CN113656270B - Method, device, medium and computer program product for testing application performance - Google Patents

Method, device, medium and computer program product for testing application performance Download PDF

Info

Publication number
CN113656270B
CN113656270B CN202110879239.0A CN202110879239A CN113656270B CN 113656270 B CN113656270 B CN 113656270B CN 202110879239 A CN202110879239 A CN 202110879239A CN 113656270 B CN113656270 B CN 113656270B
Authority
CN
China
Prior art keywords
similarity
performance
application
curve
fluctuation
Prior art date
Legal status (The legal status 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 status listed.)
Active
Application number
CN202110879239.0A
Other languages
Chinese (zh)
Other versions
CN113656270A (en
Inventor
许治华
陈运权
金清华
罗伟涌
陈睿扬
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
China Merchants Bank Co Ltd
Original Assignee
China Merchants Bank Co Ltd
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 China Merchants Bank Co Ltd filed Critical China Merchants Bank Co Ltd
Priority to CN202110879239.0A priority Critical patent/CN113656270B/en
Publication of CN113656270A publication Critical patent/CN113656270A/en
Application granted granted Critical
Publication of CN113656270B publication Critical patent/CN113656270B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
    • G06F11/3409—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment for performance assessment
    • G06F11/3414—Workload generation, e.g. scripts, playback
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
    • G06F11/3452—Performance evaluation by statistical analysis
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Abstract

The application discloses a method, equipment, medium and computer program product for testing application performance, which are used for acquiring a node monitoring index curve of an application to be tested, and performing similarity calculation on the node monitoring index curve to obtain first similarity information; determining whether the application to be tested has performance fluctuation or not based on the first similarity information; if the performance fluctuation exists in the application to be tested, determining an abnormal node where the performance fluctuation is located according to the first similarity information; acquiring an index data curve of the abnormal node, and performing similarity calculation on the index data curve to obtain second similarity information; determining a performance bottleneck point in the abnormal node according to the second similarity information; and determining the performance fluctuation root cause of the application to be tested according to the performance bottleneck point. The method and the device can rapidly and accurately position the performance fluctuation root cause of the application to be tested, reduce the skill requirements on application related personnel, have higher accuracy in the aspect of positioning the performance fluctuation problem of the application, and improve the efficiency of application performance test.

Description

Method, device, medium and computer program product for testing application performance
Technical Field
The present invention relates to the field of communications technologies, and in particular, to a method, an apparatus, a medium, and a computer program product for testing application performance.
Background
With the development of internet technology, especially mobile internet technology, the user quantity and access quantity of computer software systems are exponentially increased, and the non-functional requirements on the reliability, stability, compatibility and the like of the software are increasingly urgent. In order to meet the nonfunctional requirements of the system, the architecture of the software system is gradually changed from the simple single application to the complex architecture of distributed and micro-service, and the like, and great challenges are brought to the analysis of the performance fluctuation of the software system. The existing solution to the problem of fluctuation of the performance of the application software depends on manual experience judgment, has no definite rule guidance and system support, cannot accurately cover the root cause of the fluctuation of the performance of the software, and has lower accuracy of the performance test of the application at present.
Disclosure of Invention
The main purpose of the application is to provide a method, a device, a medium and a computer program product for testing application performance, which aim to solve the technical problem of lower accuracy of performance testing of the application at present.
In order to achieve the above object, an embodiment of the present application provides a method for testing application performance, where the method for testing application performance includes:
acquiring a node monitoring index curve of an application to be tested, and performing similarity calculation on the node monitoring index curve to obtain first similarity information;
determining whether performance fluctuation exists in the application to be tested or not based on the first similarity information;
if the performance fluctuation exists in the application to be tested, determining an abnormal node where the performance fluctuation exists according to the first similarity information;
acquiring an index data curve of the abnormal node, and performing similarity calculation on the index data curve to obtain second similarity information;
determining a performance bottleneck point in the abnormal node according to the second similarity information;
and determining the performance fluctuation root cause of the application to be tested according to the performance bottleneck point.
Preferably, the step of calculating the similarity of the node monitoring indicator curve to obtain first similarity information includes:
carrying out quantization treatment on the node monitoring index curve to obtain a node monitoring index curve after the quantization treatment;
determining a concurrent pressure curve of a source in the node monitoring index curve and a performance gold index as a reference curve pair;
And respectively carrying out similarity calculation on the monitoring index curves of other nodes in the node monitoring index curves and the reference curve pair to obtain a plurality of first similarities and form first similarity information.
Preferably, the step of determining whether the application to be tested has performance fluctuation based on the first similarity information includes:
comparing each first similarity in the first similarity information with a first preset similarity threshold value respectively;
and if the intermediate similarity smaller than the first preset similarity threshold exists in each first similarity, determining that the application to be tested has performance fluctuation.
Preferably, the step of determining the abnormal node where the performance fluctuation is located according to the first similarity information includes:
obtaining each intermediate similarity smaller than the first preset similarity threshold in the first similarity information;
performing numerical comparison on each intermediate similarity to determine a first target similarity with the smallest numerical value in each intermediate similarity;
and determining the node corresponding to the first target similarity as an abnormal node where the performance fluctuation is located.
Preferably, the step of determining a performance bottleneck point in the abnormal node according to the second similarity information includes:
Performing numerical comparison on each second similarity in the second similarity information to determine a second target similarity with the smallest numerical value in each second similarity;
and determining the index item corresponding to the second target similarity as a performance bottleneck point in the abnormal node.
Preferably, the step of determining the root cause of the performance fluctuation of the application to be tested according to the performance bottleneck point includes:
obtaining a bottleneck index data curve of the performance bottleneck point, and performing similarity calculation on the bottleneck index data curve and a preset characteristic model to obtain a third similarity;
comparing the third similarity with a second preset similarity threshold;
and if the third similarity is greater than or equal to the second preset similarity threshold, determining a performance fluctuation root cause of the application to be tested based on the third similarity.
Preferably, the step of determining a root cause of performance fluctuation of the application to be tested based on the third similarity comprises:
determining a preset feature model corresponding to the third similarity as a target feature model;
indexing a performance fluctuation root cause corresponding to the target feature model in a preset fluctuation model library;
And determining the performance fluctuation root cause corresponding to the target feature model as the performance fluctuation root cause of the application to be tested.
Further, in order to achieve the above object, the present application further provides an application performance test device, where the application performance test device includes a memory, a processor, and an application performance test program stored in the memory and executable on the processor, where the application performance test program is executed by the processor to implement the steps of the application performance test method.
Further, in order to achieve the above object, the present application further provides a storage medium, on which a test program for application performance is stored, the test program for application performance implementing the steps of the above test method for application performance when executed by a processor.
Further, to achieve the above object, the present application further provides a computer program product, which includes a computer program, and the computer program when executed by a processor implements the steps of the method for testing application performance described above.
The embodiment of the application provides a method, equipment, medium and computer program product for testing application performance, which are used for acquiring a node monitoring index curve of an application to be tested, and performing similarity calculation on the node monitoring index curve to obtain first similarity information; determining whether performance fluctuation exists in the application to be tested or not based on the first similarity information; if the performance fluctuation exists in the application to be tested, determining an abnormal node where the performance fluctuation exists according to the first similarity information; acquiring an index data curve of the abnormal node, and performing similarity calculation on the index data curve to obtain second similarity information; determining a performance bottleneck point in the abnormal node according to the second similarity information; and determining the performance fluctuation root cause of the application to be tested according to the performance bottleneck point. According to the method and the device, whether the performance fluctuation exists in the application to be tested is automatically and accurately determined by carrying out similarity calculation on the node monitoring index curve of the application to be tested, when the performance fluctuation exists in the application to be tested, the abnormal node where the performance fluctuation exists is determined, then the performance bottleneck point in the abnormal node is determined, finally, the root cause of the performance fluctuation of the application to be tested is rapidly and accurately positioned through similarity matching between the curves, the skill requirements of relevant personnel of the application are reduced, the accuracy is higher in the aspect of positioning the performance fluctuation problem of the application, and the efficiency of the application performance test is effectively improved.
Drawings
FIG. 1 is a schematic structural diagram of a hardware operating environment related to an embodiment of a test method for application performance of the present application;
FIG. 2 is a flowchart of a first embodiment of a testing method for application performance of the present application;
FIG. 3 is a flow chart of a second embodiment of a testing method for application performance of the present application;
FIG. 4 is a flow chart of a third embodiment of a testing method for application performance of the present application;
FIG. 5 is a flowchart of a fourth embodiment of a testing method for application performance of the present application;
FIG. 6 is a flowchart of a fifth embodiment of a testing method for application performance of the present application;
fig. 7 is a flowchart of a sixth embodiment of a testing method for application performance of the present application.
The realization, functional characteristics and advantages of the present application will be further described with reference to the embodiments, referring to the attached drawings.
Detailed Description
It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the present application.
The embodiment of the application provides a method, equipment, medium and computer program product for testing application performance, which are used for acquiring a node monitoring index curve of an application to be tested, and performing similarity calculation on the node monitoring index curve to obtain first similarity information; determining whether performance fluctuation exists in the application to be tested or not based on the first similarity information; if the performance fluctuation exists in the application to be tested, determining an abnormal node where the performance fluctuation exists according to the first similarity information; acquiring an index data curve of the abnormal node, and performing similarity calculation on the index data curve to obtain second similarity information; determining a performance bottleneck point in the abnormal node according to the second similarity information; and determining the performance fluctuation root cause of the application to be tested according to the performance bottleneck point. According to the method and the device, whether the performance fluctuation exists in the application to be tested is automatically and accurately determined by carrying out similarity calculation on the node monitoring index curve of the application to be tested, when the performance fluctuation exists in the application to be tested, the abnormal node where the performance fluctuation exists is determined, then the performance bottleneck point in the abnormal node is determined, finally, the root cause of the performance fluctuation of the application to be tested is rapidly and accurately positioned through similarity matching between the curves, the skill requirements of relevant personnel of the application are reduced, the accuracy is higher in the aspect of positioning the performance fluctuation problem of the application, and the efficiency of the application performance test is effectively improved.
As shown in fig. 1, fig. 1 is a schematic structural diagram of a test device for application performance of a hardware running environment according to an embodiment of the present application.
In the following description, suffixes such as "module", "component", or "unit" for representing elements are used only for facilitating the description of the present application, and are not of specific significance per se. Thus, "module," "component," or "unit" may be used in combination.
The test device for application performance of the embodiment of the application can be a PC, or can be a mobile terminal device such as a tablet computer, a portable computer and the like.
As shown in fig. 1, the test apparatus for application performance may include: a processor 1001, such as a CPU, a network interface 1004, a user interface 1003, a memory 1005, a communication bus 1002. Wherein the communication bus 1002 is used to enable connected communication between these components. The user interface 1003 may include a Display, an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may further include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a wireless interface (e.g., WI-FI interface). The memory 1005 may be a high-speed RAM memory or a stable memory (non-volatile memory), such as a disk memory. The memory 1005 may also optionally be a storage device separate from the processor 1001 described above.
Those skilled in the art will appreciate that the application performance testing apparatus structure shown in fig. 1 does not constitute a limitation of the application performance testing apparatus, and may include more or fewer components than shown, or may combine certain components, or may be arranged in different components.
As shown in fig. 1, a memory 1005, which is a storage medium, may include an operating system, a network communication module, a user interface module, and a test program for application performance.
In the device shown in fig. 1, the network interface 1004 is mainly used for connecting to a background server, and performing data communication with the background server; the user interface 1003 is mainly used for connecting a client (user side) and performing data communication with the client; and the processor 1001 may be configured to call a test program for application performance stored in the memory 1005, and perform the following operations:
acquiring a node monitoring index curve of an application to be tested, and performing similarity calculation on the node monitoring index curve to obtain first similarity information;
determining whether performance fluctuation exists in the application to be tested or not based on the first similarity information;
if the performance fluctuation exists in the application to be tested, determining an abnormal node where the performance fluctuation exists according to the first similarity information;
Acquiring an index data curve of the abnormal node, and performing similarity calculation on the index data curve to obtain second similarity information;
determining a performance bottleneck point in the abnormal node according to the second similarity information;
and determining the performance fluctuation root cause of the application to be tested according to the performance bottleneck point.
Further, the step of performing similarity calculation on the node monitoring index curve to obtain first similarity information includes:
carrying out quantization treatment on the node monitoring index curve to obtain a node monitoring index curve after the quantization treatment;
determining a concurrent pressure curve of a source in the node monitoring index curve and a performance gold index as a reference curve pair;
and respectively carrying out similarity calculation on the monitoring index curves of other nodes in the node monitoring index curves and the reference curve pair to obtain a plurality of first similarities and form first similarity information.
Further, the step of determining whether the application to be tested has performance fluctuation based on the first similarity information includes:
comparing each first similarity in the first similarity information with a first preset similarity threshold value respectively;
And if the intermediate similarity smaller than the first preset similarity threshold exists in each first similarity, determining that the application to be tested has performance fluctuation.
Further, the step of determining the abnormal node where the performance fluctuation is located according to the first similarity information includes:
obtaining each intermediate similarity smaller than the first preset similarity threshold in the first similarity information;
performing numerical comparison on each intermediate similarity to determine a first target similarity with the smallest numerical value in each intermediate similarity;
and determining the node corresponding to the first target similarity as an abnormal node where the performance fluctuation is located.
Further, the step of determining the performance bottleneck point in the abnormal node according to the second similarity information includes:
performing numerical comparison on each second similarity in the second similarity information to determine a second target similarity with the smallest numerical value in each second similarity;
and determining the index item corresponding to the second target similarity as a performance bottleneck point in the abnormal node.
Further, the step of determining the root cause of the performance fluctuation of the application to be tested according to the performance bottleneck point comprises the following steps:
Obtaining a bottleneck index data curve of the performance bottleneck point, and performing similarity calculation on the bottleneck index data curve and a preset characteristic model to obtain a third similarity;
comparing the third similarity with a second preset similarity threshold;
and if the third similarity is greater than or equal to the second preset similarity threshold, determining a performance fluctuation root cause of the application to be tested based on the third similarity.
Further, the step of determining a root cause of the performance fluctuation of the application to be tested based on the third similarity includes:
determining a preset feature model corresponding to the third similarity as a target feature model;
indexing a performance fluctuation root cause corresponding to the target feature model in a preset fluctuation model library;
and determining the performance fluctuation root cause corresponding to the target feature model as the performance fluctuation root cause of the application to be tested.
In order that the above-described aspects may be better understood, exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
In order to better understand the above technical solutions, the following detailed description will refer to the accompanying drawings and specific embodiments.
Referring to fig. 2, a first embodiment of the present application provides a flowchart of a testing method for application performance. In this embodiment, the method for testing application performance includes the following steps:
step S10, acquiring a node monitoring index curve of an application to be tested, and performing similarity calculation on the node monitoring index curve to obtain first similarity information;
the method for testing the application performance is applied to a performance fluctuation analysis system based on graph trend matching, and the performance fluctuation analysis system based on graph trend matching in the embodiment can be deployed on a server or an intelligent terminal to realize the function of testing the application performance. It will be appreciated that for convenience of description, the performance fluctuation analysis system based on pattern trend matching will be hereinafter simply referred to as a system.
It will be appreciated that from the theory of energy conduction, energy is conducted from the centre of energy to the periphery along the path along which the energy can be transferred during the diffusion process, there is some similarity between waveforms on the path along which the energy can be transferred, and obstruction during the conduction process can interfere with the conduction waveform, resulting in reduced similarity between the waveform and the source waveform during the conduction process. Similarly, in the performance pressure measurement scenario of an application, business transaction pressures are conducted back from the source along the critical path of the transaction. If a certain node on the critical path has a performance bottleneck, pressure conduction abnormality is caused, and then the similarity between the index curve of the abnormal node and the concurrent pressure curve and the gold index curve of the pressure source is reduced. Based on the theory, the method starts from the source of concurrent pressure generation, along the business transaction critical path, the association relation between the monitoring index curves on the pressure conduction path is analyzed from front to back, the abnormal node is determined, the reason for generating the fluctuation of the positioning performance is determined based on the abnormal node, the method has certain prejudgement, and the method can assist software testing, development and operation and maintenance personnel to quickly position the performance bottleneck of application.
Further, after software development is completed to obtain a software application, in order to ensure that the application has better performance when being online so as to improve the use experience of a user, performance test is required to be performed on the application, whether the application has a performance bottleneck is determined, and if the performance bottleneck does not exist, the application can be online; if the application is found to have the performance bottleneck after the test, the performance bottleneck generated by the application is required to be positioned to determine the corresponding performance problem, then an application performance optimization scheme is formed aiming at the performance problem, and finally the application performance optimization scheme is pushed to a performance tester and a developer for reference, so that the performance tester and the developer can optimize the application.
Specifically, for an application (i.e., an application to be tested) that needs performance test, the application to be tested is deployed and run on a server, and then a scene simulation is performed on a running target application, so that the performance of the target application is tested according to running information of the target application in the scene. For example: and deploying and running a shopping application on the server, and simulating a business pressure environment generated by operations such as shopping and order submitting by a plurality of people to test whether the application to be tested has a performance bottleneck in the environment.
Further, the system acquires a node monitoring index curve of an application to be tested, wherein the node monitoring index curve comprises a concurrent pressure curve and a performance gold index curve of a source on a pressure transmission critical path, and a server resource consumption curve and a response time trend curve of other nodes on the pressure transmission critical path, and the concurrent pressure curve is a curve formed by concurrent pressures within a certain time (the concurrent pressures can be understood as pressures generated when a certain number of virtual users use the application in the embodiment); the performance gold index curve can be a curve formed by indexes of flow rate, error (success) rate, response time and capacity; the server resource consumption curve is a curve formed by resources consumed by the server in a certain time; the response time curve is a curve formed by the time required for responding to a certain operation. Specifically, the system can acquire a concurrent pressure curve and a performance gold index curve of a source on a pressure transmission critical path and a server resource consumption curve and a response time trend curve of other nodes on the pressure transmission critical path of the application to be tested in a simulated environment, so as to acquire a node monitoring index curve of the application to be tested.
Further, the system carries out quantization processing on the node monitoring index curve, then determines a concurrent pressure curve of a source and a performance gold index in the node monitoring index curve after the quantization processing as a reference curve pair, wherein the reference curve pair is a curve pair serving as a reference and is used for comparing with other curves, and the quantization processing is used for converting the node monitoring index curve into a data form capable of being calculated. Further, the system calculates the similarity between the monitoring index curves of other nodes in the node monitoring index curves and the concurrent pressure curve and the performance gold index serving as the reference curve pair respectively to obtain a plurality of first similarities, and forms first similarity information by the plurality of first similarities, wherein the calculation of the similarity between the monitoring index curves and the concurrent pressure curve and the performance gold index serving as the reference curve pair in the embodiment specifically calculates a Hamming Distance (Hamming Distance) between the two curves, specifically performs a difference operation on fluctuation characteristics of the two curves, so that Hamming Distance between the two curves is obtained, and therefore quantization processing is needed to be performed on the node monitoring index curves to obtain fluctuation characteristics of each curve. By calculating the similarity between curves of all nodes (including sources) of the application to be tested and forming first similarity information, whether performance fluctuation exists in the application to be tested or not can be conveniently and subsequently determined according to the first similarity information, and performance bottleneck points can be rapidly positioned based on the first similarity information when the performance fluctuation exists in the application to be tested is determined, so that the efficiency of application performance testing is improved.
Step S20, determining whether performance fluctuation exists in the application to be tested or not based on the first similarity information;
after the first similarity information is obtained, the system determines whether the performance fluctuation exists in the application to be tested according to a plurality of first similarities contained in the first similarity information, specifically, the system detects whether intermediate similarities smaller than a first preset similarity threshold exist in the plurality of first similarities, and if the intermediate similarities exist, the system determines that the performance fluctuation exists in the application to be tested; conversely, if there is no intermediate similarity, it is determined that there is no performance fluctuation of the application to be tested. The first preset similarity threshold is a value that can be set based on the actual environment and the requirements. Whether the application to be tested has performance fluctuation or not can be rapidly and accurately determined through the first similarity information, and the efficiency of the application performance test can be improved.
Step S30, if the application to be tested has performance fluctuation, determining an abnormal node where the performance fluctuation is located according to the first similarity information;
further, if it is determined that the application to be tested has performance fluctuation, the system acquires all intermediate similarities in the first similarity information, which are smaller than a first preset similarity threshold, determines the intermediate similarity with the smallest value from the intermediate similarity thresholds as a first target similarity, and takes a node corresponding to the first target similarity as an abnormal node where the performance fluctuation is located. After the performance fluctuation of the application to be tested is determined, the abnormal node where the performance fluctuation is located can be rapidly and accurately determined through the first similarity information, so that the subsequent step-by-step drilling and positioning of the performance fluctuation generating root cause based on the abnormal node are facilitated, and the efficiency of the application performance test can be improved.
Step S40, obtaining an index data curve of the abnormal node, and performing similarity calculation on the index data curve to obtain second similarity information;
after detecting that the application to be tested has performance fluctuation and determining an abnormal node where the performance fluctuation is located, the system acquires an index data curve of the abnormal node, and specifically, can acquire curves such as a pressure measurement data fluctuation curve of the abnormal node at a press end, a resource consumption data curve of a server end, a performance data curve of an application program layer and the like as the index data curve of the abnormal node. Further, the system carries out quantization processing on the index data curve to obtain a quantized index data curve, carries out similarity calculation on the index data curve and a concurrent pressure curve and a performance gold index curve serving as a reference curve pair respectively to obtain a plurality of second similarities, and forms second similarity information by the plurality of second similarities. Specifically, the system calculates the similarity between the pressure measurement data fluctuation curve of the quantized abnormal node at the press end and the concurrent pressure curve and the performance golden index curve respectively, averages the two obtained similarities, and then calculates by combining the weight of the index item to obtain the second similarity of the index item; the quantized resource consumption data curve of the server is respectively calculated with the concurrent pressure curve and the performance gold index curve, the obtained two similarities are averaged, and then the calculation is carried out by combining the weight of the index item, so that the second similarity of the index item is obtained; and calculating the similarity between the quantized performance data of the application program layer and the concurrent pressure curve and the performance golden index curve, averaging the obtained two similarities, and then calculating by combining the weight of the index item to obtain the second similarity of the index item. And the similarity between the index data curve and the reference curve pair is calculated, and second similarity information is formed, so that the performance bottleneck point can be conveniently and quickly positioned according to the second similarity information, the performance fluctuation root cause of the application to be tested is determined according to the performance bottleneck point, and the efficiency of the application performance test is improved.
Step S50, determining a performance bottleneck point in the abnormal node according to the second similarity information;
after the second similarity information is obtained through calculation, the system determines the second similarity with the smallest value from the second similarity thresholds as second target similarity, and takes the index item corresponding to the second target similarity as a performance bottleneck point in the abnormal node. After the abnormal node where the performance fluctuation is located is determined, the performance bottleneck point in the abnormal node can be quickly and accurately drilled and positioned through the second similarity information, the root cause generated by the performance fluctuation can be conveniently and subsequently determined based on the performance bottleneck point, and the efficiency of the application performance test can be improved.
And step S60, determining the performance fluctuation root cause of the application to be tested according to the performance bottleneck point.
After determining the performance bottleneck point, the system firstly acquires a bottleneck index data curve of the performance bottleneck point, and performs similarity calculation on the bottleneck index data curve and a preset feature model to obtain third similarity information; comparing each third similarity in the third similarity information with a second preset similarity threshold value respectively; if the third target similarity with the value larger than or equal to the second preset similarity threshold exists in each third similarity, determining a preset feature model corresponding to the third target similarity as a target feature model, and indexing a performance fluctuation root cause corresponding to the target feature model in a preset fluctuation model library to serve as the performance fluctuation root cause of the application to be tested. By calculating the third similarity information between the bottleneck index data curve of the performance bottleneck point and the preset feature model, the performance fluctuation root cause of the application to be tested can be rapidly and accurately determined according to the third similarity information, and the efficiency of the application performance test can be improved.
In a specific embodiment of the present application, when a pressure end initiates pressure, a pressure source is taken as an example, a pressure source concurrent pressure curve and a performance golden index curve model are described, a numerical matrix of curve indexes such as wave crest, amplitude, duration and the like is taken as model features, hamming distances between feature matrices are calculated as similarity, if the model feature charging similarity between curve pairs is lower than 0.3, the curve model pairs are considered to be unmatched, it can be determined that the pressure is not normally transmitted, then a bottleneck point of the performance is located at the pressure end, and the problem of pressure initiation can be further located by checking scripts, test data, parameter configuration and the like; in the case where the similarity of the concurrent pressure curve at the pressure end and the reference curve is higher than 0.8, then the bottleneck point on the non-pressure conductive path at the pressure end is considered. Further, for each node on the pressure conduction path, the manner of operation is similar to the pressure source. And extracting model characteristic values of the resource consumption curves of all the nodes one by one along the pressure conduction path, performing similarity matching with a concurrent pressure curve and a gold index curve of the pressure source, and identifying the nodes with the fluctuation curve model characteristic similarity lower than 0.3 as performance bottleneck points. Further calculating the similarity between the bottleneck point fluctuation curve and various fluctuation models in the historical feature library, and if the model feature similarity is higher than 0.8, confirming the fluctuation classification of the node index curve so as to identify the fluctuation rule classification of the curve. Based on the classification of the fluctuation rule of the curve, the root cause can be further confirmed in combination with the following means. Firstly, whether a gold index curve of a pressure end source belongs to regular fluctuation classification is confirmed, for the gold index curve of the pressure end source with regular fluctuation, whether the fluctuation curves of system resource consumption, network delay end and database end resource consumption exist on each node on a key path or not can be further judged, if the similar fluctuation curves exist, performance bottlenecks exist on the responding nodes and index points can be judged, and the resource consumption or application state of a single system can be further analyzed to determine bottleneck root cause. If the fluctuation curve without similarity exists in the regular fluctuation, the fluctuation amplitude of the golden index curve at a certain time node is reduced or tends to be converged, and the monitoring of a server can be assisted to judge that the resource consumption reaches the system limit or the system parameter setting is unreasonable. If the curve trend analysis shows that the golden index curve fluctuation is irregular and the system monitoring indexes are normal, further analyzing the resource condition of the application end, and if the memory utilization rate of the application end stack shows a trend of fluctuation upwards, primarily judging that the application has the problem of memory leakage; if the thread number trend at the application end is normal and the cpu consumption shows a trend of fluctuation downwards, whether the application has data cache or not can be preliminarily judged, and the problems of data volume reduction and the like of the application can be solved.
The embodiment provides a method, equipment, medium and computer program product for testing application performance, which are used for acquiring a node monitoring index curve of an application to be tested, and performing similarity calculation on the node monitoring index curve to obtain first similarity information; determining whether performance fluctuation exists in the application to be tested or not based on the first similarity information; if the performance fluctuation exists in the application to be tested, determining an abnormal node where the performance fluctuation exists according to the first similarity information; acquiring an index data curve of the abnormal node, and performing similarity calculation on the index data curve to obtain second similarity information; determining a performance bottleneck point in the abnormal node according to the second similarity information; and determining the performance fluctuation root cause of the application to be tested according to the performance bottleneck point. According to the method and the device, whether the performance fluctuation exists in the application to be tested is automatically and accurately determined by carrying out similarity calculation on the node monitoring index curve of the application to be tested, when the performance fluctuation exists in the application to be tested, the abnormal node where the performance fluctuation exists is determined, then the performance bottleneck point in the abnormal node is determined, finally, the root cause of the performance fluctuation of the application to be tested is rapidly and accurately positioned through similarity matching between the curves, the skill requirements of relevant personnel of the application are reduced, the accuracy is higher in the aspect of positioning the performance fluctuation problem of the application, and the efficiency of the application performance test is effectively improved.
Further, referring to fig. 3, a second embodiment of the method for testing application performance of the present application is provided based on a first embodiment of the method for testing application performance of the present application, in the second embodiment, the step of performing similarity calculation on the node monitoring indicator curve to obtain first similarity information includes:
s11, carrying out quantization processing on the node monitoring index curve to obtain a node monitoring index curve after the quantization processing;
step S12, determining a concurrent pressure curve of a source in the node monitoring index curve and a performance gold index as a reference curve pair;
and S13, performing similarity calculation on the monitoring index curves of other nodes in the node monitoring index curves and the reference curve pair respectively to obtain a plurality of first similarities and form first similarity information.
It will be appreciated that the present embodiment is based on the concurrent pressure curve C p And a golden index curve C of performance ta Respectively extractingWithin a unit voltage time period [ t ] b ，t e ]The fluctuation characteristic factors of all relevant fluctuation curves mainly comprise fluctuation frequency W f Amplitude W of fluctuation a And duration T p The function relation between the wave characteristics F of the extracted wave curve c Establishing a reference curve pair model M based on the functional relation and the fluctuation characteristics c ＝F(W f ,W a ,T p ) And quantizing the acquired curve according to the reference curve, so that the quantized curve can be subjected to similarity calculation. The node monitoring index curves are quantized through the reference curve, so that the node monitoring index curves on the key path in the current pressure measurement scene can be analyzed in a unit time period [ ct ] b ,ct e ]The fluctuation characteristics in the pressure transmission critical path are calculated and matched with the model characteristics of a reference curve pair formed by the concurrent pressure curve and the performance gold index curve of the source one by one from the source of the pressure transmission critical path f To identify index curve pairs on the critical path that are outside the similarity threshold range with the concurrent pressure curve, the fluctuating characteristic of the performance golden index curve, and the model.
Since the concurrent pressure is conducted backwards along the business transaction critical path from the pressure source, the monitoring index of each node on the critical path must have a certain similarity with the performance gold index under normal conditions. Therefore, the system inputs the obtained node monitoring index curve into a reference curve pair model, extracts the fluctuation characteristics of each node monitoring index curve according to the function relation among the fluctuation frequency, the fluctuation amplitude and the duration of the node monitoring index curve through the reference curve pair model, realizes the quantization processing of the node monitoring index curve, and outputs the node monitoring index curve after the quantization processing. Further, the system predefines the concurrent pressure curve and the performance golden index curve of the source as a reference curve pair, and takes the fluctuation characteristics of the concurrent pressure curve and the performance golden index curve of the source as the calculation content of the reference curve pair.
Further, the system calculates the similarity between the monitoring index curves of other nodes except the source node in the quantized node monitoring index curves and the reference curve pair respectively, specifically, because the monitoring index curves of all the nodes comprise server resource consumption curves and response time trend curves, the similarity between the server resource consumption curves and the concurrent pressure curves in the reference curve pair, the similarity between the server resource consumption curves and the performance gold index curves in the reference curve pair, the similarity between the response time trend curves and the concurrent pressure curves in the reference curve pair, and the similarity between the response time trend curves and the performance gold index curves in the reference curve pair are calculated respectively, when the similarity calculation is performed, the fluctuation characteristics of the two curves are obtained respectively, the Hamming distance between the two curves is calculated according to the fluctuation characteristics of the two curves, and the calculated Hamming distance is taken as the similarity between the two curves. After obtaining the plurality of similarities, the system may take an average value of the plurality of similarities as a first similarity between the node monitoring index curve and the reference curve pair. Or the system can also obtain the similarity weight preset for the node according to the position of the node in the link, and calculate according to the average value of the similarity weight and a plurality of similarities to obtain the first similarity between the node monitoring index curve and the reference curve pair.
In the embodiment, the node monitoring index curves are quantized, the similarity between each curve and the reference curve pair in the node monitoring index curves (including the source) of the application to be tested is calculated, and first similarity information is formed, so that whether performance fluctuation exists in the application to be tested can be automatically and accurately determined according to the first similarity information, and performance bottleneck points can be rapidly positioned based on the first similarity information when the performance fluctuation exists in the application to be tested, and the efficiency of application performance test is improved.
Further, referring to fig. 4, a third embodiment of the testing method of application performance of the present application is provided based on the first embodiment of the testing method of application performance of the present application, and in the third embodiment, the step of determining whether the application to be tested has performance fluctuation based on the first similarity information includes:
step S21, comparing each first similarity in the first similarity information with a first preset similarity threshold value respectively;
step S22, if there is an intermediate similarity smaller than the first preset similarity threshold in each of the first similarities, determining that the application to be tested has performance fluctuation.
After the first similarity information formed by the plurality of first similarities is obtained, in order to determine whether the performance fluctuation exists in the application to be tested, the system respectively obtains each first similarity in the first similarity information, and compares each first similarity with a first preset similarity threshold value respectively so as to determine the magnitude relation between each first similarity and the first preset similarity threshold value. Specifically, the system may take each first similarity as a subtracted number, take a first preset similarity threshold as a subtracted number, perform a difference operation on each first similarity and the first preset similarity threshold through a computing device, for example, a calculator, so as to obtain a plurality of differences, and determine whether an intermediate similarity smaller than the first preset similarity threshold exists in each first similarity according to the plurality of differences. Further, if it is determined that, according to the plurality of differences, there is an intermediate similarity smaller than the first similarity threshold in each first similarity, for example, if one or more differences in the differences are negative, that is, if the first similarity is smaller than the first preset similarity threshold, it is determined that there is performance fluctuation in the application to be tested. Conversely, if the differences are all values not smaller than 0, which indicates that the first similarities are all greater than or equal to the first preset similarity threshold, it is determined that the application to be tested has no performance test, and the test flow of the application performance can be ended.
According to the method and the device for testing the performance of the application, the limitation of manual testing can be avoided according to the similarity among curves in the node monitoring index curve of the application to be tested, whether the application to be tested has performance fluctuation or not can be automatically and accurately determined, and the efficiency of the application performance testing can be improved.
Further, referring to fig. 5, a fourth embodiment of the testing method for application performance of the present application is provided based on the first embodiment of the testing method for application performance of the present application, in the fourth embodiment, the step of determining, according to the first similarity information, an abnormal node where the performance fluctuation is located includes:
step S31, obtaining each intermediate similarity smaller than the first preset similarity threshold in the first similarity information;
step S32, carrying out numerical comparison on each intermediate similarity to determine a first target similarity with the smallest numerical value in each intermediate similarity;
and step S33, determining the node corresponding to the first target similarity as an abnormal node where the performance fluctuation is located.
When determining the abnormal node where the performance fluctuation is located according to the first similarity information, the system firstly identifies all first similarities smaller than a first preset similarity threshold value in the first similarity information, and takes all first similarities smaller than the first preset similarity threshold value as intermediate similarities. Further, the system compares the values of the intermediate similarities, specifically, the values of the intermediate similarities may be compared one by one, and the intermediate similarity with the smallest value is determined from the plurality of intermediate similarities to be used as the first target similarity.
Further, the system identifies a monitoring index curve corresponding to the first target similarity, locates a node where the monitoring index curve corresponding to the first target similarity is located, and determines the node as an abnormal node where the performance fluctuation is located. It can be understood that if there are multiple intermediate similarities with the smallest value among the multiple intermediate similarities, that is, there are multiple intermediate similarities with the same value and the smallest value, the nodes where the monitoring index data corresponding to the intermediate similarities are located are determined respectively, and then the node closest to the source in the nodes is determined as the abnormal node.
In the embodiment, the numerical comparison is performed on each intermediate similarity smaller than the first preset similarity threshold in the first similarity information, so that the node with the smallest similarity distance (namely the numerical value of the similarity) in the pressure conduction process is rapidly and accurately identified as the abnormal node where the performance fluctuation is located, and the efficiency of the application performance test can be improved.
Further, referring to fig. 6, a fifth embodiment of the test method for application performance of the present application is provided based on the first embodiment of the test method for application performance of the present application, in the fifth embodiment, the step of determining the performance bottleneck point in the abnormal node according to the second similarity information includes:
Step S51, comparing the values of the second similarities in the second similarity information, and determining a second target similarity with the smallest value in the second similarities;
and step S52, determining the index item corresponding to the second target similarity as a performance bottleneck point in the abnormal node.
After the second similarity information is obtained, the system compares the values of the second similarities included in the second similarity information, specifically, the values of the second similarities may be compared one by one, and the second similarity with the smallest value is determined from the plurality of second similarities to be the second target similarity. Further, the system identifies an index data curve corresponding to the second target similarity, locates an index item of the index data curve corresponding to the second target similarity, and determines the index item as a performance bottleneck point in the abnormal node.
According to the embodiment, after the abnormal node where the performance fluctuation is located is determined, the performance bottleneck point in the abnormal node is rapidly and accurately positioned by analyzing each second similarity in the second similarity information, so that the efficiency of application performance test can be improved.
Further, referring to fig. 7, based on the first embodiment of the testing method of application performance of the present application, a sixth embodiment of the testing method of application performance of the present application is provided, in the sixth embodiment, the step of determining, according to the performance bottleneck point, a root cause of performance fluctuation of the application to be tested includes:
Step S61, obtaining a bottleneck index data curve of the performance bottleneck point, and performing similarity calculation on the bottleneck index data curve and a preset feature model to obtain a third similarity;
step S62, comparing the third similarity with a second preset similarity threshold;
step S63, if the third similarity is greater than or equal to the second preset similarity threshold, determining a performance fluctuation root cause of the application to be tested based on the third similarity.
After determining the performance bottleneck point in the abnormal node, the system acquires a bottleneck index data curve of the performance bottleneck point, and it can be understood that the bottleneck index data curve of the performance bottleneck point in this embodiment is an index data curve of one index item in the index data curve of the abnormal node, that is, the bottleneck index data curve of the performance bottleneck point includes one of a pressure measurement data fluctuation curve of the abnormal node at the press end, a resource consumption data curve of the server end, a performance data curve of the application program layer, and the like. Further, the system carries out quantization processing on the bottleneck index data curve of the performance bottleneck point through the reference curve, then carries out similarity calculation on the bottleneck index data curve after the quantization processing (or directly obtains the index data curve after the abnormal node quantization) and a preset feature model, specifically obtains the preset feature model from a preset fluctuation model library, carries out Hamming distance calculation on the bottleneck index data curve and the preset feature model, and takes the Hamming distance obtained through calculation as a third similarity between the bottleneck index data curve and the preset feature model, wherein the preset feature model comprises index data curves corresponding to the performance bottleneck points of different index items when the performance fluctuation exists in application, and therefore the similarity between the bottleneck index data curve of the current performance bottleneck point and the index data curve when the performance fluctuation exists can be determined through similarity calculation on the bottleneck index data curve and the preset feature model. Further, the system compares the third similarity with a second preset similarity threshold, specifically, the third similarity can be used as a subtracted number, the second preset similarity threshold is used as a subtracted number to perform a difference value operation, whether the difference value obtained by the operation is smaller than 0 is determined, if so, the bottleneck index data curve of the index item is not matched with the index data curve in the case of performance fluctuation; if the calculated difference is not less than 0, the third similarity is larger than or equal to a second preset similarity threshold, the bottleneck index data curve of the index item is matched with the index data curve in the process of performance fluctuation, and the root cause of application performance fluctuation can be found through a preset feature model. The system determines a root cause of performance fluctuation of the application under test based on the third similarity. The second preset similarity threshold is a value that can be set based on the actual environment and the requirements. Moreover, the first preset similarity threshold and the second preset similarity threshold can be dynamically adjusted, and the first preset similarity threshold and the second preset similarity threshold can be the same or different before adjustment or after adjustment. It can be understood that, the present application performs analysis modeling on various concurrent pressure curves and gold index curves in the existing performance test scenario, describes the fluctuation characteristics of each type of performance fluctuation curve pair, marks the corresponding performance root cause thereof, stores the corresponding performance root cause in a fluctuation model pair-root cause resource library, and establishes a preset fluctuation model library containing the index relationship between each type of fluctuation model pair and the performance bottleneck root cause causing such fluctuation, for example, the index type may be in the form as shown below:
Index([Model ta ,Model app ],App pb ,Para[hd,conf…])
Further, the step of determining a root cause of the performance fluctuation of the application to be tested based on the third similarity includes:
step S631, determining the preset feature model corresponding to the third similarity as a target feature model;
step S632, indexing the performance fluctuation root cause corresponding to the target feature model in a preset fluctuation model library;
and step S633, determining the performance fluctuation root cause corresponding to the target feature model as the performance fluctuation root cause of the application to be tested.
It can be appreciated that this application enumerates several common performance golden index curve models, the first being a concurrent number curve, presenting a pattern that increases with time under compression; the normal performance response time or TPS curve can be increased or reduced stepwise along with the increase of the concurrency number, after the concurrency number is stable, regular small-amplitude fluctuation appears, the fluctuation amplitude and the frequency of the small-amplitude fluctuation are both in a threshold range, and the fluctuation of the curve keeps a stable extending trend along with the duration of the concurrency time. Similarly, based on the pressure conduction process, the performance monitoring index curve in the subsequent node is similar to the performance gold index curve, the stepwise change occurs along with the concurrent increase, and after the concurrent number is stable, the monitoring index also has stable fluctuation and extending trend in the threshold range. The abnormal fluctuation conditions are relatively more, the step increase or decrease possibly occurs along with the increase of the concurrency number, but the large-amplitude regular or irregular fluctuation occurs after the concurrency number is stable, the fluctuation amplitude or frequency of the abnormal fluctuation exceeds the threshold range, and the abnormal fluctuation with obvious upward or downward trend or no obvious trend exists.
When the performance fluctuation root cause of the application to be tested is determined based on the third similarity, the system firstly determines a preset feature model corresponding to the third similarity as a target feature model, and because the preset fluctuation model library contains an index relation between each type of fluctuation model pair and the performance bottleneck root cause causing the fluctuation, the system can index the performance fluctuation root cause through the index relation by inputting the target feature model into the preset fluctuation model library, and determines the performance fluctuation root cause corresponding to the target feature model as the performance fluctuation root cause of the application to be tested, thereby completing analysis of the performance fluctuation root cause. Furthermore, the system can also output a diagnosis report through the performance fluctuation root cause so as to assist software testing, development and operation and maintenance personnel to quickly locate software performance bottlenecks. The root cause classification for performance fluctuation in this embodiment includes: the method comprises the steps that a pressure end [ script, parameter data ], a link side [ network, peripheral system, parameter configuration ], a server end [ resource consumption, application performance and parameter configuration ] can be subdivided and explained for each type of fluctuation root cause, wherein in script type fluctuation, in a common Jmeter script example, the use of a bell shell or the use of various press resource consumption plug-ins can be compared with a reference curve or trend of a template script, and the risk is prompted to occur for a user; the judgment logic of the network fluctuation starts from the fluctuation frequency, and whether the fluctuation of the network delay/packet loss/jitter is consistent with the fluctuation of the golden curve or not; for fluctuation of operating system parameters, network protocol related parameters are common, and whether regular fluctuation occurs can be judged by combining a thread number change curve; response time fluctuations caused by peripheral system performance problems, the fluctuation curves of which are substantially consistent with golden curve fluctuations, where the database performance problems are similar to those of the peripheral system; the fluctuation caused by the aspect of server resources mostly appears at the bottleneck point of resource consumption, and under the condition of resource shortage, the golden curve has no relation to the fluctuation of the step increment of the concurrency number. In general, there is no obvious similarity relation between the resource consumption of the application server, the number of application threads and the concurrency number, and the golden index curve tends to the performance upper limit, and the fluctuation amplitude does not change with the increase of the concurrency number.
According to the method and the device for testing the performance of the application, similarity calculation is conducted according to the bottleneck index data curve of the performance bottleneck point and the preset feature model, the performance fluctuation root cause of the application to be tested is determined through the obtained third similarity, the root cause causing performance fluctuation can be rapidly and accurately located according to the similarity between the bottleneck index curve of the performance bottleneck point and the preset feature model, and efficiency of testing the application performance is improved.
In addition, the application further provides a medium, preferably a computer readable storage medium, on which a test program of application performance is stored, wherein the test program of application performance realizes the steps of the embodiments of the test method of application performance when being executed by a processor.
Furthermore, the present application provides a computer program product comprising a computer program which, when executed by a processor, implements the steps of the embodiments of the test method for application performance described above.
In the embodiments of the application performance test apparatus, the computer readable medium and the computer program product of the present application, all technical features of each embodiment of the application performance test method are included, and description and explanation contents are basically the same as those of each embodiment of the application performance test method, which are not described herein.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The foregoing embodiment numbers of the present application are merely for describing, and do not represent advantages or disadvantages of the embodiments.
From the above description of the embodiments, it will be clear to those skilled in the art that the above-described embodiment method may be implemented by means of software plus a necessary general hardware platform, but of course may also be implemented by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (such as ROM/RAM, magnetic disk, optical disk), comprising several instructions for causing a terminal device (which may be a fixed terminal, such as an intelligent device for internet of things, including intelligent home such as an intelligent air conditioner, an intelligent electric lamp, an intelligent power supply, an intelligent router, or a mobile terminal, including a smart phone, a wearable internet AR/VR device, an intelligent sound box, an automatic driving car, or many other internet devices) to perform the method described in the embodiments of the present application.
The foregoing description is only of the preferred embodiments of the present application, and is not intended to limit the scope of the claims, and all equivalent structures or equivalent processes using the descriptions and drawings of the present application, or direct or indirect application in other related technical fields are included in the scope of the claims of the present application.

Claims (8)

1. The application performance testing method is characterized by comprising the following steps of:
acquiring a node monitoring index curve of an application to be tested, and performing similarity calculation on the node monitoring index curve to obtain first similarity information;
determining whether performance fluctuation exists in the application to be tested or not based on the first similarity information;
if the performance fluctuation exists in the application to be tested, determining an abnormal node where the performance fluctuation exists according to the first similarity information;
acquiring an index data curve of the abnormal node, and performing similarity calculation on the index data curve to obtain second similarity information;
determining a performance bottleneck point in the abnormal node according to the second similarity information;
determining a performance fluctuation root cause of the application to be tested according to the performance bottleneck point;
The step of obtaining first similarity information includes:
carrying out quantization treatment on the node monitoring index curve to obtain a node monitoring index curve after the quantization treatment;
determining a concurrent pressure curve of a source in the node monitoring index curve and a performance gold index as a reference curve pair;
respectively carrying out similarity calculation on monitoring index curves of other nodes in the node monitoring index curves and the reference curve pair to obtain a plurality of first similarities and form first similarity information;
the step of calculating the similarity of the index data curve to obtain second similarity information comprises the following steps:
carrying out quantization processing on the index data curve to obtain a quantized index data curve;
and respectively carrying out similarity calculation on the index data curve after the quantization processing and the reference curve pair to obtain a plurality of second similarity and form second similarity information.
2. The method for testing performance of an application according to claim 1, wherein the step of determining whether the application to be tested has performance fluctuations based on the first similarity information includes:
Comparing each first similarity in the first similarity information with a first preset similarity threshold value respectively;
and if the intermediate similarity smaller than the first preset similarity threshold exists in each first similarity, determining that the application to be tested has performance fluctuation.
3. The method for testing application performance according to claim 1, wherein the step of determining the abnormal node where the performance fluctuation is located according to the first similarity information includes:
obtaining each intermediate similarity smaller than a first preset similarity threshold in the first similarity information;
performing numerical comparison on each intermediate similarity to determine a first target similarity with the smallest numerical value in each intermediate similarity;
and determining the node corresponding to the first target similarity as an abnormal node where the performance fluctuation is located.
4. The method for testing application performance according to claim 1, wherein the step of determining the performance bottleneck point in the abnormal node according to the second similarity information includes:
performing numerical comparison on each second similarity in the second similarity information to determine a second target similarity with the smallest numerical value in each second similarity;
And determining the index item corresponding to the second target similarity as a performance bottleneck point in the abnormal node.
5. The method for testing performance of an application according to claim 1, wherein the step of determining a root cause of performance fluctuation of the application to be tested according to the performance bottleneck point comprises:
obtaining a bottleneck index data curve of the performance bottleneck point, and performing similarity calculation on the bottleneck index data curve and a preset characteristic model to obtain a third similarity;
comparing the third similarity with a second preset similarity threshold;
and if the third similarity is greater than or equal to the second preset similarity threshold, determining a performance fluctuation root cause of the application to be tested based on the third similarity.
6. The method for testing performance of an application according to claim 5, wherein the step of determining a root cause of performance fluctuation of the application to be tested based on the third similarity comprises:
determining a preset feature model corresponding to the third similarity as a target feature model;
indexing a performance fluctuation root cause corresponding to the target feature model in a preset fluctuation model library;
and determining the performance fluctuation root cause corresponding to the target feature model as the performance fluctuation root cause of the application to be tested.
7. An application performance testing apparatus, characterized in that the application performance testing apparatus comprises a memory, a processor and an application performance testing program stored on the memory and executable on the processor, which application performance testing program, when executed by the processor, implements the steps of the application performance testing method according to any one of claims 1-6.
8. A medium, which is a computer-readable storage medium, wherein the computer-readable storage medium has stored thereon a test program for application performance, which when executed by a processor, implements the steps of the method for testing application performance according to any one of claims 1-6.
CN202110879239.0A 2021-07-30 2021-07-30 Method, device, medium and computer program product for testing application performance Active CN113656270B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110879239.0A CN113656270B (en) 2021-07-30 2021-07-30 Method, device, medium and computer program product for testing application performance

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110879239.0A CN113656270B (en) 2021-07-30 2021-07-30 Method, device, medium and computer program product for testing application performance

Publications (2)

Publication Number Publication Date
CN113656270A CN113656270A (en) 2021-11-16
CN113656270B true CN113656270B (en) 2024-03-08

Family

ID=78490194

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110879239.0A Active CN113656270B (en) 2021-07-30 2021-07-30 Method, device, medium and computer program product for testing application performance

Country Status (1)

Country Link
CN (1) CN113656270B (en)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111001275A (en) * 2018-10-08 2020-04-14 上海洁鹿环保科技有限公司 Data processing method and device for waste gas treatment system and storable medium
CN111160329A (en) * 2019-12-27 2020-05-15 深圳前海微众银行股份有限公司 Root cause analysis method and device
CN111811567A (en) * 2020-07-21 2020-10-23 北京中科五极数据科技有限公司 Equipment detection method based on curve inflection point comparison and related device
CN112882796A (en) * 2021-02-25 2021-06-01 深信服科技股份有限公司 Abnormal root cause analysis method and apparatus, and storage medium
CN113010805A (en) * 2021-02-23 2021-06-22 腾讯科技（深圳）有限公司 Index data processing method, device, equipment and storage medium

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10909018B2 (en) * 2015-09-04 2021-02-02 International Business Machines Corporation System and method for end-to-end application root cause recommendation
US10318366B2 (en) * 2015-09-04 2019-06-11 International Business Machines Corporation System and method for relationship based root cause recommendation

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111001275A (en) * 2018-10-08 2020-04-14 上海洁鹿环保科技有限公司 Data processing method and device for waste gas treatment system and storable medium
CN111160329A (en) * 2019-12-27 2020-05-15 深圳前海微众银行股份有限公司 Root cause analysis method and device
CN111811567A (en) * 2020-07-21 2020-10-23 北京中科五极数据科技有限公司 Equipment detection method based on curve inflection point comparison and related device
CN113010805A (en) * 2021-02-23 2021-06-22 腾讯科技（深圳）有限公司 Index data processing method, device, equipment and storage medium
CN112882796A (en) * 2021-02-25 2021-06-01 深信服科技股份有限公司 Abnormal root cause analysis method and apparatus, and storage medium

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
压力抗血栓装置气泵性能测试系统设计;史清清;廖跃华;尚昆;徐晶伟;;北京生物医学工程(第05期);78-83 *
基于LoadRunner的企业级应用性能测试设计与实现;王淑满;潘丽云;田雨;马宏全;钟坚飞;;电脑知识与技术(第25期);215-217 *
王淑满 ; 潘丽云 ; 田雨 ; 马宏全 ; 钟坚飞 ; .基于LoadRunner的企业级应用性能测试设计与实现.电脑知识与技术.2020,(第25期),215-217. *

Also Published As

Publication number Publication date
CN113656270A (en) 2021-11-16

Similar Documents

Publication Publication Date Title
CN107341098B (en) Software performance testing method, platform, equipment and storage medium
CN109491894B (en) Interface test method and equipment
WO2018059402A1 (en) Method and apparatus for determining fault type
CN108345670B (en) Service hotspot discovery method for 95598 power work order
CN106713011B (en) Method and system for obtaining test data
CN113553267A (en) Application performance testing method, device, medium, and computer program product
CN109462517A (en) A kind of method, system and the equipment of the data monitoring towards digital electric network business
US7617313B1 (en) Metric transport and database load
CN115562978A (en) Performance test system and method based on service scene
CN105868956A (en) Data processing method and device
US20120226484A1 (en) Calculation simulation system and method thereof
CN115016766A (en) Internet and cloud computing software development method
JP2023534331A (en) CONTENT DISTRIBUTION NETWORK PROCESSING METHOD, APPARATUS, ELECTRONIC DEVICE, COMPUTER READABLE STORAGE MEDIUM AND COMPUTER PROGRAM
CN113127356A (en) Pressure measurement method and device, electronic equipment and storage medium
CN113656270B (en) Method, device, medium and computer program product for testing application performance
CN107257290B (en) Test method and system for open SOA service-oriented architecture
CN116561542B (en) Model optimization training system, method and related device
CN111061800A (en) STATCOM transient response analysis system and method based on fault recording file
CN115952098A (en) Performance test tuning scheme recommendation method and system
CN115080412A (en) Software update quality evaluation method, device, equipment and computer storage medium
CN112035366B (en) Test case generation method, device and equipment
CN114997574A (en) Power distribution station area elastic resource management method and device based on service middling station
CN113064812A (en) Project development process quality defect prediction method, device and medium
CN114610599A (en) Test method and system
CN102256271B (en) The acquisition methods and device of capacity expansion indication information

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant