root cause categories, which can include: page availability exceptions (decreased success rate of opening), rendering exceptions (component template or field mapping failure), interface exceptions (timeout / rate limiting), interaction exceptions (key button clicks are unresponsive), transaction exceptions (order placement or payment redirection failure), resource exceptions (products are removed from the shelves or coupons are invalid), etc.
[0060] For example, when page availability is abnormal, a lightweight fallback page should be switched first and high-risk components should be hidden; when rendering is abnormal, circuit breaking should be applied at the component level and the previous field-compatible version should be rolled back; when interface is abnormal, a backup interface should be switched or cache snapshots should be enabled and request concurrency should be reduced; when interaction is abnormal, the problem button should be temporarily taken offline while the main transaction path should be retained; when transaction is abnormal, the volume release should be frozen immediately, the activity entry should be closed, and the previous stable configuration should be rolled back; when resources are abnormal, associated cards should be automatically removed and operations should be notified to replenish resources.
[0061] S406, if the abnormal threshold is not exceeded and the threshold is less than the set threshold, the gray scale ratio in the gray scale strategy is increased to obtain an updated gray scale strategy, and the configuration data of the target service page is distributed according to the updated gray scale strategy until full release is achieved.
[0062] In step S406 of some embodiments, if the abnormal threshold is not exceeded and the threshold is less than the set threshold, the grayscale ratio in the grayscale strategy is increased to obtain an updated grayscale strategy. The configuration data of the target service page is then distributed according to the updated grayscale strategy until full release is achieved. It can be understood that, under the premise that the indicators are continuously stable and there are no abnormal fluctuations, the full version coverage can be smoothly achieved by gradually increasing the volume according to the predetermined step size. In one feasible implementation, when the real-time running business indicators are stable within the safe threshold range and there are no new abnormal errors or business failures, the system gradually increases the grayscale ratio according to the volume increase step size corresponding to the current risk level, iteratively updates the grayscale strategy and continuously distributes configuration data; the indicator monitoring and volume increase process is executed cyclically until the grayscale traffic covers all users, and the full release of the target service page configuration is completed.
[0063] Optionally, after the full release is completed, the configuration version, verification results, grayscale indicators, and rollback reasons used for each target service page update can be saved to provide a basis for reusing configuration templates, optimizing release rules, and assessing activity risks for subsequent target services.
[0064] In this embodiment, after the configuration data of the target service page is distributed according to the canary release strategy, the operational business metrics of the target service page are collected to determine whether the operational business metrics exceed the abnormal threshold. If the abnormal threshold is exceeded, the canary release of the target service page is suspended, and the corresponding handling strategy is executed according to the abnormal type of the operational business metrics. If the abnormal threshold is not exceeded and is less than the set threshold, the canary ratio in the canary release strategy is increased to obtain an updated canary release strategy. The configuration data of the target service page is distributed according to the updated canary release strategy until full release is achieved. Using this method, page technical metrics and core business metrics are continuously collected during the canary release process, the online operating status is monitored in real time, and various online faults are accurately identified through preset abnormal thresholds. When an abnormality occurs, automatic degradation or rollback is performed, significantly reducing the probability of online faults in the configuration update of the mini-program activity page and improving the stability and intelligence level of the canary release of configurable pages.
[0065] Next, please see Figure 6 , Figure 6 This is a schematic diagram of the overall process of a canary release method for a target service provided in an embodiment of this application, based on... Figure 6 This application describes the canary release method for the target service provided in this embodiment. First, the configuration data of the target service page is obtained and parsed to obtain standardized configuration objects. A configuration dependency graph is constructed based on the association relationships of the configuration objects. Multiple verification execution paths are obtained based on the configuration dependency graph, and path verification is completed. Based on the verification results, it is determined whether there are any blocking issues. If a blocking issue exists, the release is directly stopped. If no blocking issue exists, scenario-based simulation test cases are extracted and executed to obtain the corresponding simulation results. Based on this, a release risk score is calculated and generated by combining the verification results, simulation results, interface risk data, the number of critical business paths, and estimated access volume. According to different gradients of the release risk score, a differentiated hierarchical canary release strategy is adaptively matched to determine the corresponding initial canary release ratio, observation period, and volume increase step size. Page configuration data is then distributed according to the canary release strategy. During the canary release process, the operational business indicators of the page are continuously collected to determine whether the indicators exceed preset abnormal thresholds. If the indicators are abnormal, the canary release is paused and the corresponding handling strategy is executed. If the indicators are stable and normal, the canary release ratio is gradually increased, the canary release strategy is iteratively updated, and configuration data is continuously distributed until full release is completed.
