Full-link pressure testing method and system based on artificial intelligence

Abstract

The invention provides a full-link pressure test method and system based on artificial intelligence. The method comprises the following steps: determining a data type according to a test target defined by a user; generating simulation traffic according to the data type and a business type traffic proportion obtained based on a historical access log; generating pressure measurement execution information according to the simulated flow and a preset fuzzy control model; when the pressure measurement execution information is executed, obtaining a performance index corresponding to the pressure measurement execution information in an execution process; generating node topology information according to the performance index, the voltage measurement execution information and a preset hierarchical fusing condition; and determining a full-link voltage measurement scheme according to the node topology information and a preset voltage measurement index. A test environment consistent with the production environment is constructed, and it is ensured that a test result truly reflects the performance of the production environment; large-scale test data of a real scene is generated, a real business scene is simulated, and the reliability of a test result is improved.

Description

Technical Field

[0001] This application relates to the field of software testing, and particularly to an end-to-end stress testing method and system based on artificial intelligence. Background Art

[0002] End-to-end stress testing is an important means to verify the overall performance and stability of a system under high-concurrency and high-load scenarios. In the prior art, end-to-end stress testing is usually carried out by building a test environment, generating test data, and performing stress testing. However, there are many deficiencies in the prior art, making it difficult to truly reflect the performance of the system in the production environment.

[0003] First, there are differences between the test environment and the production environment, such as inconsistent hardware resources, network topologies, middleware versions, etc., resulting in test results that cannot truly reflect the performance of the production environment. Second, there is a lack of large-scale test data in real scenarios, including data volume, data diversity, data access patterns, etc., leading to distorted test results. In addition, the prior art often only focuses on the core business link and ignores edge scenarios or dependent systems, such as third-party interfaces, message queues, caches, etc., presenting potential risks. On the other hand, resource monitoring is not comprehensive or the analysis ability is insufficient, making it difficult to accurately locate the performance bottleneck position. Only surface indicators such as CPU and memory are concerned, lacking in-depth exploration of the root causes, such as slow SQL, cache breakdown, lock competition, or downstream service traffic limiting. Finally, end-to-end stress testing has high requirements for hardware and human costs, insufficient automation, and low efficiency in repeated testing.

[0004] In summary, there are many problems in the existing end-to-end stress testing technology, such as inconsistent test environment and production environment, insufficient data preparation, incomplete link coverage, misjudgment of performance bottlenecks, lack of in-depth analysis of test results, resource and cost limitations, etc. There is an urgent need for a new end-to-end stress testing method to solve the above problems. Summary of the Invention

[0005] In view of the above problems, this application is proposed to provide an end-to-end stress testing method and system based on artificial intelligence that overcomes or at least partially solves the above problems, including: An end-to-end stress testing method based on artificial intelligence, the method includes: Determine the data type according to the test target defined by the user, where the data type includes stress test data and / or real data; Generate simulated traffic according to the data type and the traffic ratio of the business type obtained from the historical access logs; Generate stress test execution information according to the simulated traffic and a preset fuzzy control model; When the stress test execution information is in execution, obtain the performance indicators corresponding to the stress test execution information during the execution process; Generate node topology information based on the performance metrics, the stress test execution information, and the preset hierarchical fusing conditions; Determine the full-link stress test plan based on the node topology information and the preset stress test metrics.

[0006] Further, the step of determining the data type based on the user-defined test target, where the data type includes stress test data and / or real data, includes: Generate a stress test task based on the test target; Determine the target link nodes based on the stress test task; Determine parameter information based on the stress test task and the target link nodes; Determine the data type based on the parameter information and the preset mirror production library, where the data type includes stress test data and / or real data.

[0007] Further, the step of generating simulated traffic based on the data type and the traffic ratio of the service type obtained from the historical access logs includes: Determine the target conditions based on the data type; Determine the stress test rules based on the target conditions and the data type; Perform data desensitization on the real data according to the stress test rules to generate a stress test data set; Generate simulated traffic based on the stress test data set and the traffic ratio of the service type obtained from the historical access logs.

[0008] Further, the step of generating stress test execution information based on the simulated traffic and the preset fuzzy control model includes: Determine the respective traffic weight values based on the simulated traffic and the preset fuzzy control model; Generate a weight label list based on the respective traffic weight values; Generate full-link nodes based on the weight label list and each distributed node; Generate stress test execution information by covering within the full-link nodes according to the simulated traffic.

[0009] Further, the step of generating node topology information based on the performance metrics, the stress test execution information, and the preset hierarchical fusing conditions includes: Determine the processing mechanism based on the performance metrics and the preset hierarchical fusing conditions; Generate node topology information based on the processing mechanism and the stress test execution information; Generate a stress test result based on the node topology information and the performance metrics.

