CN116450485A - Detection method and system for application performance interference - Google Patents
Detection method and system for application performance interference Download PDFInfo
- Publication number
- CN116450485A CN116450485A CN202310678411.5A CN202310678411A CN116450485A CN 116450485 A CN116450485 A CN 116450485A CN 202310678411 A CN202310678411 A CN 202310678411A CN 116450485 A CN116450485 A CN 116450485A
- Authority
- CN
- China
- Prior art keywords
- index
- application
- indexes
- model
- kernel
- 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.)
- Granted
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 32
- 238000012360 testing method Methods 0.000 claims abstract description 46
- 238000004458 analytical method Methods 0.000 claims abstract description 33
- 238000000034 method Methods 0.000 claims abstract description 33
- 230000036541 health Effects 0.000 claims abstract description 27
- 238000012549 training Methods 0.000 claims abstract description 20
- 238000012216 screening Methods 0.000 claims abstract description 16
- 238000012545 processing Methods 0.000 claims abstract description 10
- 238000012417 linear regression Methods 0.000 claims abstract description 8
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000010998 test method Methods 0.000 claims 1
- 238000009530 blood pressure measurement Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 230000004044 response Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- LMOINURANNBYCM-UHFFFAOYSA-N metaproterenol Chemical compound CC(C)NCC(O)C1=CC(O)=CC(O)=C1 LMOINURANNBYCM-UHFFFAOYSA-N 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 241000135164 Timea Species 0.000 description 1
- 230000004931 aggregating effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000003825 pressing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/3058—Monitoring arrangements for monitoring environmental properties or parameters of the computing system or of the computing system component, e.g. monitoring of power, currents, temperature, humidity, position, vibrations
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Hardware Design (AREA)
- Computing Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a detection method and a detection system for application performance interference, which belong to the technical field of electric digital data processing, wherein the method comprises the following steps: obtaining a test data set for applying a pressure test; screening model indexes from the kernel indexes according to the correlation between the service indexes and the kernel indexes; training the test data of the model index based on a linear regression algorithm to obtain an interference analysis model; analyzing the operation data of the applied model index through an interference analysis model to obtain a predicted value of the service index; and obtaining the interference state of the application performance according to the predicted value and the health threshold value. An interference analysis model is built through part of kernel indexes, online application is analyzed and judged, the interference state of the application is found in time, the interfered application is processed in time, the normal operation of the application is ensured, and the service quality of a service is ensured; only part of kernel indexes are monitored, so that direct detection of service indexes is avoided, and intrusion into applications is avoided.
Description
Technical Field
The invention relates to the technical field of electric digital data processing, in particular to a detection method and a detection system for application performance interference.
Background
In recent years, with the rapid development of cloud computing technology, cloud services are a necessary trend. However, as the service on the cloud increases, the resource utilization rate of the system is not obviously improved. The average resource utilization of the global data center is reported to be only about 10%. The application of different workload types is mixed and deployed by means of scheduling, resource isolation and the like, and the technology of resource complementation, resource time-sharing multiplexing and the like is realized, so that the method becomes an effective method for improving the resource utilization rate of the data center. In the application mixed part scene, the data center is regarded as a super computer, and various applications such as online application, batch application, AI application and the like uniformly run on the super computer and share the resources of the super computer.
However, after the application is mixed and deployed, since a plurality of applications share uniform resources, serious resource contention is caused among the applications, and for some online applications sensitive to service indexes such as response time, the performance of the applications is reduced, the service quality is reduced, and the user experience is poor, even economic loss is caused. Therefore, after application hybrid deployment, the interference state of the application performance is perceived and monitored, and then subsequent scheduling is performed once the performance of some core services is found to be interfered, so that it is important to ensure the service quality of the core services. Application performance monitoring has long been the focus of research in both academia and industry.
