WO2023125272A1 - Radius环境下的全链路压测方法、装置、计算机设备及存储介质 - Google Patents

Radius环境下的全链路压测方法、装置、计算机设备及存储介质 Download PDF

Info

Publication number
WO2023125272A1
WO2023125272A1 PCT/CN2022/141328 CN2022141328W WO2023125272A1 WO 2023125272 A1 WO2023125272 A1 WO 2023125272A1 CN 2022141328 W CN2022141328 W CN 2022141328W WO 2023125272 A1 WO2023125272 A1 WO 2023125272A1
Authority
WO
WIPO (PCT)
Prior art keywords
test
link
traffic
stress testing
threshold
Prior art date
Application number
PCT/CN2022/141328
Other languages
English (en)
French (fr)
Inventor
江艳
陈鑫远
詹振辉
Original Assignee
天翼物联科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 天翼物联科技有限公司 filed Critical 天翼物联科技有限公司
Publication of WO2023125272A1 publication Critical patent/WO2023125272A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/50—Testing arrangements

Definitions

  • the present invention relates to the technical field of the Internet of Things, in particular to a full-link pressure measurement method, device, computer equipment and storage medium in a Radius environment.
  • Remote Authentication Dial In User Service provides users with services such as authentication, authorization, and billing, and is widely used in scenarios such as machine-card binding and user Internet access.
  • IoT services With the continuous expansion of IoT services, the number of users also increases, which further leads to the increase in the scale and complexity of IoT services. Therefore, it is becoming more and more important to ensure the stable operation of RADIUS.
  • the current stress testing method is generally to perform stress testing on a single node by node and by business scenario, that is, first create a request for service node 1 to perform stress testing. Then, based on the collected results of the node 1 request falling in the database, the node 2 request is spliced.
  • the time interval of the pressure test between the two nodes is uncertain, and the pressure test cannot be performed on multiple nodes at the same time, which is different from the actual pressure test.
  • the test results are contradictory. At the same time, the test of a single node cannot accurately express the performance of the entire link.
  • Embodiments of the present invention provide a full-link stress test method, device, computer equipment, and storage medium in a Radius environment, which can not only perform stress tests on the full link online without interrupting normal running services, but also can According to the performance data of the downstream system, the stress test can be interrupted at any time to ensure the normal operation of the business.
  • the embodiment of the present invention provides a full-link stress testing method in a Radius environment, the method comprising:
  • Target links that require stress testing according to preset stress testing targets, wherein the target links include multiple core links of different business lines;
  • test traffic includes normal traffic and pressure test traffic
  • the embodiment of the present invention also provides a full-link pressure measurement device under the Radius environment, the device includes:
  • a target link confirmation unit configured to confirm a target link that needs to be subjected to stress testing according to a preset stress testing target, wherein the target link includes a plurality of core links of different business lines;
  • the first creation unit is used to create a mirror table library of a normal business table library to obtain a shadow table library of the normal business table library;
  • the first test unit is configured to jointly debug test traffic and perform the stress test on the target link, wherein the test traffic includes normal traffic and stress test traffic;
  • the first storage unit is configured to store the normal data generated by the normal traffic during the stress test into the normal business table library, and store the stress test data generated by the stress test traffic into the shadow table library;
  • the first real-time collection unit is configured to collect the performance data of the downstream system in the target link in real time, and judge whether to interrupt the pressure measurement according to the performance data.
  • an embodiment of the present invention further provides a computer device, which includes a memory and a processor, where a computer program is stored in the memory, and the above method is implemented when the processor executes the computer program.
  • an embodiment of the present invention further provides a computer-readable storage medium, the storage medium stores a computer program, and the computer program can implement the above method when executed by a processor.
  • Embodiments of the present invention provide a full-link stress testing method, device, computer equipment, and storage medium in a Radius environment.
  • the method includes: confirming the target link that needs to be tested according to the preset pressure test target, wherein the target link includes multiple core links of different business lines; creating a mirror table library of a normal business table library To obtain the shadow table library of the normal business table library; joint debugging test traffic to carry out the pressure test on the target link, wherein the test traffic includes normal traffic and pressure measurement traffic;
  • the normal data generated by the normal traffic is stored in the normal business table database, and the pressure measurement data generated by the pressure measurement traffic is stored in the shadow table database; real-time collection of the downstream system in the target link performance data, and judge whether to interrupt the pressure measurement according to the performance data.
  • the target link can be confirmed according to the preset stress test target, and the target link is formed by integrating all core links that need to be stress tested, so that all core links that need to be stress tested can be simultaneously Pressure testing, by storing the normal data generated by normal traffic into the normal business table database, and storing the pressure measurement data generated by the pressure measurement traffic into the shadow table database, so as to ensure that the entire business system can run normally during the pressure test, and realize the online
  • the performance data in the downstream system can be detected in real time during the stress test, and it can be judged whether to interrupt the stress test according to the performance data to avoid damage to the downstream system and affect the normal business operation.
  • FIG. 1 is a schematic flowchart of a full-link stress testing method in a Radius environment provided by an embodiment of the present invention
  • FIG. 2 is a schematic block diagram of a full-link stress testing method in a Radius environment provided by an embodiment of the present invention
  • FIG. 3 is a schematic block diagram of a full-link pressure measurement device in a Radius environment provided by an embodiment of the present invention
  • Fig. 4 is a schematic block diagram of a computer device provided by an embodiment of the present invention.
  • Fig. 1 is a schematic flow chart of the full-link stress testing method under the Radius environment provided by the embodiment of the present invention
  • Fig. 2 is a full-link stress testing method under the Radius environment provided by the embodiment of the present invention
  • the full-link pressure testing method in the Radius environment of the embodiment of the present invention is applied to any computer device. As shown in FIG. 1, the method includes steps S110-S150.
  • the entire business system architecture for example, the request link from one end to the other end, the technical architecture, layered structure, module division, and messages, caches involved
  • each core link needs to build more parameters Set. Different parameter sets represent different behaviors of different users.
  • the preset stress testing target includes all the above information. Therefore, the target link to be tested is obtained by analyzing the preset pressure test target, so as to prepare for the pressure test.
  • the entire business system includes multiple business services. Different business services are accessed through their corresponding links to realize their basic functions.
  • Service 1, Service 2, and Service 3 each have a link to the streaming log system, log collection module, and real-time/offline computing module, and the framework composed of log system, log collection module, and real-time/offline computing module has two links to the performance analysis module and log alarm module respectively.
  • the core links mentioned in the embodiments of the invention include all the links mentioned above and combinations of different links.
  • the test traffic enters Service 1, Service 2 and Service 3 from the Request in Figure 2, and will flow into Service 1, Service 2 and Service 3 respectively according to the pre-configured traffic ratio, for example, Service 1, Service The traffic ratio of Service 2 and Service 3 is 1:3:6 respectively, so the pressure test traffic in all test traffic will flow into Service 1, Service 2, and Service 3 according to the traffic ratio.
  • a pressure test switch can be set in service 1, service 2, and service 3 to control the pressure test traffic entering the corresponding service.
  • the full-link stress testing method in the Radius environment further includes the following steps: acquiring preset stress testing traffic, and coloring the preset stress testing traffic to obtain the The above pressure measurement flow.
  • the pressure measurement traffic in order to distinguish between normal traffic and pressure measurement traffic, can be dyed so that each business system can identify the pressure measurement traffic. At the same time, the business system obtains the Pressure measurement data is also marked.
  • the full-link stress test is carried out online, that is, it is necessary to ensure the normal operation of the business system while performing the stress test without affecting the normal data of the business system.