[0010] Further, the step of determining a processing mechanism according to the performance index and the preset hierarchical fusing condition, where the preset hierarchical fusing condition includes a primary fusing condition and / or a secondary fusing condition, includes: Determine the index attribute of the performance index, where the index attribute includes a call success rate or a latency; When the index attribute is the call success rate and the call success rate is less than a first preset call threshold, and / or when the index attribute is the latency and the latency is greater than a first preset latency threshold, then determine that the current preset hierarchical fusing condition is the primary fusing condition, and determine that reducing the current stress test traffic to a preset percentage range of the original traffic is the processing mechanism; and / or, When the index attribute is the call success rate and the call success rate is less than a second preset call threshold, where the second preset call threshold is less than the first preset call threshold, and / or when the index attribute is the latency and the latency is greater than a second preset latency threshold, where the second preset latency threshold is greater than the first preset latency threshold, then determine that the current preset hierarchical fusing condition is the secondary fusing condition, and determine that isolating between problem nodes is the processing mechanism.

[0011] Further, the step of determining a full-link stress test plan according to the node topology information and the preset stress test index includes: Generate a stress test result according to the node topology information and the performance index; Generate an optimization plan according to the stress test result and the preset stress test index; Determine a full-link stress test plan according to the stress test result and the optimization plan.

[0012] An embodiment of the present application also discloses an artificial intelligence-based full-link stress test system, and the system includes: A first determination module, configured to determine a data type according to a test target defined by a user, where the data type includes stress test data and / or real data; A first generation module, configured to generate simulated traffic according to the data type and the traffic ratio of the service type obtained from the historical access log; A second generation module, configured to generate stress test execution information according to the simulated traffic and a preset fuzzy control model; An acquisition module, configured to, when the stress test execution information is being executed, acquire a performance index corresponding to the stress test execution information during the execution process; A third generation module, configured to generate node topology information according to the performance index, the stress test execution information, and a preset hierarchical fusing condition; A second determination module, configured to determine a full-link stress test plan according to the node topology information and a preset stress test index.

[0013] An embodiment of the present application also discloses a computer device, including a processor, a memory, and a computer program stored on the memory and capable of running on the processor. When the computer program is executed by the processor, the steps of a full-link stress test method based on artificial intelligence as described above are implemented.

[0014] An embodiment of the present application also discloses a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of a full-link stress test method based on artificial intelligence as described above are implemented.

[0015] The present application has the following advantages: In the embodiments of the present application, compared with the "existing full-link stress testing technology in the prior art, which has problems such as inconsistent test environment and production environment, insufficient data preparation, incomplete link coverage, misjudgment of performance bottlenecks, insufficient analysis of test results, and resource and cost limitations", the present application provides a solution for solving full-link stress testing based on artificial intelligence, specifically: "A full-link stress testing method based on artificial intelligence, the method includes: determining the data type according to the test target defined by the user, wherein the data type includes stress testing data and / or real data; generating simulated traffic according to the data type and the traffic ratio of business types obtained based on historical access logs; generating stress testing execution information according to the simulated traffic and a preset fuzzy control model; when the stress testing execution information is in execution, obtaining the performance indicators corresponding to the stress testing execution information during the execution process; generating node topology information according to the performance indicators, the stress testing execution information and a preset hierarchical fusing condition; determining a full-link stress testing plan according to the node topology information and a preset stress testing index". By "generating simulated traffic according to the data type and the traffic ratio of business types obtained based on historical access logs; generating stress testing execution information according to the simulated traffic and a preset fuzzy control model; when the stress testing execution information is in execution, obtaining the performance indicators corresponding to the stress testing execution information during the execution process; generating node topology information according to the performance indicators, the stress testing execution information and a preset hierarchical fusing condition; determining a full-link stress testing plan according to the node topology information and a preset stress testing index", the problems of "existing full-link stress testing technology having inconsistent test environment and production environment, insufficient data preparation, incomplete link coverage, misjudgment of performance bottlenecks, insufficient analysis of test results, and resource and cost limitations" are solved, and the effects of "constructing a test environment consistent with the production environment to ensure that the test results truly reflect the performance of the production environment; generating large-scale test data for real scenarios, simulating real business scenarios, and improving the reliability of test results; covering core business links, edge scenarios and dependent systems, discovering potential risk points, and improving the robustness of the system; comprehensively monitoring resource indicators, accurately positioning the location of performance bottlenecks; combining logs and link tracing to locate the root cause of performance bottlenecks, deeply analyzing and solving performance problems; automating the test process, reducing labor costs, improving test efficiency, and supporting regular stress testing" are achieved. Description of the Drawings

[0016] In order to more clearly illustrate the technical solutions of the present application, the drawings required for the description of the present application will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0017] Figure 1It is a flowchart of the steps of a full - link stress testing method based on artificial intelligence provided by an embodiment of the present application; Figure 2 It is a block diagram of the structure of a full - link stress testing system based on artificial intelligence provided by an embodiment of the present application; Figure 3 It is a schematic diagram of the structure of a computer device provided by an embodiment of the present invention. Detailed implementation manners

[0018] To make the objectives, features, and advantages of the present application more obvious and understandable, the present application will be further described in detail below with reference to the accompanying drawings and specific implementation manners. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts shall fall within the protection scope of the present application.

[0019] The inventors found through analyzing the prior art that: regarding solving problems such as the inconsistency between the test environment and the production environment in full - link stress testing, the lack of data volume or data diversity in real - world scenarios, the neglect of edge scenarios or dependent systems, the incomplete resource monitoring or insufficient analysis capabilities, the lack of in - depth exploration of root causes by only focusing on surface indicators, and the high hardware and labor costs with insufficient automation in full - link stress testing, there are already some prior arts. Usually, the stress testing process is carried out in an automated form, saving the time for stress testing preparation, execution, report sorting, and analysis, and supporting regular stress testing; however, there are still problems such as insufficient analysis and summary of different application scenarios, the inability to further optimize the stress testing process and rules, improve the test efficiency and accuracy, as well as the lack of an intelligent adjustment mechanism, the lack of multi - dimensional analysis indicators, and the lack of a visualization interface.

[0020] In summary, the prior art has the following deficiencies: 1. There are differences between the test environment and the production environment, such as inconsistent hardware resources, network topologies, middleware versions, etc., resulting in the test results being unable to truly reflect the performance of the production environment.

[0021] 2. There is a lack of large - scale test data in real - world scenarios, including data volume, data diversity, data access patterns, etc., resulting in distorted test results.

[0022] 3. Edge scenarios or dependent systems are neglected, such as third - party interfaces, message queues, caches, etc., posing potential risks.

[0023] 4. Resource monitoring is incomplete or the analysis capabilities are insufficient, making it difficult to accurately locate the performance bottleneck positions, such as slow SQL, cache breakdown, lock competition, or downstream service rate limiting and other problems.

[0024] 5. Only focusing on surface indicators and lacking in-depth exploration of root causes, it is impossible to effectively analyze and solve performance problems.

[0025] Refer to Figure 1 , which shows a flowchart of the steps of a full-link stress testing method based on artificial intelligence provided by an embodiment of the present application; A full-link stress testing method based on artificial intelligence, the method comprising: S110. Determine the data type according to the test target defined by the user, wherein the data type includes stress test data and / or real data; S120. Generate simulated traffic according to the data type and the traffic ratio of the service type obtained based on the historical access log; S130. Generate stress test execution information according to the simulated traffic and a preset fuzzy control model; S140. When the stress test execution information is being executed, obtain the performance indicators corresponding to the stress test execution information during the execution process; S150. Generate node topology information according to the performance indicators, the stress test execution information, and a preset hierarchical fusing condition; S160. Determine a full-link stress test plan according to the node topology information and a preset stress test index.

[0026] In the embodiments of the present application, compared with the "existing full-link stress testing technology in the prior art has problems such as inconsistent test environment and production environment, insufficient data preparation, incomplete link coverage, misjudgment of performance bottlenecks, insufficient analysis of test results, and resource and cost limitations", the present application provides a solution for full-link stress testing based on artificial intelligence, specifically: "A full-link stress testing method based on artificial intelligence, the method includes: determining the data type according to the test target defined by the user, wherein the data type includes stress testing data and / or real data; generating simulated traffic according to the data type and the traffic ratio of business types obtained based on historical access logs; generating stress testing execution information according to the simulated traffic and a preset fuzzy control model; when the stress testing execution information is in execution, obtaining the performance indicators corresponding to the stress testing execution information during the execution process; generating node topology information according to the performance indicators, the stress testing execution information and a preset hierarchical fusing condition; determining the full-link stress testing plan according to the node topology information and a preset stress testing index". By "generating simulated traffic according to the data type and the traffic ratio of business types obtained based on historical access logs; generating stress testing execution information according to the simulated traffic and a preset fuzzy control model; when the stress testing execution information is in execution, obtaining the performance indicators corresponding to the stress testing execution information during the execution process; generating node topology information according to the performance indicators, the stress testing execution information and a preset hierarchical fusing condition; determining the full-link stress testing plan according to the node topology information and a preset stress testing index", the problems of "existing full-link stress testing technology has problems such as inconsistent test environment and production environment, insufficient data preparation, incomplete link coverage, misjudgment of performance bottlenecks, insufficient analysis of test results, and resource and cost limitations" are solved, and the effects of "constructing a test environment consistent with the production environment to ensure that the test results truly reflect the performance of the production environment; generating large-scale test data for real scenarios, simulating real business scenarios, and improving the reliability of test results; covering core business links, edge scenarios and dependent systems, discovering potential risk points, and improving the robustness of the system; comprehensively monitoring resource indicators and accurately positioning the location of performance bottlenecks; combining logs and link tracing to locate the root cause of performance bottlenecks, deeply analyzing and solving performance problems; automating the test process, reducing labor costs, improving test efficiency, and supporting regular stress testing" are achieved.

[0027] Next, a full-link stress testing method based on artificial intelligence in this exemplary embodiment will be further described.

[0028] As described in step S110, determine the data type according to the test target defined by the user, wherein the data type includes stress testing data and / or real data.

[0029] In an embodiment of the present invention, the specific process of "determining the data type according to the test target defined by the user, where the data type includes stress test data and / or real data" in step S110 can be further described in combination with the following description.

[0030] As described in the following steps, S210. Generate a stress test task according to the test target; S220. Determine the target link nodes according to the stress test task; S230. Determine the parameter information according to the stress test task and the target link nodes; S240. Determine the data type according to the parameter information and the preset mirror production library, where the data type includes stress test data and / or real data.

[0031] It should be noted that the stress test task is defined according to the test target, and the core link nodes are determined; a shadow table library mirror production library, that is, the preset mirror production library, is created to isolate the stress test data and the real data.

[0032] As an example, the test target includes but is not limited to the target traffic and the business type ratio; the core link nodes include but are not limited to the registration center, distributed nodes, order service, payment service, and database nodes.

[0033] In a specific implementation, the test target is clarified and determined through performance indicators, business scenarios, and traffic models; the performance indicators include but are not limited to defining the target throughput, response time, and success rate; the business scenario is to determine the core links to be covered, including but not limited to the e-commerce order placement link and the payment link; the traffic model is to analyze the traffic ratio of business types according to historical logs, including but not limited to query requests and write requests.

[0034] As described in step S120, simulate traffic is generated according to the data type and the traffic ratio of business types obtained from the historical access logs.

[0035] In an embodiment of the present invention, the specific process of "generating simulated traffic according to the data type and the traffic ratio of business types obtained from the historical access logs" in step S120 can be further described in combination with the following description.

[0036] As described in the following steps, S310. Determine the target conditions according to the data type; S320. Determine the stress test rules according to the target conditions and the data type; S330. Perform data desensitization on the real data according to the stress test rules to generate a stress test data set; S340. Generate simulated traffic based on the stress test data set and the traffic ratio of business types obtained from historical access logs.

[0037] It should be noted that the stress test components are deployed to the online system according to the data type, intercept the stress test traffic and route it to the shadow environment; extract real data from the shadow database and desensitize it to generate the stress test data set; simulate the mixed traffic (i.e., real traffic and stress test traffic) according to the traffic ratio of business types.

[0038] As an example, the stress test parameters can be divided into static parameters and dynamic parameters. The static parameters include, but are not limited to, the stress test duration and the list of core link nodes; the dynamic parameters include, but are not limited to, the stress test traffic gradient and the fuse threshold; the stress test execution can use JMeter, LoadRunner or a self-developed stress test engine; the link tracing can be integrated with SkyWalking and Zipkin for core link node identification; use database tools such as mysqldump of MySQL or Data Pump of Oracle to export the production database Schema and data, that is, the target conditions; create a shadow database, that is, a preset mirror production database that prohibits production services from accessing; create an independent Topic for the message queue to isolate the stress test messages; use an independent cluster or prefix isolation for the cache; deploy at the service gateway and identify the stress test traffic through the request header; implant routing logic in the business service layer to route requests with stress test identifiers to the shadow database; gray-release the stress test components to the test environment to verify the interception and routing functions; dynamically manage the stress test switch through the configuration center; desensitize fields such as user IDs and mobile phone numbers in the shadow database; automatically empty the shadow database after the stress test to avoid residual data affecting subsequent tests; sample real request data from the production database and transfer it to the shadow database; use script tools to generate the stress test data set to cover boundary scenarios; configure parameterized requests in JMeter to dynamically replace fields such as user IDs and session Tokens; confirm the stress test traffic path through the link tracing tool; verify whether the stress test Mock service of the downstream system is ready; perform low-pressure tests to verify the environmental stability and data isolation effectiveness; check the logs and monitoring to ensure that there is no production data leakage or abnormal alarm; allocate an independent account for the stress test environment and restrict the operation permissions of the production environment; encrypt the stress test scripts and data storage to avoid unauthorized access; preset an automated fuse script to immediately terminate the stress test and trigger an alarm when a core service exception is monitored; arrange for operation and maintenance personnel to be on duty in real time and intervene at any time to handle uncovered abnormal scenarios, that is, stress test rules.

[0039] In a specific implementation, first retrieve the data of real users from the standby database, mask the information involving privacy, and make it into test-specific data; according to the real access ratio of the past website, mix the real user requests and test requests together, so that the measured effect is closer to the real situation.

[0040] As described in step S130, pressure test execution information is generated based on the simulated traffic and a preset fuzzy control model.

[0041] In an embodiment of the present invention, the specific process of "generating pressure test execution information based on the simulated traffic and a preset fuzzy control model" described in step S130 can be further described in combination with the following description.

[0042] As described in the following steps, S410. Determine each traffic weight value based on the simulated traffic and the preset fuzzy control model; S420. Generate a weight label list based on each of the traffic weight values; S430. Generate full-link nodes based on the weight label list and each distributed node; S440. Generate pressure test execution information by covering within the full-link nodes according to the simulated traffic.

[0043] It should be noted that the traffic information of the service instance is collected in real time. The traffic information includes but is not limited to real-time traffic, target traffic difference, and difference change amount. Calculate the traffic weight value of each instance; update the weight label through the instance management server, and dynamically allocate the pressure test request load to the distributed nodes to avoid overloading a single node; inject pressure test traffic through the pressure test component, and upload the pressure test identifier to the full link; the pressure test traffic is transmitted in the service simulation system to cover all core link nodes; support multi-protocol testing, and coordinate the pressure applicator and the test cluster through a unified control machine.

[0044] As an example, by staring at the status of each server in real time: how many requests there are currently, how much is still short of the target, and whether the recent traffic has increased or decreased; the pressure test component includes but is not limited to Tuxedo services and HTTP servers; multi-protocol testing includes but is not limited to Tuxedo domain connection and HTTP requests.

[0045] In a specific implementation, according to these health indicators, automatically classify the capacity that each server can bear, that is, the weight label. The task allocator will assign more test tasks to the servers with high bearing capacity, while ensuring that no server is overwhelmed; use the pressure test component to inject simulated pressure test traffic into the system; transmit a unified pressure test identifier in the request header or protocol, such as marking is_stress_test=true, to ensure the isolation of the pressure test traffic from the actual business; the pressure test traffic needs to completely cover the core nodes of the full link; track the traffic transmission path through the pressure test identifier to ensure that all key nodes receive the pressure test request; support multi-protocol testing such as Tuxedo domain connection, HTTP / S, RPC, etc.; use a unified control machine to coordinate the pressure applicator and the test cluster to ensure the synchronization and logical consistency of different protocol requests.

[0046] As described in step S150, node topology information is generated based on the performance metrics, the stress test execution information, and the preset hierarchical fusing conditions.

[0047] In an embodiment of the present invention, the specific process of "generating node topology information based on the performance metrics, the stress test execution information, and the preset hierarchical fusing conditions" described in step S150 can be further described in combination with the following description.

[0048] As described in the following steps, S510. Determine the processing mechanism based on the performance metrics and the preset hierarchical fusing conditions; S520. Generate node topology information based on the processing mechanism and the stress test execution information; S530. Generate stress test results based on the node topology information and the performance metrics.

[0049] As an example, a node topology diagram is generated according to the processing mechanism and the stress test execution information; a full-link node topology diagram is drawn according to the stress test traffic transfer path; nodes with high latency or high error rate are identified; the resource utilization rate is associated with the performance metrics.

[0050] In a specific implementation, the full-link node topology diagram may be, but is not limited to, order → payment → inventory → logistics; the node with high latency or high error rate has a timeout when the TPS of a certain service reaches 5000; the association between the resource utilization rate and the performance metrics may be, but is not limited to, a sharp increase in latency caused by full CPU load.

[0051] As described in step S510, the processing mechanism is determined based on the performance metrics and the preset hierarchical fusing conditions.

[0052] In an embodiment of the present invention, the specific process of "determining the processing mechanism based on the performance metrics and the preset hierarchical fusing conditions" described in step S510 can be further described in combination with the following description.

[0053] As described in the following steps, S610. Determine the metric attribute of the performance metrics, where the metric attribute includes the call success rate or latency; S620. When the metric attribute is the call success rate and the call success rate is less than the first preset call threshold, and / or when the metric attribute is the latency and the latency is greater than the first preset latency threshold, then determine the current preset hierarchical fusing condition as the first-level fusing condition, and determine reducing the current stress test traffic to a preset percentage range of the original traffic as the processing mechanism; and / or, S630. When the metric attribute is the call success rate and the call success rate is less than a second preset call threshold, where the second preset call threshold is less than the first preset call threshold, and / or when the metric attribute is the latency and the latency is greater than a second preset latency threshold, where the second preset latency threshold is greater than the first preset latency threshold, then determine that the current preset hierarchical fusing condition is the secondary fusing condition, and determine the isolation between problem nodes as the processing mechanism.

[0054] It should be noted that the downstream system is monitored, and the call success rate, latency, TPS (transactions per second), and resource utilization rate are collected in real time; the infrastructure data of the stress test nodes is reported to the monitoring platform in real time.

[0055] As an example, the latency can be but is not limited to TP99; the resource utilization rate can be but is not limited to CPU, memory, disk I / O, and network bandwidth; the infrastructure data can be but is not limited to CPU load > 80%, memory usage > 90%, and network packet loss rate > 1%.

[0056] In a specific implementation, for the primary fusing condition (i.e., traffic degradation), the call success rate is less than the first preset call threshold (such as 95%) or the latency is greater than the first preset latency threshold (such as 1 second); then the stress test traffic is automatically reduced to 50% of the original traffic, and the recovery of the metrics is continuously monitored. For the secondary fusing condition (i.e., link isolation), the call success rate is less than the second preset call threshold (such as 80%) or the latency is greater than the second preset latency threshold (such as 3 seconds); then the stress test switch of the abnormal link is turned off, and the problem nodes can be isolated, for example, by removing the problem instances through gray release, so as to avoid affecting the backbone service.

[0057] As described in step S160, determine the full-link stress test plan according to the node topology information and the preset stress test metrics.

[0058] In an embodiment of the present invention, the specific process of "determining the full-link stress test plan according to the node topology information and the preset stress test metrics" described in step S160 can be further described in combination with the following description.

[0059] As described in the following steps, S710. Generate stress test results according to the node topology information and the performance metrics; S720. Generate an optimization plan according to the stress test results and the preset stress test metrics; S730. Determine the full-link stress test plan according to the stress test results and the optimization plan.

[0060] It should be noted that an optimization plan is obtained by comparing the stress test results with the preset metrics, and finally a full-link stress test report is generated.

[0061] As an example, the optimization plan can be short-term and long-term. In the short term, instances are scaled up according to resource bottlenecks, including but not limited to increasing the number of payment service instances; in the long term, code optimization includes but not limited to reducing the printing of invalid logs; database optimization includes but not limited to adding indexes to frequently queried fields and sharding databases and tables; and asynchronous transformation includes but not limited to converting synchronous calls to message queues.

[0062] In a specific implementation, the full-link stress test report contains performance bottlenecks and optimization suggestions; updated system capacity planning and emergency response plans.

[0063] For the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple. For related parts, refer to the partial description of the method embodiment.

[0064] Refer to Figure 2 , which shows a structural block diagram of a full-link stress test system based on artificial intelligence provided by an embodiment of the present application; A full-link stress test system based on artificial intelligence, the system specifically includes: A first determination module 210, configured to determine a data type according to a test target defined by a user, where the data type includes stress test data and / or real data; A first generation module 220, configured to generate simulated traffic according to the data type and the traffic ratio of the service type obtained from the historical access logs; A second generation module 230, configured to generate stress test execution information according to the simulated traffic and a preset fuzzy control model; An acquisition module 240, configured to, when the stress test execution information is in execution, acquire performance metrics corresponding to the stress test execution information during the execution process; A third generation module 250, configured to generate node topology information according to the performance metrics, the stress test execution information, and a preset hierarchical fusing condition; A second determination module 260, configured to determine a full-link stress test plan according to the node topology information and a preset stress test index.

[0065] In an embodiment of the present invention, the first determination module 210 includes: A first generation sub-module, configured to generate a stress test task according to the test target; A first determination sub-module, configured to determine target link nodes according to the stress test task; A second determination sub-module, configured to determine parameter information according to the stress test task and the target link nodes; A third determination sub-module, configured to determine a data type according to the parameter information and a preset mirror production library, where the data type includes stress test data and / or real data.

[0066] In an embodiment of the present invention, the first generation module 220 includes: A fourth determination sub-module, configured to determine a target condition according to the data type; A fifth determination sub-module, configured to determine a stress test rule according to the target condition and the data type; A second generation sub-module, configured to perform data desensitization on the real data according to the stress test rule to generate a stress test data set; A third generation sub-module, configured to generate simulated traffic according to the stress test data set and the traffic proportion of the service type obtained based on the historical access log.

[0067] In an embodiment of the present invention, the second generation module 230 includes: A sixth determination sub-module, configured to determine respective traffic weight values according to the simulated traffic and a preset fuzzy control model; A fourth generation sub-module, configured to generate a weight label list according to the respective traffic weight values; A fifth generation sub-module, configured to generate full-link nodes according to the weight label list and each distributed node; A sixth generation sub-module, configured to generate stress test execution information by covering within the full-link nodes according to the simulated traffic.

[0068] In an embodiment of the present invention, the third generation module 250 includes: A seventh determination sub-module, configured to determine a processing mechanism according to the performance index and the preset hierarchical fusing condition; A seventh generation sub-module, configured to generate node topology information according to the processing mechanism and the stress test execution information; An eighth generation sub-module, configured to generate a stress test result according to the node topology information and the performance index.

[0069] In an embodiment of the present invention, the seventh determination sub-module includes: A determination unit, configured to determine an index attribute of the performance index, where the index attribute includes a call success rate or a delay; A first processing unit, configured to determine that the current preset hierarchical fusing condition is the first-level fusing condition, and determine that reducing the current stress test traffic to a preset percentage range of the original traffic is the processing mechanism when the index attribute is the call success rate and the call success rate is less than a first preset call threshold, and / or when the index attribute is the delay and the delay is greater than a first preset delay threshold; and / or A second processing unit, configured to determine that the current preset hierarchical fusing condition is the secondary fusing condition and isolate between problem nodes as the processing mechanism when the metric attribute is the call success rate and the call success rate is less than a second preset call threshold, where the second preset call threshold is less than the first preset call threshold, and / or when the metric attribute is the latency and the latency is greater than a second preset latency threshold, where the second preset latency threshold is greater than the first preset latency threshold.

[0070] In an embodiment of the present invention, the second determination module 260 includes: A ninth generation sub-module, configured to generate a stress test result according to the node topology information and the performance metric; A tenth generation sub-module, configured to generate an optimization plan according to the stress test result and a preset stress test metric; An eighth determination sub-module, configured to determine a full-link stress test plan according to the stress test result and the optimization plan.

[0071] Referring to Figure 3 , a computer device for an artificial intelligence-based full-link stress test method of the present invention is shown, which may specifically include the following: The above computer device 12 is presented in the form of a general computing device. The components of the computer device 12 may include, but are not limited to: one or more processors or processing units 16, a system memory 28, and a bus 18 connecting different system components (including the system memory 28 and the processing unit 16).

[0072] The bus 18 represents one or more of several types of bus 18 structures, including a memory bus 18 or a memory controller, a peripheral bus 18, a graphics acceleration port, a processor, or a local bus 18 using any bus 18 structure in a variety of bus 18 structures. For example, these architectures include, but are not limited to, an Industry Standard Architecture (ISA) bus 18, a Micro Channel Architecture (MAC) bus 18, an Enhanced ISA bus 18, a Video Electronics Standards Association (VESA) local bus 18, and a Peripheral Component Interconnect (PCI) bus 18.

[0073] The computer device 12 typically includes a variety of computer system-readable media. These media can be any available media accessible by the computer device 12, including volatile and non-volatile media, removable and non-removable media.

[0074] System memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. The computing device 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 34 can be used for reading from and writing to non-removable, non-volatile magnetic media (commonly referred to as a "hard disk drive"). Although Figure 3 not shown in FIG. Figure 3 , a disk drive for reading from and writing to a removable, non-volatile disk (such as a "floppy disk"), and an optical disk drive for reading from and writing to a removable, non-volatile optical disk (e.g., CD-ROM, DVD-ROM, or other optical media) can be provided. In these instances, each drive can be connected to the bus 18 by one or more data media interfaces. The memory may include at least one program product having a set (e.g., at least one) of program modules 42 that are configured to carry out the functions of various embodiments of the present invention.

[0075] A program / utility 40 having a set (at least one) of program modules 42 can be stored, for example, in the memory. Such program modules 42 include—but are not limited to—an operating system, one or more application programs, other program modules 42, and program data, and an implementation of a network environment may be included in each or some combination of these examples. The program modules 42 generally carry out the functions and / or methods in the embodiments described in the present invention.

[0076] The computing device 12 can also communicate with one or more external devices 14 (such as a keyboard, a pointing device, a display 24, a camera, etc.), and can also communicate with one or more devices that enable a user to interact with the computing device 12, and / or with any device that enables the computing device 12 to communicate with one or more other computing devices (such as a network card, a modem, etc.). Such communication can be through an input / output (I / O) interface 22. Also, the computing device 12 can communicate with one or more networks (such as a local area network (LAN)), a wide area network (WAN), and / or a public network (such as the Internet) through a network adapter 20. As shown, the network adapter 20 communicates with other modules of the computing device 12 through the bus 18. It should be understood that although Figure 3 not shown in FIG. Figure 3 , other hardware and / or software modules can be used in conjunction with the computing device 12, including but not limited to: microcode, device drivers, a redundant processing unit 16, an external disk drive array, a RAID system, a tape drive, and a data backup storage system 34, etc.

[0077] The processing unit 16 executes various functional applications and data processing by running the programs stored in the system memory 28, for example, implementing a full-link stress testing method based on artificial intelligence provided by the embodiments of the present invention.

[0078] That is, when the above-mentioned processing unit 16 executes the above-mentioned program, it realizes: determining the data type according to the test target defined by the user, wherein the data type includes stress test data and / or real data; generating simulated traffic according to the data type and the traffic ratio of the service type obtained based on the historical access log; generating stress test execution information according to the simulated traffic and a preset fuzzy control model; when the stress test execution information is in execution, obtaining the performance indicators corresponding to the stress test execution information during the execution process; generating node topology information according to the performance indicators, the stress test execution information and a preset hierarchical fusing condition; determining a full-link stress test plan according to the node topology information and a preset stress test index.

[0079] In the embodiments of the present invention, the present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it realizes a full-link stress testing method based on artificial intelligence provided by all embodiments of the present application: That is, when the program is executed by a processor, it realizes: determining the data type according to the test target defined by the user, wherein the data type includes stress test data and / or real data; generating simulated traffic according to the data type and the traffic ratio of the service type obtained based on the historical access log; generating stress test execution information according to the simulated traffic and a preset fuzzy control model; when the stress test execution information is in execution, obtaining the performance indicators corresponding to the stress test execution information during the execution process; generating node topology information according to the performance indicators, the stress test execution information and a preset hierarchical fusing condition; determining a full-link stress test plan according to the node topology information and a preset stress test index.

[0080] Any combination of one or more computer-readable media may be employed. The computer-readable media may be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium may be, for example - but not limited to - an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium include: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present document, a computer-readable storage medium may be any tangible medium that contains or stores a program which can be used by or in connection with an instruction execution system, apparatus, or device.

[0081] A computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which computer-readable program code is carried. Such a propagated data signal may take many forms, including - but not limited to - an electromagnetic signal, an optical signal, or any suitable combination of the foregoing. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device.

[0082] The computer program code for performing the operations of the present invention may be written in one or more programming languages or combinations thereof, including object-oriented programming languages - such as Java, Smalltalk, C++ - and also including conventional procedural programming languages - such as the "C" language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer, or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network - including a local area network (LAN) or a wide area network (WAN) - or, alternatively, may be connected to an external computer (e.g., through the Internet using an Internet service provider). Each embodiment in this specification is described in a progressive manner, with each embodiment focusing on the differences from other embodiments, and the same or similar parts among the embodiments may be referred to each other.

[0083] Although the preferred embodiments of the embodiments of the present application have been described, those skilled in the art can make additional changes and modifications once they learn the basic creative concepts. Therefore, the appended claims are intended to be construed to include the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present application.

[0084] Finally, it should also be noted that in this text, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or terminal device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or elements inherent to such process, method, article or terminal device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or terminal device comprising the element.

[0085] The above provides a detailed introduction to an artificial intelligence-based full-link stress testing method and system provided by the present application. Specific examples are used in this text to elaborate on the principles and implementation manners of the present application. The description of the above embodiments is only used to help understand the method and its core idea of the present application; at the same time, for those of ordinary skill in the art, according to the idea of the present application, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to the present application.

Claims

1. An all-link stress testing method based on artificial intelligence, characterized in that, The method includes: Determining a data type based on a user-defined test objective, where the data type includes stress test data and / or real data; Generating simulated traffic based on the data type and the traffic proportion of service types obtained from historical access logs; Generating stress test execution information based on the simulated traffic and a preset fuzzy control model; When the stress test execution information is in execution, obtaining performance metrics corresponding to the stress test execution information during the execution process; Generating node topology information based on the performance metrics, the stress test execution information, and preset hierarchical fusing conditions; Determining a full-link stress test plan based on the node topology information and preset stress test metrics.

2. The method according to claim 1, wherein The step of determining a data type based on a user-defined test objective, where the data type includes stress test data and / or real data, includes: Generating a stress test task based on the test objective; Determining target link nodes based on the stress test task; Determining parameter information based on the stress test task and the target link nodes; Determining a data type based on the parameter information and a preset mirror production library, where the data type includes stress test data and / or real data.

3. The method according to claim 1, wherein The step of generating simulated traffic based on the data type and the traffic proportion of service types obtained from historical access logs includes: Determining target conditions based on the data type; Determining stress test rules based on the target conditions and the data type; Performing data desensitization on the real data according to the stress test rules to generate a stress test data set; Generating simulated traffic based on the stress test data set and the traffic proportion of service types obtained from historical access logs.

4. The method according to claim 1, wherein The step of generating stress test execution information based on the simulated traffic and a preset fuzzy control model includes: Determining respective traffic weight values based on the simulated traffic and the preset fuzzy control model; Generating a weight label list based on the respective traffic weight values; Generating full-link nodes based on the weight label list and each distributed node; Generating stress test execution information by covering within the full-link nodes according to the simulated traffic.

5. The method according to claim 1, wherein The step of generating node topology information based on the performance metrics, the stress test execution information, and preset hierarchical fusing conditions includes: Determining a processing mechanism based on the performance metrics and the preset hierarchical fusing conditions; Generating node topology information based on the processing mechanism and the stress test execution information; Generating a stress test result based on the node topology information and the performance metrics.

6. The method according to claim 5, wherein The step of determining a processing mechanism based on the performance metrics and the preset hierarchical fusing conditions, where the preset hierarchical fusing conditions include a first-level fusing condition and / or a second-level fusing condition, includes: Determining the metric attribute of the performance metrics, where the metric attribute includes call success rate or latency; When the metric attribute is the call success rate and the call success rate is less than a first preset call threshold, and / or when the metric attribute is the latency and the latency is greater than a first preset latency threshold, then determining the current preset hierarchical fusing condition as the first-level fusing condition, and determining reducing the current stress test traffic to a preset percentage range of the original traffic as the processing mechanism; and / or, When the metric attribute is the call success rate and the call success rate is less than a second preset call threshold, where the second preset call threshold is less than the first preset call threshold, and / or when the metric attribute is the latency and the latency is greater than a second preset latency threshold, where the second preset latency threshold is greater than the first preset latency threshold, then determine that the current preset hierarchical fuse condition is the secondary fuse condition, and the isolation between problem nodes is determined as the processing mechanism.

7. The method according to claim 1, characterized in that, The step of determining the full-link stress test plan according to the node topology information and the preset stress test metrics includes: Generating a stress test result according to the node topology information and the performance metrics; Generating an optimization plan according to the stress test result and the preset stress test metrics; Determining the full-link stress test plan according to the stress test result and the optimization plan.

8. An artificial intelligence-based full-link stress testing system, characterized in that, The system includes: A first determination module, configured to determine a data type according to a test target defined by a user, where the data type includes stress test data and / or real data; A first generation module, configured to generate simulated traffic according to the data type and the traffic proportion of the service type obtained from the historical access log; A second generation module, configured to generate stress test execution information according to the simulated traffic and a preset fuzzy control model; An acquisition module, configured to acquire the performance metrics corresponding to the stress test execution information during the execution process when the stress test execution information is in execution; A third generation module, configured to generate node topology information according to the performance metrics, the stress test execution information, and a preset hierarchical fuse condition; A second determination module, configured to determine a full-link stress test plan according to the node topology information and the preset stress test metrics.

9. A computer device, characterized in that, It includes a processor, a memory, and a computer program stored on the memory and capable of running on the processor. When the computer program is executed by the processor, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium. When the computer program is executed by the processor, the method according to any one of claims 1 to 7 is implemented.

Images