[0066] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.
[0067] In one embodiment, a gray-scale deployment apparatus for a target service is provided, which corresponds one-to-one with the gray-scale deployment method for the target service described in the above embodiments. For example... Figure 7As shown, the canary deployment device 700 for the target service includes a configuration module 10, a graph construction module 20, a verification module 30, a test case testing module 40, and a canary execution module 50. Detailed descriptions of each functional module are as follows: Configuration module 10 is used to obtain configuration data of the target service page of the target service to be published, and parse the configuration data to obtain a configuration object; Graph construction module 20 is used to convert the configuration object into nodes and establish directed edges between the nodes according to the dependency relationship of the configuration object in the configuration data to obtain a configuration dependency graph; Verification module 30 is used to execute verification rules according to the verification execution path indicated by the configuration dependency graph and obtain verification results; The test case module 40 is used to extract simulated test cases from the configuration dependency graph, execute the simulated test cases, and obtain the simulation results of the simulated test cases; The grayscale execution module 50 is used to determine the grayscale strategy based on the verification result and the simulation result, and to distribute the configuration data of the target service page according to the grayscale strategy.
[0068] In one embodiment, the test case module 40 is specifically used for: Determine the user profile corresponding to the applicable users of the target service page; the user profile includes one profile or a combination of multiple profiles; Determine the business constraints corresponding to the user profile; Based on the business constraints, the business execution path corresponding to the user profile is determined from the configuration dependency graph; Based on the business execution path, generate simulated use cases corresponding to the user profile; Execute the simulated test case to obtain the simulation results.
[0069] In one embodiment, the verification module 30 is specifically used for: Starting from the page entry node in the configuration dependency graph, a reachability traversal is performed along the directed edges between component nodes, data source nodes, interface nodes, business resource nodes, and route nodes to obtain multiple verification execution paths; The verification rules are executed on the verification execution path to obtain the verification result.
[0070] In one embodiment, the grayscale execution module 50 is specifically used for: A release risk score is generated based on the verification results, simulation results, interface risk data, number of critical business paths, and estimated access volume. The gray-scale strategy for the target service page is determined based on the published risk score, and the configuration data for the target service page is distributed according to the gray-scale strategy.
[0071] In one embodiment, the grayscale execution module 50 is specifically used for: If the release risk score is greater than the first threshold, then the initial grayscale ratio is determined to be the first ratio, the observation period is the first duration, and the volume increase step size is the first step size. If the release risk score is less than the first threshold and greater than or equal to the second threshold, then the initial grayscale ratio is determined to be the second ratio, the observation period is determined to be the second duration, and the volume increase step size is determined to be the second step size; the second ratio is greater than the first ratio, the second duration is less than the first duration, and the second step size is greater than the first step size. If the release risk score is less than the second threshold, then the initial grayscale ratio is determined to be the third ratio, the observation period is determined to be the third duration, and the volume increase step size is determined to be the third step size; the third ratio is greater than the second ratio, the third duration is less than the second duration, and the third step size is greater than the second step size.
[0072] In one embodiment, the grayscale execution module 50 is further configured to: Collect operational business metrics of the target service page; Determine whether the operational business metrics exceed the abnormal threshold; If the abnormal threshold is exceeded, the gray-scale release of the target service page will be suspended and the corresponding handling strategy will be executed according to the abnormality type of the operational business indicator. If the abnormal threshold is not exceeded and the threshold is less than the set threshold, the gray scale ratio in the gray scale strategy is increased to obtain an updated gray scale strategy. The configuration data of the target service page is then distributed according to the updated gray scale strategy until full release is achieved.
[0073] In one embodiment, the test case module 40 is further configured to: If the verification result indicates that there is no blocking problem with the configuration data, then a simulated test case is extracted from the configuration dependency graph, the simulated test case is executed, and the simulation result of the simulated test case is obtained; If the verification result indicates that there is a blocking problem with the configuration data, a problem list is generated and the release of the configuration data is stopped.
[0074] This application provides a canary release device for a target service. By automatically collecting software data from the target software and inputting it into a pre-trained model, it can quickly identify target risk defects whose risk assessment indicators exceed preset thresholds, effectively improving defect identification efficiency and accuracy. Then, using the selected target risk defects as target nodes, feature data is extracted from the software data to construct related nodes and build a causal dependency graph. Based on a graph model causal inference algorithm combined with the causal dependency graph, related nodes are filtered to identify highly correlated candidate root causes of defects. This causal dependency graph is only built for identified high-risk defects, eliminating the need for modeling based on the full amount of software data, significantly reducing the data processing volume for graph construction. On this basis, the associated data corresponding to the candidate root causes is retrieved to complete confidence verification, further ensuring the accuracy and reliability of the final root cause location result, and improving the overall efficiency of software daily operation and maintenance and defect management.
[0075] Specific limitations regarding the canary deployment apparatus for the target service can be found in the limitations of the canary deployment method for the target service described above, and will not be repeated here. Each module in the aforementioned canary deployment apparatus for the target service can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.
[0076] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 8 As shown. The computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used to communicate with external clients via a network connection. When the computer program is executed by the processor, it implements the functions or steps of a canary release method for a target service on the server side.
[0077] In one embodiment, a computer device is provided, which may be a client, and its internal structure diagram may be as follows: Figure 9As shown, the computer device includes a processor, memory, network interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used to communicate with an external server via a network connection. When the computer program is executed by the processor, it implements client-side functions or steps of a canary release method for a target service.
[0078] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps: Obtain the configuration data of the target service page of the target service to be published, and parse the configuration data to obtain the configuration object; The configuration object is converted into a node, and directed edges are established between the nodes based on the dependencies of the configuration object in the configuration data to obtain a configuration dependency graph; The verification rules are executed according to the verification execution path indicated by the configuration dependency graph to obtain the verification result of the configuration data; Extract simulated use cases from the configuration dependency graph, execute the simulated use cases, and obtain the simulation results of the simulated use cases; Based on the verification results and the simulation results, a grayscale strategy is determined, and the configuration data of the target service page is distributed according to the grayscale strategy.
[0079] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor: Obtain the configuration data of the target service page of the target service to be published, and parse the configuration data to obtain the configuration object; The configuration object is converted into a node, and directed edges are established between the nodes based on the dependencies of the configuration object in the configuration data to obtain a configuration dependency graph; The verification rules are executed according to the verification execution path indicated by the configuration dependency graph to obtain the verification result of the configuration data; Extract simulated use cases from the configuration dependency graph, execute the simulated use cases, and obtain the simulation results of the simulated use cases; Based on the verification results and the simulation results, a grayscale strategy is determined, and the configuration data of the target service page is distributed according to the grayscale strategy.
[0080] It should be noted that the functions or steps that can be implemented by the computer-readable storage medium or computer device described above can be referred to the relevant descriptions on the server side and client side in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.
[0081] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.
[0082] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.
[0083] It should be noted that any software tools or components not belonging to this company appearing in the embodiments of this application are merely illustrative examples and do not represent actual use. All user personal information involved in the embodiments of this application has been authorized (with knowledge and consent) by the relevant parties or has been fully authorized by all parties, and the executing entity may obtain it through various legal and compliant means. The collection, storage, use, processing, transmission, provision, and disclosure of the information, data, and signals involved all comply with relevant laws and regulations and do not violate public order and good morals.
[0084] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.
Claims
1. A method for canary release of a target service, characterized in that, include: Obtain the configuration data of the target service page of the target service to be published, and parse the configuration data to obtain the configuration object; The configuration object is converted into a node, and directed edges are established between the nodes based on the dependencies of the configuration object in the configuration data to obtain a configuration dependency graph; The verification rules are executed according to the verification execution path indicated by the configuration dependency graph to obtain the verification result of the configuration data; Extract simulated use cases from the configuration dependency graph, execute the simulated use cases, and obtain the simulation results of the simulated use cases; Based on the verification results and the simulation results, a grayscale strategy is determined, and the configuration data of the target service page is distributed according to the grayscale strategy.
2. The canary release method for the target service as described in claim 1, characterized in that, The step of extracting simulated use cases from the configuration dependency graph, executing the simulated use cases, and obtaining the simulation results of the simulated use cases includes: Determine the user profile corresponding to the applicable users of the target service page; the user profile includes one profile or a combination of multiple profiles; Determine the business constraints corresponding to the user profile; Based on the business constraints, the business execution path corresponding to the user profile is determined from the configuration dependency graph; Based on the business execution path, generate simulated use cases corresponding to the user profile; Execute the simulated test case to obtain the simulation results.
3. The canary release method for the target service as described in claim 1, characterized in that, The step of executing the verification rules according to the verification execution path indicated by the configuration dependency graph to obtain the verification result includes: Starting from the page entry node in the configuration dependency graph, a reachability traversal is performed along the directed edges between component nodes, data source nodes, interface nodes, business resource nodes, and route nodes to obtain multiple verification execution paths; The verification rules are executed on the verification execution path to obtain the verification result.
4. The canary release method for the target service as described in claim 1, characterized in that, The step of determining a grayscale strategy based on the verification result and the simulation result, and issuing the configuration data of the target service page according to the grayscale strategy, includes: A release risk score is generated based on the verification results, simulation results, interface risk data, number of critical business paths, and estimated access volume. The gray-scale strategy for the target service page is determined based on the published risk score, and the configuration data for the target service page is distributed according to the gray-scale strategy.
5. The canary release method for the target service as described in claim 4, characterized in that, The grayscale strategy includes an initial grayscale ratio, observation period, and incremental step size; determining the grayscale strategy for the target service page based on the release risk score includes: If the release risk score is greater than the first threshold, then the initial grayscale ratio is determined to be the first ratio, the observation period is the first duration, and the volume increase step size is the first step size. If the release risk score is less than the first threshold and greater than or equal to the second threshold, then the initial grayscale ratio is determined to be the second ratio, the observation period is determined to be the second duration, and the volume increase step size is determined to be the second step size; the second ratio is greater than the first ratio, the second duration is less than the first duration, and the second step size is greater than the first step size. If the release risk score is less than the second threshold, then the initial grayscale ratio is determined to be the third ratio, the observation period is determined to be the third duration, and the volume increase step size is determined to be the third step size; the third ratio is greater than the second ratio, the third duration is less than the second duration, and the third step size is greater than the second step size.
6. The canary release method for the target service as described in claim 1, characterized in that, After determining the grayscale strategy based on the verification result and the simulation result, and distributing the configuration data of the target service page according to the grayscale strategy, the method further includes: Collect operational business metrics of the target service page; Determine whether the operational business metrics exceed the abnormal threshold; If the abnormal threshold is exceeded, the gray-scale release of the target service page will be suspended and the corresponding handling strategy will be executed according to the abnormality type of the operational business indicator. If the abnormal threshold is not exceeded and the threshold is less than the set threshold, the gray scale ratio in the gray scale strategy is increased to obtain an updated gray scale strategy. The configuration data of the target service page is then distributed according to the updated gray scale strategy until full release is achieved.
7. The canary release method for the target service as described in claim 1, characterized in that, The step of extracting simulated use cases from the configuration dependency graph, executing the simulated use cases, and obtaining the simulation results of the simulated use cases includes: If the verification result indicates that there is no blocking problem with the configuration data, then a simulated test case is extracted from the configuration dependency graph, the simulated test case is executed, and the simulation result of the simulated test case is obtained; If the verification result indicates that there is a blocking problem with the configuration data, a problem list is generated and the release of the configuration data is stopped.
8. A grayscale publishing device for a target service, characterized in that, include: The configuration module is used to obtain the configuration data of the target service page of the target service to be published, and parse the configuration data to obtain the configuration object; The graph construction module is used to convert the configuration object into nodes and establish directed edges between the nodes according to the dependency relationship of the configuration object in the configuration data to obtain a configuration dependency graph. The verification module is used to execute verification rules according to the verification execution path indicated by the configuration dependency graph and obtain verification results; The test case module is used to extract simulated test cases from the configuration dependency graph, execute the simulated test cases, and obtain the simulation results of the simulated test cases; The grayscale execution module is used to determine the grayscale strategy based on the verification results and the simulation results, and to distribute the configuration data of the target service page according to the grayscale strategy.
9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the canary release method for the target service as described in any one of claims 1 to 7.
10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the canary release method for the target service as described in any one of claims 1 to 7.