  • the shadow table library can store normal data and pressure test data separately to ensure the normal operation of the business system.
  • test traffic includes normal traffic and stress test traffic.
  • the step 120 may also include the following steps:
  • the step of jointly debugging the test traffic and performing separate stress testing on each core link in the target link may also include the following steps:
  • test traffic described in the joint debugging section performs a separate pressure test on each core link in the target link
  • the step of jointly debugging the test traffic and performing an overall pressure test on the target link may include the following steps:
  • test traffic described in the joint debugging part performs an overall pressure test on the target link
  • the joint debugging traffic includes stress testing traffic and normal traffic.
  • the normal traffic needs to be returned to the business system after the stress testing is over.
  • the stress testing traffic can be deleted directly after the testing is over, so it can When the flow rate and the pressure measurement flow are used, different flow marks are marked on the normal flow and the pressure measurement flow to distinguish the normal flow and the pressure measurement flow.
  • the joint debugging test traffic When the joint debugging test traffic is used to test the target link, the specific process is as follows: First, the joint debugging part of the test traffic performs a separate pressure test on each core link, that is, tests whether each link can be accessed normally through small traffic , after completing the small traffic test on each core link, you can jointly adjust all the test traffic to perform stress testing on each core link, so as to test the maximum performance of a single core link without interference and competition. Load and related baseline data.
  • test traffic After completing the test of a single core link, you can further jointly debug part of the test traffic to perform pressure testing on the entire target link, so as to test whether each core link can access normally when concurrent access occurs, and finally, joint debugging All test traffic, and then according to the traffic ratio of different core links, all test traffic is jointly dispatched to the corresponding core link, so as to test the resource consumption level of different core links in the case of mutual interference and competition and bottlenecks.
  • the normal flow and the pressure measurement flow have their own flow labels, and the normal flow and the pressure measurement flow can be identified through the flow labels, and the normal flow is stored in the normal business table database, and the pressure measurement flow is stored in the shadow
  • the table library is convenient for deleting the shadow table library after the pressure test is completed. Among them, the pressure measurement data obtained after the dyed pressure measurement traffic is processed by the business system is marked to distinguish it from normal data.
  • S150 Collect performance data of a downstream system in the target link in real time, and judge whether to interrupt the pressure measurement according to the performance data.
  • the step S150 may include the following steps:
  • the downstream system can refer to the performance analysis module and the log alarm module
  • the downstream interface refers to the framework composed of the streaming log system, the log collection module, and the real-time/offline computing module, as well as the performance analysis module and the log alarm module. Interface between modules.
  • the preset call success rate tolerable exception threshold and the preset delay tolerable exception threshold are set by the user. For example, the preset call success rate tolerable exception threshold is 99%, and the preset delay tolerable exception threshold is 40ms.
  • the performance analysis framework When the success rate of the performance analysis framework invoking the data of the streaming log system is greater than 99%, and the delay of invoking data is less than 40ms, the performance analysis framework is in a normal state and has not yet reached full capacity. On the contrary, it indicates that the performance framework is abnormal, and it is necessary to interrupt the pressure test or reduce the pressure test traffic entering service 1, service 2, or service 3 to avoid problems in the performance analysis framework.
  • the preset call success rate tolerable exception threshold includes a call success rate first threshold and call success rate second threshold
  • the preset delay tolerable exception threshold includes time The first delay threshold and the second delay threshold, the size of the tolerable exception threshold according to the call success rate and the call success rate, the delay index and the preset delay tolerable exception threshold
  • the first threshold of call success rate may be 99.9%
  • the second threshold of call success rate may be 99%
  • the first threshold of delay may be 40ms
  • the second threshold of delay may be 50ms. If the service The call success rate of the performance analysis framework in 1 is between 99% and 99.9%, and the delay index is between 40ms and 50ms, indicating that the performance analysis framework is in an abnormal state and needs to reduce the pressure measurement traffic entering service 1. If calling If the success rate is below 99%, and the delay index is above 50ms, it indicates that the performance analysis framework has been overloaded, and the pressure test needs to be interrupted, or the pressure test traffic entering service 1 should be closed. You can control the pressure measurement traffic to enter the corresponding service through the pressure measurement switch set in each service to ensure that the normal operation of the entire business system is not affected.
  • the moving average method can be used to dynamically identify whether the call success rate and delay index of the downstream system have changed drastically, so as to judge whether it is necessary to interrupt the pressure test or reduce the access to the downstream System pressure measurement flow.
  • the specific process is as follows: Suppose there are N i concurrent requests for pressure measurement at the same time, the corresponding call success rate is C i , and the delay index is T i . If the size of the sliding window is M, it can be updated by moving the sliding window M The moving average of call success rate C i and the moving standard deviation of the calling success rate Ci The specific calculation formula is as follows:
  • the moving average of the delay index can be obtained and moving standard deviation when and and , the pressure measurement data is reduced to enter the corresponding service at this time; when and or , at this time, turn off the pressure test switch of the corresponding service, where a is the first threshold of call success rate, aa is the second threshold of call success rate, b is the first threshold of delay, and bb is the second threshold of delay.
  • the full-link stress testing method under the Radius environment further includes the following steps: monitor the performance of all hardware devices involved in the stress testing process in real time to obtain the The performance index data; real-time monitoring of the time-consuming processing of the test traffic by each service node during the stress testing process to obtain a time index.
  • the hardware device refers to devices such as computers, and the performance index data may be the status of indicators such as CPU, memory, and disk of the computer device, for example, CPU usage, memory usage, disk usage, and the like.
  • the performance analysis framework receives the test traffic, it needs to call the CPU, memory, and disk to process the received test traffic.
  • the performance analysis module can process the CPU usage of the received traffic, memory Utilization rate and disk usage rate, as well as the time consumption of the entire processing process, so that when optimizing, you can know exactly which aspect to optimize.
  • FIG. 3 is a schematic block diagram of a full-link stress testing device 100 in a Radius environment provided by an embodiment of the present invention.
  • the present invention also provides a full-link stress testing device 100 in the Radius environment.
  • the full-link stress testing device 100 under the Radius environment includes a unit for performing the above-mentioned full-link stress testing method under the Radius environment.
  • the full-link pressure testing device 100 under the Radius environment includes a target link confirmation unit 110, a first creation unit 120, a first test unit 130, a first storage unit 140, and a first real-time collection Unit 150.
  • the target link confirmation unit 110 is configured to confirm the target link that needs to be subjected to stress test according to the preset stress test target, wherein the target link includes multiple core links of different business lines;
  • the first creation unit 120 uses To create a mirror table library of a normal business table library to obtain a shadow table library of the normal business table library;
  • the first test unit 130 is used to jointly debug test traffic and perform the pressure test on the target link, wherein the The test flow includes normal flow and pressure measurement flow;
  • the first storage unit 140 is used to store the normal data generated by the normal flow during the pressure test into the normal business table library, and store the normal data generated by the pressure measurement flow
  • the pressure measurement data is stored in the shadow table database;
  • the first real-time collection unit 150 is used to collect the performance data of the downstream system in the target link in real time, and judge whether to interrupt the pressure measurement according to the performance data.
  • the first testing unit 130 includes a first testing unit and a second testing unit.
  • the first test unit is used for joint debugging of the test traffic to carry out separate pressure test on each core link in the target link;
  • the second test unit is used for joint debugging of the test traffic to the target link Perform an overall pressure test.
  • the first testing unit includes a third testing unit and a fourth testing unit.
  • the third test unit is used for jointly debugging part of the test traffic to perform a separate pressure test on each core link in the target link; the fourth test unit is used for joint debugging of all the test traffic to the target Each core link in the link is subjected to a separate stress test.
  • the second testing unit includes a fifth testing unit and a sixth testing unit.
  • the fifth test unit is used for jointly debugging part of the test traffic to perform an overall pressure test on the target link; the sixth test unit is used to jointly debug all the test traffic to perform an overall pressure test on the target link.
  • the first real-time collection unit 150 includes a second real-time collection unit and a first acquisition unit.
  • the second real-time collection unit is used to collect the call success rate and delay index of the downstream interface in the downstream system in real time;
  • the first acquisition unit is used to obtain the preset call success rate tolerable abnormal threshold and the preset time delay Tolerant exception threshold, judging whether to interrupt the pressure test according to the call success rate and the tolerable exception threshold of the call success rate, the delay index and the preset delay tolerable exception threshold.
  • the preset call success rate tolerable exception threshold includes a call success rate first threshold and call success rate second threshold
  • the preset delay tolerable exception threshold includes time Delayed by the first threshold and delayed by the second threshold
  • the first acquisition unit includes a first processing unit, a first confirmation unit, and a second processing unit.
  • the first processing unit is configured to if the call success rate is less than the first threshold of the call success rate and greater than the second threshold of the call success rate, and the delay index is greater than the first delay threshold and less than The time delay second threshold, then reduce the pressure measurement flow to reduce the pressure of the downstream system; the first confirmation unit is used to if the call success rate is less than the second call success rate threshold, and the time If the delay index is greater than the second delay threshold, then confirm the core link corresponding to the downstream interface; the second processing unit is used to close the preset pressure measurement switch in the core link to avoid the pressure measurement traffic into the core chain.
  • Another embodiment of the present invention also provides a full-link pressure measurement device under the Radius environment.
  • the full-link pressure measurement device under the Radius environment in this embodiment is based on the above-mentioned embodiment.
  • the second acquisition unit is added. .
  • the second acquiring unit is configured to acquire a preset pressure measurement flow, and color the preset pressure measurement flow to obtain the pressure measurement flow.
  • Another embodiment of the present invention also provides a full-link pressure measurement device under the Radius environment.
  • the full-link pressure measurement device under the Radius environment in this embodiment is based on the above-mentioned embodiment.
  • the second real-time monitoring unit is used for real-time monitoring the performance of all hardware devices involved in the stress testing process to obtain the performance index data of the hardware devices; the second real-time monitoring unit is used for real-time monitoring during the stress testing process The time consumed by each service node in processing the test traffic to obtain the time index.
  • the above-mentioned full-link stress testing device under the Radius environment can be implemented in the form of a computer program, and the computer program can be run on the computer equipment shown in FIG. 4 .
  • FIG. 4 is a schematic block diagram of a computer device provided by an embodiment of the present application.
  • the computer device 500 includes a processor 502 connected through a system bus 501 , a memory and an interface 507 , wherein the memory may include a non-volatile storage medium 503 and an internal memory 504 .
  • the non-volatile storage medium 503 can store an operating system 5031 and a computer program 5032 .
  • the computer program 5032 When executed, it can cause the processor 502 to execute a full-link pressure testing method under the Radius environment.
  • the processor 502 is used to provide calculation and control capabilities to support the operation of the entire computer device 500 .
  • the internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503.
  • the processor 502 can execute the above-mentioned full-link stress testing method under the Radius environment. any of the examples.
  • the interface 505 is used to communicate with other devices.
  • the structure shown in FIG. 4 is only a block diagram of a partial structure related to the solution of this application, and does not constitute a limitation to the computer device 500 on which the solution of this application is applied.
  • the specific computer device 500 may include more or fewer components than shown, or combine certain components, or have a different arrangement of components.
  • the processor 502 is configured to run a computer program 5032 stored in the memory, so as to realize the following steps:
  • Target links that require stress testing according to preset stress testing targets, wherein the target links include multiple core links of different business lines;
  • test traffic includes normal traffic and pressure test traffic
  • processor 502 when the processor 502 implements the step of performing the pressure test on the target link with the joint debugging test traffic, it specifically implements the following steps:
  • processor 502 when the processor 502 implements the step of jointly debugging the test traffic and performing individual stress testing on each core link in the target link, the following steps are specifically implemented:
  • test traffic described in the joint debugging section performs a separate pressure test on each core link in the target link
  • processor 502 when the processor 502 implements the step of jointly debugging the test traffic and performing an overall pressure test on the target link, it specifically implements the following steps:
  • test traffic described in the joint debugging part performs an overall pressure test on the target link
  • the processor 502 realizes the step of collecting the performance data of the downstream system in the target link in real time, and judging whether to interrupt the pressure measurement according to the performance data, the following steps are specifically implemented:
  • the processor 502 realizes the value of the tolerable exception threshold according to the call success rate and the call success rate, the delay index and the preset delay tolerable exception threshold When judging whether to interrupt the pressure measurement step, the following steps are specifically implemented:
  • processor 502 also implements the following steps:
  • processor 502 also implements the following steps:
  • the processor 502 may be a central processing unit (Central Processing Unit, CPU), and the processor 502 may also be other general-purpose processors, digital signal processors (Figal Signal Processor, FSP), Application Specific IntegrateF Circuit (ASIC), off-the-shelf programmable gate array (FielF-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.
  • the general-purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
  • the computer program can be stored in a storage medium, which is a computer-readable storage medium.
  • the computer program is executed by at least one processor in the computer system to implement the process steps of the above method embodiments.
  • the present invention also provides a storage medium.
  • the storage medium may be a computer readable storage medium.
  • the storage medium stores a computer program. When the computer program is executed by the processor, any embodiment of the above-mentioned full-link pressure testing method under the Radius environment is realized.
  • the storage medium can be various computer-readable storage media that can store program codes such as U disk, mobile hard disk, read-only memory (ReaF-Only Memory, ROM), magnetic disk or optical disk.
  • program codes such as U disk, mobile hard disk, read-only memory (ReaF-Only Memory, ROM), magnetic disk or optical disk.
  • the disclosed devices and methods can be implemented in other ways.
  • the device embodiments described above are illustrative only.
  • the division of each unit is only a logical function division, and there may be another division method in actual implementation.
  • several units or components may be combined or integrated into another system, or some features may be omitted, or not implemented.
  • each functional unit in each embodiment of the present invention may be integrated into one processing unit, each unit may exist separately physically, or two or more units may be integrated into one unit.
  • the integrated unit is realized in the form of a software function unit and sold or used as an independent product, it can be stored in a storage medium.
  • the technical solution of the present invention is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium Among them, several instructions are included to make a computer device execute all or part of the steps of the methods described in the various embodiments of the present invention.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Environmental & Geological Engineering (AREA)
  • Investigating Strength Of Materials By Application Of Mechanical Stress (AREA)
  • Measuring Volume Flow (AREA)
  • Measuring Fluid Pressure (AREA)