The current mainstream method is to construct a performance monitoring model based on service indexes such as request response time and request error rate of an application, automatically calculate a threshold value by manually setting the threshold value or based on an algorithm, then compare the service index value with the threshold value in real time, and consider that the application performance is interfered once the threshold value is frequently exceeded. However, for batch processing, AI and other types of applications, this method cannot collect service indexes such as request response time, request error rate and the like. And therefore cannot be effectively detected for these applications.
Disclosure of Invention
Aiming at the technical problems in the prior art, the invention provides a detection method and a detection system for application performance interference, which are used for realizing effective detection of service indexes and avoiding direct monitoring of complex service indexes by analyzing part of kernel indexes and predicting the service indexes.
The invention discloses a detection method for application performance interference, which comprises the following steps: obtaining a test data set of an application pressure test, wherein the test data set comprises test data of a kernel index and a service index; screening model indexes from the kernel indexes according to the correlation between the service indexes and the kernel indexes; training the test data of the model index based on a linear regression algorithm to obtain an interference analysis model; acquiring running data of an applied model index; analyzing the operation data through an interference analysis model to obtain a predicted value of a business index; acquiring a health threshold of a business index; and obtaining the interference state of the application performance according to the predicted value and the health threshold value. An interference analysis model is built through part of kernel indexes, online application is analyzed and judged, the interference state of the application is found in time, the interfered application is processed in time, the normal operation of the application is ensured, and the service quality of a service is ensured; only part of kernel indexes are monitored, so that direct detection of service indexes is avoided, and intrusion into applications is avoided.
Wherein the method of obtaining a test dataset comprises: building a pressure test environment based on Kubernetes; deploying an application container and a side vehicle container; collecting test data of service indexes through a side car container; and collecting test data of the kernel index through a pressure detection tool.
The method for screening the model index comprises the following steps: training test data by using a service index as a data tag and using a plurality of kernel indexes as data characteristics based on a method of extreme gradient lifting to obtain the correlation between the kernel indexes and the service index; and screening model indexes based on the correlation. After training, the correlation between the kernel index and the business index is obtained by calling the feature_importances interface.
The interference analysis model is expressed as:
wherein,,business_indicator 1 represented as a predicted value of a traffic index,ja sequence number denoted as a model index,Mexpressed as the total number of model indicators,kernel_indicator j denoted as the firstjModel indexes;w j represent the firstjThe weight of the individual model indicators is determined,brepresented as a bias value. The loss function of the interference analysis model is expressed as:
l(W,b) = 1/2 (business_indicator
1
-business_indicator)
2
wherein,,business_indicatorrepresented as the actual value of the traffic indicator,l(W,b)error values expressed as business index true values and predicted values.
The health threshold may be calculated based on a quartile method, the calculation formula of which is expressed as:
business_threshold = Q3 + 1.5×(Q3-Q1)
wherein,,business_thresholda health threshold value expressed as a traffic indicator,Q3indicating the index value at the 3/4 th position after the service index data is sequenced from small to large,Q1indicating the index value at 1/4 of the index data after sorting from small to large.
The invention also comprises a method for alarming: judging whether the following conditions are satisfied: predictive value continuation of business indexNA detection period exceeding a health threshold, whereinNIs a natural number; if yes, generating and sending an alarm when the performance of the application is in an interfered state; if the model index is not satisfied, continuously detecting the operation data of the model index.
In a specific embodiment, the first application and the second application are deployed in a mixed mode, and the alarm processing method comprises the following steps: obtaining priorities of a first application and a second application, wherein the priority of the first application is lower than that of the second application; if the first application is interfered, generating an alarm; and if the second application is interfered, the first application is evicted.
Compared with the prior art, the invention has the beneficial effects that: an interference analysis model is built through part of kernel indexes, online application is analyzed and judged, the interference state of the application is found in time, the interfered application is processed in time, the normal operation of the application is ensured, and the service quality of a service is ensured; only part of kernel indexes are monitored, so that direct detection of service indexes is avoided, and intrusion into applications is avoided.
Drawings
FIG. 1 is a flow chart of a method for detecting application performance interference of the present invention;
FIG. 2 is a logic block diagram of a pressure test;
FIG. 3 is a flow chart of a detection method based on an interference analysis model;
fig. 4 is a system logic block diagram of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The invention is described in further detail below with reference to the attached drawing figures:
as shown in fig. 1, the basic flow of the detection method for application performance interference of the present invention is as follows:
step 101: and obtaining a test data set for applying the pressure test, wherein the test data set comprises test data of the kernel index and the service index. The traffic index may include, but is not limited to, a request response time, a request error rate, a task execution duration, and the like.
Step 102: and screening the model index from the kernel indexes according to the correlation between the service indexes and the kernel indexes.
Step 103: and training the test data of the model index based on a linear regression algorithm to obtain an interference analysis model.
Step 104: and acquiring the running data of the model index of the application.
Step 105: and analyzing the operation data through an interference analysis model to obtain a predicted value of the service index.
Step 106: and acquiring a health threshold of the business index.
Step 107: and obtaining the interference state of the application performance according to the predicted value and the health threshold value.
Step 108: and if the application is in the interfered state, performing interference processing.
An interference analysis model is built through part of kernel indexes, online application is analyzed and judged, the interference state of the application is found in time, the interfered application is processed in time, the normal operation of the application is ensured, and the service quality of a service is ensured; only part of kernel indexes are monitored, so that direct detection of service indexes is avoided, and intrusion into applications is avoided. The method solves the problems of difficult collection of service indexes, coarse granularity and the like of the traditional application performance analysis method, shortens the interfered discovery time and improves the service quality of the application.
The manner of the pressure test in step 101 is as follows: as shown in fig. 2, the pressure test environment is built based on Kubernetes, and the application is deployed by adopting a container mode. When the application is deployed, the application container is injected into a Sidecar container (Sidecar), the Sidecar container is used for intercepting tasks or requests related to the application, recording the execution time of the tasks or requests of the application, and transmitting the test data of the acquired service indexes to an Agent module (Agent) on a Node (Node) in an interface mode. In the test, the load number of the application should be kept consistent with the load under the real condition as much as possible. The application is subjected to pressure measurement based on a jmeter isobaric measurement tool.
Meanwhile, data of kernel indexes in the running process of the application are collected through cadvisor, perf and other tools, the collection period size can be configured, the default value is 1 second, and the kernel indexes comprise 120 indexes in total in all dimensions such as CPI, L3 Cache, memory bandwidth, CPU utilization rate and memory utilization rate.
The Agent stores the acquired data in the file of the node, and can only store the data acquired at the latest time in an overlaying mode, but is not limited to the method. The Agent exposes the RPC interface to the outside, and the data convergence storage component Prometa calls the RPC interface of the Agent on each node periodically to acquire test data, and stores the test data into the InfluxDB time sequence database. The prothenes pull cycle is configurable, defaulting to 1s, and the historical data stored in the influxdb is also configurable from time to time, defaulting to 30 days.
The mode of screening the model index in step 102 is as follows:
1) Data preprocessing: acquiring test data of a service index and a kernel index of an application pressure test from an influxDB, aggregating the data, processing the data into one piece per minute, and filling by adopting a mode if a missing value exists; the final result is a data set, with a single data format as follows:
wherein,,timea time stamp representing the test data is presented,business_indicatora traffic index value representing the data,k_indicator i representing the first of the dataiThe value of the kernel index.
2) Correlation analysis:
step 201: and training the test data by using the service indexes as data labels and using a plurality of kernel indexes as data characteristics based on an extreme gradient lifting (XGBoost, eXtreme Gradient Boosting) method to obtain the correlation between the kernel indexes and the service indexes. After training, the correlation/importance of the kernel index and the business index is obtained by calling features_importances.
Step 202: and screening model indexes based on the correlation. Specifically, the correlation is sorted from large to small, the first M kernel indexes are taken as model indexes, and M is a natural number, for example, 10. The set of model metrics may be expressed as: (kernel_ indicator 1 ,..., kernel_indicator j ,...,kernel_indicator M ). Wherein,, ja sequence number denoted as a model index,Mexpressed as the total number of model indicators,kernel_indicator j denoted as the firstjAnd (5) model indexes.
In step 103, the interference analysis model is obtained as follows: taking a model index as an independent variable, a business index as an independent variable, and constructing an applied performance interference detection model based on a linear regression algorithm, wherein the initial model is as follows:
(1)
wherein,,business_indicator 1 represented as a predicted value of a traffic index,ja sequence number denoted as a model index,Mexpressed as the total number of model indicators,kernel_indicator j denoted as the firstjModel indexes;w j represent the firstjThe weight of the individual model indicators is determined,brepresented as a bias value. The value in the initial state may be set to 1, high to 0. When M takes 10, equation 1 can also be written as:
training is performed based on the test data of the model indexes, and the weight and bias value of each model index can be obtained. The square error function is adopted as a loss function in the training process, the smaller the error value is, the smaller the loss is, and the loss in perfect prediction is 0. The loss function is expressed as:
l(W,b) = 1/2 (business_indicator 1 -business_indicator) 2 (3)
wherein,,business_indicatorrepresented as the actual value of the traffic indicator,l(W,b)error values expressed as business index true values and predicted values. After training, interference analysis modelAnd stored in a file.
In step 106, a health threshold may be empirically set; the health threshold may also be calculated based on a quartile method, and the calculation formula of the health threshold is expressed as:
business_threshold = Q3 + 1.5×(Q3-Q1) (4)
wherein,,business_thresholdthe health threshold value expressed as the business index is that Q3 represents the index value at the 3/4 th after the business index data is sorted from small to large, and Q1 represents the index value at the 1/4 th after the index data is sorted from small to large.
In step 107, for the deployment online application, whether the application performance is interfered can be judged in real time through a performance interference analysis model, and the specific steps are as follows:
step 301: after the application is deployed online, the running data of the kernel index in the running process of the application are collected in real time through the Agent running on the node and are converged into Prometa.
Step 302: and periodically (the default value is 1 s) acquiring the operation data of the model index from Prometa, and calculating the predicted value of the service index through the interference analysis model.
Step 303: comparing the predicted value with a health threshold, if the predicted value is larger than the health threshold of the service index, judging that the application is possibly in a disturbed state, and if the predicted value is smaller than or equal to the health threshold of the service index, judging that the application is in a normal state.
In step 108, if the application is in an interfered state, an alert is generated. In order to ensure the accuracy of the alarm, a buffer mechanism is added:
judging whether the following conditions are satisfied: the predicted value of the traffic indicator exceeds the health threshold for N consecutive detection periods, where N is a natural number, which may be set to 3.
If yes, the performance of the application is in an interfered state, and an alarm is generated and sent out.
If the model index is not satisfied, continuously detecting the operation data of the model index.
More specifically, corresponding interference alarm processing strategies can also be adopted for different applications. For example, the application includes a first application and a second application deployed in a hybrid manner, and the method for processing the alarm includes:
step 501: the method comprises the steps of obtaining priorities of a first application and a second application, wherein the priorities of the first application are lower than the priorities of the second application.
Step 502: and if the first application is interfered, generating an alarm. The first application is not specially treated in the case that the interference has little influence on the service quality.
Step 503: and if the second application is interfered, the first application is evicted. To ensure efficient operation of the higher priority second application.
In a specific embodiment, core services such as online transaction, online payment and the like of a certain bank are frequently preempted and interfered by other application resources deployed in the same cluster, and the service quality of the core services is easy to interfere, so that the use experience of the application is poor. Referring to fig. 3, after obtaining the interference analysis model, the detection method is as follows:
step 601: and collecting the kernel index of the core service in real time.
Step 602: and obtaining the value of the model index.
Step 603: and calculating a predicted value of the service index based on the interference analysis model.
Step 604: and judging whether the predicted value exceeds a health threshold.
If not, go to step 601.
If yes, go to step 605: and judging whether the predicted value exceeds the health threshold value 3 times continuously.
If 3 consecutive passes are made, step 606 is performed: the core service is interfered and an alarm is generated.
If not, step 601 is performed.
After transformation, whether the performance of the core service is interfered is detected in real time, so that the rapid detection and processing of the interference are realized, and the stability and the service quality of the core service are improved.
If the core service has a plurality of service indexes, an interference detection model can be respectively constructed for each service index, the predicted value of each service index is respectively detected, and if one or a part of the predicted values of the service indexes exceed or exceed the health threshold value for 3 times continuously, the core service is considered to be interfered.
The invention also provides a detection system for realizing the detection method, as shown in fig. 4, which comprises an acquisition module 1, a screening module 2, a training module 3 and an interference analysis module 4, wherein the acquisition module 1 is used for acquiring a test data set for applying pressure test; the screening module 2 is used for screening the model index from the kernel indexes according to the correlation between the service indexes and the kernel indexes; the training module 3 is used for training the test data of the model index based on a linear regression algorithm to obtain an interference analysis model; the interference analysis module 4 is used for analyzing the operation data of the applied model index through an interference analysis model to obtain a predicted value of the service index, and obtaining an interference state of the application performance according to the predicted value and the health threshold.
In a pressure measurement environment, the Sidecar technology based on Kubernetes can automatically inject a Sidecar container when pressure measurement application deployment is realized, and the Sidecar container is used for intercepting the initiation and the termination of all application tasks or requests, so that the execution time of the application tasks or requests is calculated, and the unified acquisition of business index data of various types of applications is realized; based on the combination of XGBoost algorithm, linear regression algorithm and quartile algorithm, constructing an applied kernel index-based applied performance interference analysis model, and performing model training based on applied pressure measurement data to determine parameters of the model; the application performance interference detection model based on the kernel index analyzes and judges the application after the deployment online in real time, timely discovers the interfered condition of the application, and can process according to a preset strategy to ensure timely discovery and processing of the interference.
The above is only a preferred embodiment of the present invention, and is not intended to limit the present invention, but various modifications and variations can be made to the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A method for detecting application performance interference, comprising:
obtaining a test data set of an application pressure test, wherein the test data set comprises test data of a kernel index and a service index;
screening model indexes from the kernel indexes according to the correlation between the service indexes and the kernel indexes;
training the test data of the model index based on a linear regression algorithm to obtain an interference analysis model;
acquiring running data of an applied model index;
analyzing the operation data through an interference analysis model to obtain a predicted value of a business index;
acquiring a health threshold of a business index;
and obtaining the interference state of the application performance according to the predicted value and the health threshold value.
2. The method of claim 1, wherein the method of screening the model metrics comprises:
training test data by using a service index as a data tag and using a plurality of kernel indexes as data characteristics based on a method of extreme gradient lifting to obtain the correlation between the kernel indexes and the service index;
and screening model indexes based on the correlation.
3. The method of claim 2, wherein,
after training, the correlation between the kernel index and the business index is obtained by calling the feature_importances interface.
4. The detection method according to claim 1, wherein the interference analysis model is expressed as:
,
wherein,,business_indicator 1 represented as a predicted value of a traffic index,ja sequence number denoted as a model index,Mexpressed as the total number of model indicators,kernel_indicator j denoted as the firstjModel indexes;w j represent the firstjThe weight of the individual model indicators is determined,brepresented as a bias value.
5. The method of claim 4, wherein the loss function of the interference analysis model is expressed as:
l(W,b) = 1/2 (business_indicator
1
-business_indicator)
2
wherein,,business_indicatorrepresented as the actual value of the traffic indicator,l(W,b)error values expressed as business index true values and predicted values.
6. The detection method according to claim 1, wherein the health threshold is calculated based on a quartile method, and a calculation formula of the health threshold is expressed as:
business_threshold = Q3 + 1.5×(Q3-Q1)
wherein,,business_thresholda health threshold value expressed as a traffic indicator,Q3indicating the index value at the 3/4 th position after the service index data is sequenced from small to large,Q1indicating the index value at 1/4 of the index data after sorting from small to large.
7. The method of detecting according to claim 1, further comprising a method of alerting:
judging whether the following conditions are satisfied: predictive value continuation of business indexNA detection period exceeding a health threshold, whereinNIs a natural number;
if yes, generating and sending an alarm when the performance of the application is in an interfered state;
if the model index is not satisfied, continuously detecting the operation data of the model index.
8. The method of detecting according to claim 7, wherein the applications include a first application and a second application deployed in a hybrid, the method of alert processing comprising:
obtaining priorities of a first application and a second application, wherein the priority of the first application is lower than that of the second application;
if the first application is interfered, generating an alarm;
and if the second application is interfered, the first application is evicted.
9. The method of testing according to claim 1, wherein the method of obtaining the test dataset comprises:
building a pressure test environment based on Kubernetes;
deploying an application container and a side vehicle container;
collecting test data of service indexes through a side car container;
and collecting test data of the kernel index through a pressure detection tool.
10. A detection system for application performance interference, which is used for realizing the detection method according to any one of claims 1-9, and comprises an acquisition module, a screening module, a training module and an interference analysis module,
the acquisition module is used for acquiring a test data set for applying a pressure test;
the screening module is used for screening model indexes from the kernel indexes according to the correlation between the service indexes and the kernel indexes;
the training module is used for training the test data of the model index based on a linear regression algorithm to obtain an interference analysis model;
the interference analysis module is used for analyzing the operation data of the applied model indexes through an interference analysis model to obtain predicted values of service indexes, and obtaining interference states of application performances according to the predicted values and health thresholds.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310678411.5A CN116450485B (en) | 2023-06-09 | 2023-06-09 | Detection method and system for application performance interference |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310678411.5A CN116450485B (en) | 2023-06-09 | 2023-06-09 | Detection method and system for application performance interference |
Publications (2)
Publication Number | Publication Date |
---|---|
CN116450485A true CN116450485A (en) | 2023-07-18 |
CN116450485B CN116450485B (en) | 2023-11-10 |
Family
ID=87125909
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310678411.5A Active CN116450485B (en) | 2023-06-09 | 2023-06-09 | Detection method and system for application performance interference |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116450485B (en) |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107659950A (en) * | 2017-11-16 | 2018-02-02 | 中国联合网络通信集团有限公司 | A kind of method and base station based on wireless backhaul links transmission data |
CN110311987A (en) * | 2019-07-24 | 2019-10-08 | 中南民族大学 | Node scheduling method, apparatus, equipment and the storage medium of microserver |
CN113467886A (en) * | 2021-06-25 | 2021-10-01 | 四川新网银行股份有限公司 | Pressure measurement engine cluster construction method based on containerized scheduling |
US20210383295A1 (en) * | 2020-06-05 | 2021-12-09 | Beijing Baidu Netcom Science And Technology Co., Ltd. | Method and apparatus for stress testing, electronic device and computer readable medium |
CN114297008A (en) * | 2021-12-17 | 2022-04-08 | 苏州浪潮智能科技有限公司 | Cloud host performance prediction method and device, terminal and storage medium |
CN116069591A (en) * | 2023-02-03 | 2023-05-05 | 北京蓝色星际科技股份有限公司 | Interface performance monitoring method, device, equipment and storage medium |
CN116149846A (en) * | 2022-12-05 | 2023-05-23 | 中国科学院深圳先进技术研究院 | Application performance optimization method and device, electronic equipment and storage medium |
-
2023
- 2023-06-09 CN CN202310678411.5A patent/CN116450485B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107659950A (en) * | 2017-11-16 | 2018-02-02 | 中国联合网络通信集团有限公司 | A kind of method and base station based on wireless backhaul links transmission data |
CN110311987A (en) * | 2019-07-24 | 2019-10-08 | 中南民族大学 | Node scheduling method, apparatus, equipment and the storage medium of microserver |
US20210383295A1 (en) * | 2020-06-05 | 2021-12-09 | Beijing Baidu Netcom Science And Technology Co., Ltd. | Method and apparatus for stress testing, electronic device and computer readable medium |
CN113467886A (en) * | 2021-06-25 | 2021-10-01 | 四川新网银行股份有限公司 | Pressure measurement engine cluster construction method based on containerized scheduling |
CN114297008A (en) * | 2021-12-17 | 2022-04-08 | 苏州浪潮智能科技有限公司 | Cloud host performance prediction method and device, terminal and storage medium |
CN116149846A (en) * | 2022-12-05 | 2023-05-23 | 中国科学院深圳先进技术研究院 | Application performance optimization method and device, electronic equipment and storage medium |
CN116069591A (en) * | 2023-02-03 | 2023-05-05 | 北京蓝色星际科技股份有限公司 | Interface performance monitoring method, device, equipment and storage medium |
Non-Patent Citations (3)
Title |
---|
BRUNO L. DALMAZO: "Triple-Similarity Mechanism for alarm management in the cloud", COMPUTERS & SECURITY, vol. 78 * |
张世琼;: "财政平台一体化应用系统性能测试", 计算机系统应用, no. 07 * |
李红辉;关婷婷;杨芳南;: "云计算平台状态监控技术研究与应用", 软件, no. 01 * |
Also Published As
Publication number | Publication date |
---|---|
CN116450485B (en) | 2023-11-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
KR101984730B1 (en) | Automatic predicting system for server failure and automatic predicting method for server failure | |
CN110888783B (en) | Method and device for monitoring micro-service system and electronic equipment | |
CN110912737A (en) | Dynamic perception performance early warning method based on hybrid model | |
CN106487601B (en) | Resource monitoring method, device and system | |
WO2020172852A1 (en) | Computing resource scheduling method, scheduler, internet of things system, and computer readable medium | |
CN109062769B (en) | Method, device and equipment for predicting IT system performance risk trend | |
CN107579858A (en) | The alarm method and device of cloud main frame, communication system | |
CN112486767B (en) | Intelligent monitoring method, system, server and storage medium for cloud resources | |
CN112527448A (en) | Openstack-based dynamic load adjustment method and system | |
WO2021040584A1 (en) | Entity and method performed therein for handling computational resources | |
CN112148561A (en) | Service system running state prediction method and device and server | |
CN111339052A (en) | Unstructured log data processing method and device | |
CN113934535A (en) | Mass point cloud data processing method, device, server and system | |
CN112118127B (en) | Service reliability guarantee method based on fault similarity | |
CN116010228B (en) | Time estimation method and device for network security scanning | |
CN116450485B (en) | Detection method and system for application performance interference | |
CN111125195A (en) | Data anomaly detection method and device | |
CN112749003A (en) | Method, apparatus and computer-readable storage medium for system optimization | |
CN113703394A (en) | Cutter monitoring and managing method and system based on edge calculation | |
CN114676002A (en) | PHM technology-based system operation and maintenance method and device | |
CN106888237A (en) | A kind of data dispatching method and system | |
CN104883273A (en) | Method and system for processing service influence model in virtualized service management platform | |
CN111258866A (en) | Computer performance prediction method, device, equipment and readable storage medium | |
CN115495231B (en) | Dynamic resource scheduling method and system under high concurrency task complex scene | |
CN116541252B (en) | Computer room fault log data processing method and device |
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 |