Abstract

本发明实施例公开了一种Radius环境下的全链路压测方法、装置、计算机设备及存储介质，所述方法包括：根据预设压测目标确认需要进行压测的目标链路，其中，所述目标链路包括多个不同业务线的核心链路；创建正常业务表库的镜像表库以获得所述正常业务表库的影子表库；联调测试流量对所述目标链路进行所述压测，其中，所述测试流量包括正常流量和压测流量；将在压测过程中由所述正常流量所产生的正常数据存入所述正常业务表库，将由所述压测流量所产生的压测数据存入所述影子表库；实时采集所述目标链路中的下游系统的性能数据，并根据所述性能数据判断是否中断所述压测。本发明能够在线上实现全链路压测，提高压测的准确性。

Description

Radius环境下的全链路压测方法、装置、计算机设备及存储介质 技术领域
本发明涉及物联网技术领域，尤其涉及一种Radius环境下的全链路压测方法、装置、计算机设备及存储介质。
背景技术
用户远程拨入认证服务(Remote Authentication Dial In User Service，RADIUS)为用户提供认证、授权和计费等服务，广泛用于机卡绑定和用户上网等场景。随着物联网业务的不断扩张，用户数量也随之提升，进一步导致物联网的业务规模和复杂性随之增加，因此，确保RADIUS能够稳定运行也越来越重要。
为了确保RADIUS能够稳定运行，需要在RADIUS环境下，进行压测，目前的压测方法一般是分节点、分业务场景进行单个节点的压测，也即先制造业务节点1的请求进行压测，然后再通过收集到的节点1请求落在数据库里的结果，进行拼接节点2请求，2个节点之间压测的时间间隔不确定，且不能同时对多个节点进行压测，与实际的压测结果有悖。同时，单个节点的测试并不能准确地表达出全链路的性能，在实际运用中，一般是多个链路串起来同时工作，单个节点的测试并不能测出多个链路同时工作时的情况，且当多链路同时运行时，只要其中一个链路挂掉，会引起整个链路的崩溃。
发明内容
本发明实施例提供了一种Radius环境下的全链路压测方法、装置、计算机设备及存储介质，不仅能够通过在线上对全链路进行压力测试，无需中断正常运行的业务，而且还可以根据下游系统的性能数据随时中断压力测试，以确保业务的正常运行。
第一方面，本发明实施例提供了一种Radius环境下的全链路压测方法，该方法包括：
根据预设压测目标确认需要进行压测的目标链路，其中，所述目标链路包括多个不同业务线的核心链路；
创建正常业务表库的镜像表库以获得所述正常业务表库的影子表库；
联调测试流量对所述目标链路进行所述压测，其中，所述测试流量包括正常流量和压测流量；
将在压测过程中由所述正常流量所产生的正常数据存入所述正常业务表库，将由所述压测流量所产生的压测数据存入所述影子表库；
实时采集所述目标链路中的下游系统的性能数据，并根据所述性能数据判断是否中断所述压测。
第二方面，本发明实施例还提供了一种Radius环境下的全链路压测装置，该装置包括：
目标链路确认单元，用于根据预设压测目标确认需要进行压测的目标链路，其中，所述目标链路包括多个不同业务线的核心链路；
第一创建单元，用于创建正常业务表库的镜像表库以获得所述正常业务表库的影子表库；
第一测试单元，用于联调测试流量对所述目标链路进行所述压测，其中，所述测试流量包括正常流量和压测流量；
第一存储单元，用于将在压测过程中由所述正常流量所产生的正常数据存入所述正常业务表库，将由所述压测流量所产生的压测数据存入所述影子表库；
第一实时采集单元，用于实时采集所述目标链路中的下游系统的性能数据，并根据所述性能数据判断是否中断所述压测。
第三方面，本发明实施例还提供了一种计算机设备，其包括存储器及处理器，所述存储器上存储有计算机程序，所述处理器执行所述计算机程序时实现上述方法。
第四方面，本发明实施例还提供了一种计算机可读存储介质，所述存储介质存储有计算机程序，所述计算机程序当被处理器执行时可实现上述方法。
本发明实施例提供了一种Radius环境下的全链路压测方法、装置、计算机设备及存储介质。其中，所述方法包括：根据预设压测目标确认需要进行压测的目标链路，其中，所述目标链路包括多个不同业务线的核心链路；创建正常业务表库的镜像表库以获得所述正常业务表库的影子表库；联调测试流量对所述目标链路进行所述压测，其中，所述测试流量包括正常流量和压测流量；将在压测过程中由所述正常流量所产生的正常数据存入所述正常业务表库，将由所述压测流量所产生的压测数据存入所述影子表库；实时采集所述目标链路中的下游系统的性能数据，并根据所述性能数据判断是否中断所述压测。本发明实施例可以根据预设压测目标确认目标链路，而目标链路是由所有需要进行压测的核心链路整合而成，从而可以实现对所有需要进行压测的核心链路同时进行压测，通过将正常流量产生的正常数据存入至正常业务表库，将压测流量产生的压测数据存入影子表库，从而确保整个业务系统在压测的同时可以正常运行，实现线上压测，同时，在压测过程中还可以实时检测下游系统中的性能数据，并且根据性能数据判断是否需要中断压测，避免对下游系统造成损害影响正常业务的运行，通过上述方法，不仅能够实现在线上进行全链路压测的同时不影响正常业务的运行，而且还可以获得更高的测试精度。
附图说明
为了更清楚地说明本发明实施例技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1是本发明实施例提供的Radius环境下的全链路压测方法的流程示意图；
图2是本发明实施例提供的Radius环境下的全链路压测方法的方框示意图；
图3是本发明实施例提供的Radius环境下的全链路压测装置的示意性框图；
图4是本发明实施例提供的计算机设备的示意性框图。
具体实施方式
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
应当理解，当在本说明书和所附权利要求书中使用时，术语“包括”和“包含”指示所描述特征、整体、操作、元素和/或组件的存在，但并不排除一个或多个其它特征、整体、操作、元素、组件和/或其集合的存在或添加。
还应当理解，在此本发明说明书中所使用的术语仅仅是出于描述特定实施例的目的而并不意在限制本发明。如在本发明说明书和所附权利要求书中所使用的那样，除非上下文清楚地指明其它情况，否则单数形式的“一”、“一个”及“该”意在包括复数形式。还应当进一步理解，在本发明说明书和所附权利要求书中使用的术语“和/或”是指相关联列出的项中的一个或多个的任何组合以及所有可能组合，并且包括这些组合。
请参阅图1和图2，图1是本发明实施例提供的Radius环境下的全链路压测方法的流程示意图，图2是本发明实施例提供的Radius环境下的全链路压测方法的方框示意图。本发明实施例的Radius环境下的全链路压测方法应用于任一计算机设备。如图1所示，该方法包括步骤S110～S150。
S110，根据预设压测目标确认需要进行压测的目标链路，其中，所述目标链路包括多个不同业务线的核心链路。
在本发明实施例中，在进行压测之前，需要梳理整个业务系统架构，例如，一个端到另一个端的请求链路，所涉及到的技术架构、分层结构、模块划分，以及消息、缓存和数据库等中间件的使用情况，同时，还需要梳理整个业务系统的全链路核心业务模型，并且确定需要进行压测的核心链路，其中，每一个核心链路均需要构建较多的参数集合，不同的参数集合代表不同用户的不同行为，同时，还需要确认核心链路的链路范围、链路的访问量级和基础数据，便于初步掌握需要进行压测的核心链路的情况，以便于用户设置压测策略和压测目的，而预设压测目标包括上述所有信息。故通过解析预设压测目标获得需要进行压测的目标链路以便于准备进行压测。
整个业务系统包括多个业务服务，不同的业务服务通过各自对应的链路进行访问以实现其基本功能，如图2所示，服务1、服务2和服务3分别有一条链 路到流式日志系统、日志采集模块和实时/离线计算模块，而日志系统、日志采集模块和实时/离线计算模块所组成的框架有两条链路分别到性能分析模块和日志告警模块，需要注意的是，本发明实施例中所提到的核心链路包括上述提到的所有链路以及不同链路的组合。在进行压测时，测试流量从图2中Request处进入服务1、服务2和服务3，且会根据预先配置的流量配比分别流入服务1、服务2和服务3，例如，服务1、服务2和服务3的流量配比分别为1:3:6，则所有的测试流量中的压测流量会根据流量配比分别流入服务1、服务2和服务3。另外，由于整个压测是在业务系统正常运行中进行测试，业务系统正常运行的情况下会有正常流量分别访问服务1、服务2和服务3，在正常流量的基础上增加压测流量，从而形成测试流量进行压测。可以通过在服务1、服务2和服务3中分别设有一压测开关，以控制进入对应服务的压测流量。
在某些实施例，例如本实施例中，所述Radius环境下的全链路压测方法还包括如下步骤：获取预设压测流量，并对所述预设压测流量进行染色以获得所述压测流量。
在本发明实施例中，为了区别正常流量和压测流量，可以对压测流量进行染色，便于各个业务系统可以识别压测流量，同时，业务系统在处理染色后的压测流量后所获得的压测数据同样带有标记。
S120，创建正常业务表库的镜像表库以获得所述正常业务表库的影子表库。
在本发明实施例中，全链路压测是在线上进行的，也即需要在进行压测的同时，确保业务系统的正常运行，不影响业务系统的正常数据，通过创建正常业务表库的影子表库以分别存储正常数据和压测数据，确保业务系统的正常运行。
S130，联调测试流量对所述目标链路进行所述压测，其中，所述测试流量包括正常流量和压测流量。
在本发明实施例中，所述步骤120还可以包括如下步骤：
联调所述测试流量对所述目标链路中的每一个核心链路进行单独压测；
联调所述测试流量对所述目标链路进行整体压测；
其中，所述步骤联调所述测试流量对所述目标链路中的每一个核心链路进行单独压测还可以包括如下步骤：
联调部分所述测试流量对所述目标链路中的每一个核心链路进行单独压测；
联调所有所述测试流量对所述目标链路中的每一个核心链路进行单独压测；
其中，所述联调所述测试流量对所述目标链路进行整体压测的步骤可以包括如下步骤：
联调部分所述测试流量对所述目标链路进行整体压测；
联调所有所述测试流量对所述目标链路进行整体压测。
在整个压测期间，联调的流量包括压测流量和正常流量，正常流量在压测结束后需要回归到业务系统中，压测流量在测试结束后，可以直接删除，故可 以在联调正常流量和压测流量时，分别对正常流量和压测流量标记上不同的流量标，以区分正常流量和压测流量。在联调测试流量对目标链路进行测试时，具体过程如下：先联调部分测试流量分别对每个核心链路进行单独的压测，也即通过小流量测试每个链路是否可以正常访问，在完成对每个核心链路的小流量测试后，可以联调所有测试流量对每个核心链路进行压测，以此来测试单个核心链路在没有干扰，没有竞争的情况下的最大负荷以及相关基线数据。在完成对单个核心链路的测试后，可以进一步联调部分测试流量对整个目标链路进行压测，以此来测试各个核心链路出现并发访问时，是否能够正常访问，最后，再联调所有测试流量，然后根据不同核心链路的流量配比，将所有测试流量联调至对应的核心链路，以此来测试不同核心链路在存在相互干扰，相互竞争的情况下的资源消耗水平和瓶颈。
S140，将在压测过程中由所述正常流量所产生的正常数据存入所述正常业务表库，将由所述压测流量所产生的压测数据存入所述影子表库。
在本发明实施例中，正常流量和压测流量均有各自的流量标，可以通过流量标识别正常流量和压测流量，并将正常流量存入正常业务表库，将压测流量存入影子表库，便于在压测结束后，删除影子表库。其中，染色后的压测流量在经业务系统处理后获得的压测数据带有标记以区分于正常数据。
S150，实时采集所述目标链路中的下游系统的性能数据，并根据所述性能数据判断是否中断所述压测。
在本发明实施例中，所述步骤S150可以包括如下步骤：
实时采集所述下游系统中的下游接口的调用成功率和时延指标；
获取预设调用成功率可容忍异常阈值和预设时延可容忍异常阈值，根据所述调用成功率和所述调用成功率可容忍异常阈值的大小以及所述时延指标和所述预设时延可容忍异常阈值的大小判断是否中断所述压测。
如图2所示，下游系统可以是指性能分析模块和日志告警模块，而下游接口则是指流式日志系统、日志采集模块和实时/离线计算模块所组成的框架与性能分析模块和日志告警模块之间的接口。通过实时采集下游接口的调用成功率和时延指标来判断当前相对应的框架是否出现异常。预设调用成功率可容忍异常阈值和预设时延可容忍异常阈值由用户进行设定，例如，预设调用成功率可容忍异常阈值为99％，预设时延可容忍异常阈值40ms，则当性能分析框架调用流式日志系统的数据的成功率大于99％时，且调用数据的时延小于40ms时，性能分析框架处于正常状态，尚未达到满负荷运转。反之，则表明性能框架出现异常，需要中断压测或者降低进入服务1、服务2或者服务3的压测流量以避免性能分析框架出现问题。
在某些实施例，例如本实施例中，所述预设调用成功率可容忍异常阈值包括调用成功率第一阈值和调用成功率第二阈值，所述预设时延可容忍异常阈值包括时延第一阈值和时延第二阈值，所述根据所述调用成功率和所述调用成功率可容忍异常阈值的大小以及所述时延指标和所述预设时延可容忍异常阈值的大小判断是否中断所述压测的步骤包括如下步骤：
若所述调用成功率小于所述调用成功率第一阈值且大于所述调用成功率第二阈值，并且所述时延指标大于所述时延第一阈值且小于所述时延第二阈值，则降低所述压测流量以降低所述下游系统的压力；
若所述调用成功率小于所述调用成功率第二阈值，且所述时延指标大于所述时延第二阈值，则确认所述下游接口所对应的核心链路；
关闭所述核心链路中预设的压测开关以避免所述压测流量进入所述核心链路。
在本发明实施例中，调用成功率第一阈值可以是99.9％，调用成功率第二阈值可以是99％，时延第一阈值可以是40ms，时延第二阈值可以是50ms，则若服务1中的性能分析框架的调用成功率在99％和99.9％之间，且时延指标在40ms和50ms之间，表明性能分析框架处于异常状态，需要降低进入服务1的压测流量，若调用成功率在99％以下，且时延指标在50ms以上，则表明性能分析框架已经超负荷工作，需要中断压测，或者关闭进入服务1的压测流量。可以通过设置在每一个服务中的压测开关控制压测流量进入相对应的服务，以确保不影响整个业务系统的正常运行。
为了进一步准确识别下游系统的下游接口是否达到满负荷运转，可以通过移动均值法动态识别下游系统的调用成功率和时延指标是否出现剧烈变化，以此来判断是否需要中断压测或者降低进入下游系统的压测流量。具体过程如下：设同时有N i个压测并发请求，则与之对应的调用成功率为C i，时延指标为T i，若滑动窗口大小为M，则可以通过移动滑动窗口M来更新调用成功率C i的移动均值
Figure PCTCN2022141328-appb-000001
和调用成功率Ci的移动标准差
Figure PCTCN2022141328-appb-000002
具体计算公式如下：
Figure PCTCN2022141328-appb-000003
Figure PCTCN2022141328-appb-000004
同理可得时延指标的移动均值
Figure PCTCN2022141328-appb-000005
和移动标准差
Figure PCTCN2022141328-appb-000006
当
Figure PCTCN2022141328-appb-000007
且
Figure PCTCN2022141328-appb-000008
且
Figure PCTCN2022141328-appb-000009
Figure PCTCN2022141328-appb-000010
时，此时降低压测数据进入对应的服务；当
Figure PCTCN2022141328-appb-000011
且
Figure PCTCN2022141328-appb-000012
或
Figure PCTCN2022141328-appb-000013
时，此时，关闭对应服务的压测开关，其中，a为调用成功率第一阈值，aa为调用成功率第二阈值，b为时延第一阈值，bb为时延第二阈值。
在某些实施例，例如本实施例中，所述Radius环境下的全链路压测方法还包括如下步骤：实时监测在压测过程中所涉及的所有硬件设备的性能以获得所述硬件设备的性能指标数据；实时监测在所述压测过程中各个业务节点处理所述测试流量的耗时以获得时间指标。
在本发明实施例中，硬件设备是指计算机等设备，性能指标数据可以是计算机设备的CPU、内存和磁盘等指标的状态，例如，CPU使用率，内存使用率，磁盘使用率等等。如图2所示，性能分析框架在接收到的测试流量后，需要调 用CPU、内存以及磁盘对所接收到的测试流量进行处理，可以性能分析模块处理所接收到的流量的CPU使用率、内存使用率和磁盘使用率，以及整个处理过程的耗时，从而便于在进行优化时，可以准确知道从哪个方面进行优化。
图3是本发明实施例提供的一种Radius环境下的全链路压测装置100的示意性框图。如图3所示，对应于以上Radius环境下的全链路压测方法，本发明还提供一种Radius环境下的全链路压测装置100。该Radius环境下的全链路压测装置100包括用于执行上述Radius环境下的全链路压测方法的单元。具体地，请参阅图3，该Radius环境下的全链路压测装置100包括目标链路确认单元110、第一创建单元120、第一测试单元130、第一存储单元140和第一实时采集单元150。
其中，目标链路确认单元110用于根据预设压测目标确认需要进行压测的目标链路，其中，所述目标链路包括多个不同业务线的核心链路；第一创建单元120用于创建正常业务表库的镜像表库以获得所述正常业务表库的影子表库；第一测试单元130用于联调测试流量对所述目标链路进行所述压测，其中，所述测试流量包括正常流量和压测流量；第一存储单元140用于将在压测过程中由所述正常流量所产生的正常数据存入所述正常业务表库，将由所述压测流量所产生的压测数据存入所述影子表库；第一实时采集单元150用于实时采集所述目标链路中的下游系统的性能数据，并根据所述性能数据判断是否中断所述压测。
在某些实施例，例如本实施例中，所述第一测试单元130包括第一测试单元和第二测试单元。
其中，第一测试单元用于联调所述测试流量对所述目标链路中的每一个核心链路进行单独压测；第二测试单元用于联调所述测试流量对所述目标链路进行整体压测。
在某些实施例，例如本实施例中，所述第一测试单元包括第三测试单元和第四测试单元。
其中，第三测试单元用于联调部分所述测试流量对所述目标链路中的每一个核心链路进行单独压测；第四测试单元用于联调所有所述测试流量对所述目标链路中的每一个核心链路进行单独压测。
在某些实施例，例如本实施例中，所述第二测试单元包括第五测试单元和第六测试单元。
其中，第五测试单元用于联调部分所述测试流量对所述目标链路进行整体压测；第六测试单元用于联调所有所述测试流量对所述目标链路进行整体压测。
在某些实施例，例如本实施例中，所述第一实时采集单元150包括第二实时采集单元和第一获取单元。
其中，第二实时采集单元用于实时采集所述下游系统中的下游接口的调用成功率和时延指标；第一获取单元用于获取预设调用成功率可容忍异常阈值和预设时延可容忍异常阈值，根据所述调用成功率和所述调用成功率可容忍异常阈值的大小以及所述时延指标和所述预设时延可容忍异常阈值的大小判断是否 中断所述压测。
在某些实施例，例如本实施例中，所述预设调用成功率可容忍异常阈值包括调用成功率第一阈值和调用成功率第二阈值，所述预设时延可容忍异常阈值包括时延第一阈值和时延第二阈值，所述第一获取单元包括第一处理单元、第一确认单元和第二处理单元。
其中，第一处理单元用于若所述调用成功率小于所述调用成功率第一阈值且大于所述调用成功率第二阈值，并且所述时延指标大于所述时延第一阈值且小于所述时延第二阈值，则降低所述压测流量以降低所述下游系统的压力；第一确认单元用于若所述调用成功率小于所述调用成功率第二阈值，且所述时延指标大于所述时延第二阈值，则确认所述下游接口所对应的核心链路；第二处理单元用于关闭所述核心链路中预设的压测开关以避免所述压测流量进入所述核心链。
本发明另一实施例还提供了一种Radius环境下的全链路压测装置，本实施例的Radius环境下的全链路压测装置是在上述实施例的基础上增加了第二获取单元。
其中，第二获取单元用于获取预设压测流量，并对所述预设压测流量进行染色以获得所述压测流量。
本发明另一实施例还提供了一种Radius环境下的全链路压测装置，本实施例的Radius环境下的全链路压测装置是在上述实施例的基础上增加了第二实时监测单元和第二实时监测单元。
其中，第二实时监测单元用于实时监测在压测过程中所涉及的所有硬件设备的性能以获得所述硬件设备的性能指标数据；第二实时监测单元用于实时监测在所述压测过程中各个业务节点处理所述测试流量的耗时以获得时间指标。
需要说明的是，所属领域的技术人员可以清楚地了解到，上述Radius环境下的全链路压测装置和各单元的具体实现过程，可以参考前述方法实施例中的相应描述，为了描述的方便和简洁，在此不再赘述。
上述Radius环境下的全链路压测装置可以实现为一种计算机程序的形式，该计算机程序可以在如图4所示的计算机设备上运行。
请参阅图4，图4是本申请实施例提供的一种计算机设备的示意性框图。参阅图4，该计算机设备500包括通过系统总线501连接的处理器502、存储器和接口507，其中，存储器可以包括非易失性存储介质503和内存储器504。
该非易失性存储介质503可存储操作系统5031和计算机程序5032。该计算机程序5032被执行时，可使得处理器502执行一种Radius环境下的全链路压测方法。
该处理器502用于提供计算和控制能力，以支撑整个计算机设备500的运行。
该内存储器504为非易失性存储介质503中的计算机程序5032的运行提供环境，该计算机程序5032被处理器502执行时，可使得处理器502执行上述Radius环境下的全链路压测方法的任一实施例。
该接口505用于与其它设备进行通信。本领域技术人员可以理解，图4中示出的结构，仅仅是与本申请方案相关的部分结构的框图，并不构成对本申请方案所应用于其上的计算机设备500的限定，具体的计算机设备500可以包括比图中所示更多或更少的部件，或者组合某些部件，或者具有不同的部件布置。
其中，所述处理器502用于运行存储在存储器中的计算机程序5032，以实现如下步骤：
根据预设压测目标确认需要进行压测的目标链路，其中，所述目标链路包括多个不同业务线的核心链路；
创建正常业务表库的镜像表库以获得所述正常业务表库的影子表库；
联调测试流量对所述目标链路进行所述压测，其中，所述测试流量包括正常流量和压测流量；
将在压测过程中由所述正常流量所产生的正常数据存入所述正常业务表库，将由所述压测流量所产生的压测数据存入所述影子表库；
实时采集所述目标链路中的下游系统的性能数据，并根据所述性能数据判断是否中断所述压测。
在一实施例中，处理器502在实现所述联调测试流量对所述目标链路进行所述压测的步骤时，具体实现如下步骤：
联调所述测试流量对所述目标链路中的每一个核心链路进行单独压测；
联调所述测试流量对所述目标链路进行整体压测。
在一实施例中，处理器502在实现所述联调所述测试流量对所述目标链路中的每一个核心链路进行单独压测的步骤时，具体实现如下步骤：
联调部分所述测试流量对所述目标链路中的每一个核心链路进行单独压测；
联调所有所述测试流量对所述目标链路中的每一个核心链路进行单独压测。
在一实施例中，处理器502在实现所述联调所述测试流量对所述目标链路进行整体压测的步骤时，具体实现如下步骤：
联调部分所述测试流量对所述目标链路进行整体压测；
联调所有所述测试流量对所述目标链路进行整体压测。
在一实施例中，处理器502在实现所述实时采集所述目标链路中的下游系统的性能数据，并根据所述性能数据判断是否中断所述压测的步骤时，具体实现如下步骤：
实时采集所述下游系统中的下游接口的调用成功率和时延指标；
获取预设调用成功率可容忍异常阈值和预设时延可容忍异常阈值，根据所述调用成功率和所述调用成功率可容忍异常阈值的大小以及所述时延指标和所述预设时延可容忍异常阈值的大小判断是否中断所述压测。
在一实施例中，处理器502在实现所述根据所述调用成功率和所述调用成功率可容忍异常阈值的大小以及所述时延指标和所述预设时延可容忍异常阈值的大小判断是否中断所述压测的步骤时，具体实现如下步骤：
若所述调用成功率小于所述调用成功率第一阈值且大于所述调用成功率第二阈值，并且所述时延指标大于所述时延第一阈值且小于所述时延第二阈值，则降低所述压测流量以降低所述下游系统的压力；
若所述调用成功率小于所述调用成功率第二阈值，且所述时延指标大于所述时延第二阈值，则确认所述下游接口所对应的核心链路；
关闭所述核心链路中预设的压测开关以避免所述压测流量进入所述核心链路。
在一实施例中，所述处理器502还实现如下步骤：
获取预设压测流量，并对所述预设压测流量进行染色以获得所述压测流量。
在一实施例中，所述处理器502还实现如下步骤：
实时监测在压测过程中所涉及的所有硬件设备的性能以获得所述硬件设备的性能指标数据；
实时监测在所述压测过程中各个业务节点处理所述测试流量的耗时以获得时间指标。
应当理解，在本申请实施例中，处理器502可以是中央处理单元(Central Processing Unit，CPU)，该处理器502还可以是其他通用处理器、数字信号处理器(Figital Signal Processor，FSP)、专用集成电路(Application Specific IntegrateF Circuit，ASIC)、现成可编程门阵列(FielF-Programmable Gate Array，FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件等。其中，通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。
本领域普通技术人员可以理解的是实现上述实施例的方法中的全部或部分流程，是可以通过计算机程序来指令相关的硬件来完成。该计算机程序可存储于一存储介质中，该存储介质为计算机可读存储介质。该计算机程序被该计算机系统中的至少一个处理器执行，以实现上述方法的实施例的流程步骤。
因此，本发明还提供一种存储介质。该存储介质可以为计算机可读存储介质。该存储介质存储有计算机程序。该计算机程序当被处理器执行时实现上述Radius环境下的全链路压测方法的任一实施例。
所述存储介质可以是U盘、移动硬盘、只读存储器(ReaF-Only Memory，ROM)、磁碟或者光盘等各种可以存储程序代码的计算机可读存储介质。
本领域普通技术人员可以意识到，结合本文中所公开的实施例描述的各示例的单元及算法步骤，能够以电子硬件、计算机软件或者二者的结合来实现，为了清楚地说明硬件和软件的可互换性，在上述说明中已经按照功能一般性地描述了各示例的组成及步骤。这些功能究竟以硬件还是软件方式来执行，取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能，但是这种实现不应认为超出本发明的范围。
在本发明所提供的几个实施例中，应该理解到，所揭露的装置和方法，可以通过其它的方式实现。例如，以上所描述的装置实施例仅仅是示意性的。例 如，各个单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式。例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。
本发明实施例方法中的步骤可以根据实际需要进行顺序调整、合并和删减。本发明实施例装置中的单元可以根据实际需要进行合并、划分和删减。另外，在本发明各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以是两个或两个以上单元集成在一个单元中。
该集成的单元如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个存储介质中。基于这样的理解，本发明的技术方案本质上或者说对现有技术做出贡献的部分，或者该技术方案的全部或部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质中，包括若干指令用以使得一台计算机设备执行本发明各个实施例所述方法的全部或部分步骤。
在上述实施例中，对各个实施例的描述都各有侧重，某个实施例中没有详细描述的部分，可以参见其他实施例的相关描述。
显然，本领域的技术人员可以对本发明进行各种改动和变型而不脱离本发明的精神和范围。这样，尚且本发明的这些修改和变型属于本发明权利要求及其等同技术的范围之内，则本发明也意图包含这些改动和变型在内。
以上所述，仅为本发明的具体实施方式，但本发明的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本发明揭露的技术范围内，可轻易想到各种等效的修改或替换，这些修改或替换都应涵盖在本发明的保护范围之内。因此，本发明的保护范围应以权利要求的保护范围为准。

Claims (10)

  1. 一种Radius环境下的全链路压测方法，其特征在于，所述方法包括：
    根据预设压测目标确认需要进行压测的目标链路，其中，所述目标链路包括多个不同业务线的核心链路；
    创建正常业务表库的镜像表库以获得所述正常业务表库的影子表库；
    联调测试流量对所述目标链路进行所述压测，其中，所述测试流量包括正常流量和压测流量；
    将在压测过程中由所述正常流量所产生的正常数据存入所述正常业务表库，将由所述压测流量所产生的压测数据存入所述影子表库；
    实时采集所述目标链路中的下游系统的性能数据，并根据所述性能数据判断是否中断所述压测。
  2. 如权利要求1所述的Radius环境下的全链路压测方法，其特征在于，所述实时采集所述目标链路中的下游系统的性能数据，并根据所述性能数据判断是否中断所述压测的步骤，包括：
    实时采集所述下游系统中的下游接口的调用成功率和时延指标；
    获取预设调用成功率可容忍异常阈值和预设时延可容忍异常阈值，根据所述调用成功率和所述调用成功率可容忍异常阈值的大小以及所述时延指标和所述预设时延可容忍异常阈值的大小判断是否中断所述压测。
  3. 如权利要求2所述的Radius环境下的全链路压测方法，其特征在于，所述预设调用成功率可容忍异常阈值包括调用成功率第一阈值和调用成功率第二阈值，所述预设时延可容忍异常阈值包括时延第一阈值和时延第二阈值；
    所述根据所述调用成功率和所述调用成功率可容忍异常阈值的大小以及所述时延指标和所述预设时延可容忍异常阈值的大小判断是否中断所述压测的步骤，包括：
    若所述调用成功率小于所述调用成功率第一阈值且大于所述调用成功率第二阈值，并且所述时延指标大于所述时延第一阈值且小于所述时延第二阈值，则降低所述压测流量以降低所述下游系统的压力；
    若所述调用成功率小于所述调用成功率第二阈值，且所述时延指标大于所述时延第二阈值，则确认所述下游接口所对应的核心链路；
    关闭所述核心链路中预设的压测开关以避免所述压测流量进入所述核心链路。
  4. 如权利要求1所述的Radius环境下的全链路压测方法，其特征在于，所述方法还包括：
    获取预设压测流量，并对所述预设压测流量进行染色以获得所述压测流量。
  5. 如权利要求1所述的Radius环境下的全链路压测方法，其特征在于，所述联调测试流量对所述目标链路进行所述压测的步骤，包括：
    联调所述测试流量对所述目标链路中的每一个核心链路进行单独压测；
    联调所述测试流量对所述目标链路进行整体压测。
  6. 如权利要求5所述的Radius环境下的全链路压测方法，其特征在于， 所述联调所述测试流量对所述目标链路中的每一个核心链路进行单独压测的步骤，包括：
    联调部分所述测试流量对所述目标链路中的每一个核心链路进行单独压测；
    联调所有所述测试流量对所述目标链路中的每一个核心链路进行单独压测；
    所述联调所述测试流量对所述目标链路进行整体压测的步骤，包括：
    联调部分所述测试流量对所述目标链路进行整体压测；
    联调所有所述测试流量对所述目标链路进行整体压测。
  7. 如权利要求1所述的Radius环境下的全链路压测方法，其特征在于，所述方法还包括：
    实时监测在压测过程中所涉及的所有硬件设备的性能以获得所述硬件设备的性能指标数据；
    实时监测在所述压测过程中各个业务节点处理所述测试流量的耗时以获得时间指标。
  8. 一种Radius环境下的全链路压测装置，其特征在于，所述装置包括：
    目标链路确认单元，用于根据预设压测目标确认需要进行压测的目标链路，其中，所述目标链路包括多个不同业务线的核心链路；
    第一创建单元，用于创建正常业务表库的镜像表库以获得所述正常业务表库的影子表库；
    第一测试单元，用于联调测试流量对所述目标链路进行所述压测，其中，所述测试流量包括正常流量和压测流量；
    第一存储单元，用于将在压测过程中由所述正常流量所产生的正常数据存入所述正常业务表库，将由所述压测流量所产生的压测数据存入所述影子表库；
    第一实时采集单元，用于实时采集所述目标链路中的下游系统的性能数据，并根据所述性能数据判断是否中断所述压测。
  9. 一种计算机设备，其特征在于，所述计算机设备包括存储器以及与所述存储器相连的处理器；所述存储器用于存储计算机程序；所述处理器用于运行所述存储器中存储的计算机程序，以执行如权利要求1-7任一项所述方法的步骤。
  10. 一种计算机可读存储介质，其特征在于，所述存储介质存储有计算机程序，所述计算机程序被处理器执行时可于计算机设备上实现如权利要求1-7中任一项所述方法的步骤。
PCT/CN2022/141328 2021-12-29 2022-12-23 Radius环境下的全链路压测方法、装置、计算机设备及存储介质 WO2023125272A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202111636521.2 2021-12-29
CN202111636521.2A CN114338447B (zh) 2021-12-29 2021-12-29 Radius环境下的全链路压测方法、装置、计算机设备及存储介质

Publications (1)

Publication Number Publication Date
WO2023125272A1 true WO2023125272A1 (zh) 2023-07-06

Family

ID=81017586

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/141328 WO2023125272A1 (zh) 2021-12-29 2022-12-23 Radius环境下的全链路压测方法、装置、计算机设备及存储介质

Country Status (2)

Country Link
CN (1) CN114338447B (zh)
WO (1) WO2023125272A1 (zh)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114338447B (zh) * 2021-12-29 2024-07-16 天翼物联科技有限公司 Radius环境下的全链路压测方法、装置、计算机设备及存储介质
CN115065612B (zh) * 2022-06-22 2024-03-08 上海哔哩哔哩科技有限公司 全链路压测改造的测试方法及装置
CN115695241A (zh) * 2022-10-21 2023-02-03 深圳市正浩创新科技股份有限公司 通讯压力测试方法及电子设备
CN116781591B (zh) * 2023-08-23 2024-01-02 国网信息通信产业集团有限公司 一种基于流量染色和透传技术的压力测试系统、压测终端及存储介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111723017A (zh) * 2020-06-28 2020-09-29 深圳壹账通智能科技有限公司 系统容量测试方法、装置、计算机设备及存储介质
CN113254323A (zh) * 2021-07-05 2021-08-13 中邮消费金融有限公司 线上全链路压测方法、装置及计算机设备
CN113747506A (zh) * 2020-05-28 2021-12-03 华为技术有限公司 一种资源调度方法、装置和网络系统
CN114338447A (zh) * 2021-12-29 2022-04-12 天翼物联科技有限公司 Radius环境下的全链路压测方法、装置、计算机设备及存储介质

Family Cites Families (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109600280B (zh) * 2018-12-17 2021-02-02 北京车和家信息技术有限公司 压力测试系统、方法、装置及计算机可读存储介质
US10877864B2 (en) * 2018-12-20 2020-12-29 International Business Machines Corporation Controlling segment layout in a stress test for a processor memory with a link stack
CN110532168B (zh) * 2019-07-05 2024-03-29 中国平安财产保险股份有限公司 全链路接口压测方法、装置、计算机设备及存储介质
CN110958159B (zh) * 2019-11-22 2021-05-11 英业达科技有限公司 自适应网络压测的配置方法、系统、可读存储介质及设备
CN110912786B (zh) * 2019-12-27 2021-07-16 深圳市星砺达科技有限公司 网关压力测试方法、装置、计算机设备及存储介质
CN111526063B (zh) * 2020-03-04 2022-11-11 平安科技（深圳）有限公司 基于整体业务链路断点监控方法、装置、终端及存储介质
CN111966556A (zh) * 2020-08-18 2020-11-20 腾讯音乐娱乐科技（深圳）有限公司 性能压测方法、装置及服务器和计算机可读存储介质
CN112583669B (zh) * 2020-12-11 2022-06-24 北京字节跳动网络技术有限公司 线上链路压力测试方法、装置、电子设备及介质
CN113381909B (zh) * 2021-06-08 2023-04-07 中国工商银行股份有限公司 全链路压测方法及装置
CN113553267B (zh) * 2021-07-22 2024-08-13 招商银行股份有限公司 应用性能测试方法、设备、介质及计算机程序产品

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113747506A (zh) * 2020-05-28 2021-12-03 华为技术有限公司 一种资源调度方法、装置和网络系统
CN111723017A (zh) * 2020-06-28 2020-09-29 深圳壹账通智能科技有限公司 系统容量测试方法、装置、计算机设备及存储介质
CN113254323A (zh) * 2021-07-05 2021-08-13 中邮消费金融有限公司 线上全链路压测方法、装置及计算机设备
CN114338447A (zh) * 2021-12-29 2022-04-12 天翼物联科技有限公司 Radius环境下的全链路压测方法、装置、计算机设备及存储介质

Also Published As

Publication number Publication date
CN114338447A (zh) 2022-04-12
CN114338447B (zh) 2024-07-16

Similar Documents

Publication Publication Date Title
WO2023125272A1 (zh) Radius环境下的全链路压测方法、装置、计算机设备及存储介质
JP2020123951A (ja) 光モジュールの故障を予測するための方法、装置、およびデバイス
CN104584483A (zh) 用于自动确定服务质量降级的原因的方法和设备
CN112231187B (zh) 微服务异常分析方法及装置
US20200410376A1 (en) Prediction method, training method, apparatus, and computer storage medium
CN110221953A (zh) 测试结果分析方法、装置、服务器及存储介质
CN110300008A (zh) 一种确定网络设备的状态的方法及装置
CN116414717A (zh) 基于流量回放的自动测试方法、装置、设备、介质及产品
CN116208535A (zh) 一种基于大数据分析的计算机数据传输控制系统
CN108920601B (zh) 一种数据匹配方法及装置
JPWO2015182629A1 (ja) 監視システム、監視装置及び監視プログラム
CN112948262A (zh) 一种系统测试方法、装置、计算机设备和存储介质
CN112769642A (zh) 一种服务器pcie链路的测试方法及相关设备
CN111654405A (zh) 通信链路的故障节点方法、装置、设备及存储介质
CN111159009A (zh) 一种日志服务系统的压力测试方法及装置
CN116226644A (zh) 设备故障类型的确定方法、装置、电子设备及存储介质
KR20230016896A (ko) 통신 시스템에서 응용 서비스 응답 시간 예측 방법 및 장치
CN114281691A (zh) 测试用例排序方法、装置、计算设备及存储介质
CN114385398A (zh) 一种请求响应状态确定方法、装置、设备和存储介质
US20210208998A1 (en) Function analyzer, function analysis method, and function analysis program
CN117251769B (zh) 基于监控组件的异常数据识别方法、装置、设备及介质
CN116094908B (zh) 日志文件存储方法、装置、设备和存储介质
CN116882724B (zh) 一种业务流程优化方案的生成方法、装置、设备及介质
CN117056663B (zh) 一种数据处理方法、装置、电子设备及存储介质
WO2024114567A1 (zh) 一种网络问题分析方法以及相关装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22914526

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE