CN108563568A - A kind of detection of application performance bottleneck and diagnostic method based on clustering - Google Patents
A kind of detection of application performance bottleneck and diagnostic method based on clustering Download PDFInfo
- Publication number
- CN108563568A CN108563568A CN201810319392.6A CN201810319392A CN108563568A CN 108563568 A CN108563568 A CN 108563568A CN 201810319392 A CN201810319392 A CN 201810319392A CN 108563568 A CN108563568 A CN 108563568A
- Authority
- CN
- China
- Prior art keywords
- data
- analysis
- module
- application
- real
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/36—Preventing errors by testing or debugging software
- G06F11/3604—Software analysis for verifying properties of programs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/36—Preventing errors by testing or debugging software
- G06F11/362—Software debugging
- G06F11/366—Software debugging using diagnostics
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Quality & Reliability (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The detection of application performance bottleneck and diagnostic method that the present invention relates to a kind of based on clustering.The method utilizes the data collection under the true environment of mobile terminal, and with display module and on the line based on backstage, statistical analysis module three parts are realized with real-time exhibition module, offline client analysis.The present invention can carry out convenient, fast and accurate applied performance analysis and problem diagnosis, reduce unnecessary development cost；Time and the cost of optimizing phase are saved, further the stages such as the exploitation of application, test and optimization analysis are refined, so as to which unnecessary work is assigned to appropriate personnel, improve the development efficiency of application, improve the stability before application is reached the standard grade；Simultaneously with the continuous increase of data scale, more data minings and analysis can be carried out based on these data informations, good and bad degree and priority degree, the market trend etc. for analyzing application from more macroscopical angle.
Description
Technical field
The detection of application performance bottleneck and diagnostic method that the present invention relates to a kind of based on clustering.
Background technology
Under current environment, along with the increasingly hot of Unity3D engines, or else the threshold of exploitation also breaks and reduces, pour in
A large amount of various levels of developers, the application developed based on Unity3D engines, software, the quantity of game are growing day by day,
Cause the application that developer develops there may be various efficiencies, can not be well adapted to so as to cause product
The market demand.
For this demand of developer, usual way can carry out assistant analysis, based in Unity by the xcode of Mac
The real-time online of Profiler manual data collection frame by frame and analysis are set, but is inevitable in the presence of improving inessential volume
Outer development cost（Most development platforms are still windows）Possibility, difficult find potential problems, the positioning of bottleneck difficulty, people
The problems such as pain spot that work analysis efficiency is low, solution is directed to is single and deficiency.
Therefore, for problem as described above, this paper presents a whole set of to support offline, prototype real time environment, can be across flat
The data collection plan of platform and the detection of efficient application performance and diagnosis locating scheme.
The company with similar comprehensive performance analysis schemes being currently known has one --- Tencent, at present Tencent's exploitation
There are a set of UPA performance analysis tools, a subfunction is used as under WeTest platforms.UPA, that is, Unity Performance
Analysis is drawn for Unity by what the common research and development of Unity Support team and WeTest performances team of Tencent were made
The applied performance analysis tool for holding up customization is " upgrade version " of the former Cube tools of its exploitation.
It is the Comprehensive Correlation of two kinds of performance analysis schemes as shown in the table：
To sum up comparative analysis, it can be seen that Tencent focuses more on big and wide general-purpose platform, therefore the use of its tool to network,
The dependence on backstage is relatively high, and also and close friend is insufficient to developer user, most basic feature release not to
Family is free.
Invention content
The object of the present invention is to provide the diagnosis of a whole set of application performance and bottleneck location solution, mainly correspond to tripartite
The technical solution in face：
1, under non-Root patterns true mobile terminal environmental applications performance data collection, storage and exhibition scheme；
2, the data analysis display technique scheme of offline storage analysis and displaying is supported；
3, the backstage solution of true man's prototype professional test analysis is supported.
In order to realize the purpose, the present invention provides a kind of, and the application performance bottleneck based on clustering detects and diagnosis side
Method, the method utilize data collection and real-time exhibition module, offline client analysis and displaying under the true environment of mobile terminal
Statistical analysis module three parts are realized in module and line based on backstage.
Data collection and real-time exhibition module, this module provides under a whole set of real-time true environment data collection and
Exhibition scheme can collect properties data target information when operation, and i.e. while true environment runs and applies
When, easily show user's result：By application give out a contract for a project when, the data collection plug-in unit of integration packaging, you can application pacify
When shipping row, all kinds of different data are collected according to the demand of user, are shown.The implementation steps of the module are as follows：
The first step selectes each function to be detected at detection instrument interface；
Second step, according to each performance data Track modules of the unlatching of the function parallelization of selection, including：Screen capture module, Mono moulds
Block, resource module, hardware module etc.；
Third walks, and each data Track modules can call the API of corresponding function in Unity automatically, and acquisition in real time to be detected
Data information；
4th step just generates a two-dimentional buffer pool, by detection data and detection frame if enabling data real-time display function
Number is corresponding to be added in buffer pool, and calling interface refresh message, using on the data real-time update screen in buffer pool
Data are shown；
5th step is stored in local function if enabled, and each function module can be by current data, current data and present frame
The runtime datas such as several type, facility information, the system information of correspondence, current data are serialized and compress archive
It is saved in local file, while uploading to client development environment, for repetition, comparative analysis.
Offline client is analyzed and display module, and this module provides a whole set of off line data analysis tools, can be right
Data carry out off-line analysis when the application operation being collected into true environment, without sticking to fixed exploitation debugging enironment,
To simplify the work of research staff, also allow for finding to hide problem and bottleneck positioning：Pass through be collected into first part
Data upload under the arbitrary development environment configured with data analysis tool when application operation, you can the user provided according to tool
Operation interface is analyzed, checks results of performance analysis on demand.The specific implementation step of the module is as follows：
The first step, initialising subscriber operation interface.By initializing visual user base operation interface, response is based on user
The point selection operation of certain demand pops up corresponding dialog interface；The realization of user base operation interface is provided based on Unity3D engines
Programmable extension interface, inherit EditorWindow base class, realize that OnGUI methods form；
Second step, resource manager instance resource.Interface initialization after the completion of, by resource file management system carry out file,
The management of data, by loading and parse certain test result of user's selection, then by the data file under corresponding resource catalogue
Resource checksum is carried out using the hash algorithm of StrangeCRC, is unziped it using FastZip algorithms after verification is errorless, finally
The data set for the object model that all kinds of fundamental types are formed in memory is instantiated, is resolved in a manner of unserializing；
Third walks, statistical analysis and Classifying Sum.Statistical is carried out to the data set instantiated to the types of objects model of memory
The method of analysis and Classifying Sum, statistical analysis includes mean value, extreme value, normal distribution, trend analysis scheduling algorithm, thus by object mould
Type data set carries out the classification and extraction of the significant datas such as performance indicator；
4th step, shows analysis result in graphical form.The conclusion that analysis is obtained is finally direct in the form of Visual Chart
Feedback is shown to user interface and is presented to the user, and wherein Visual Chart includes that block diagram, line chart, radar map and table etc. are several
Kind common form.
Statistical analysis module on line based on backstage provides automated analysis and diagnosis scheme based on unsupervised learning
A whole set of backstage, reduce the participation and work of research staff, improve analysis and location efficiency：It is logged in and will be developed by backstage
Good is submitted to backstage using packet, that is, has professional and carry out careful test to the application, to right under black box pattern
Using careful data collection is carried out, eventually automatically by information such as the data performance indexs being collected into, with existing types of applications
Performance data information carry out comprehensive analysis, and provide detailed analytical conclusions and report.It can allow developer will more with this
More energy, which are put into, to be solved the problems, such as to navigate to, without being placed in the process of orientation problem, developer is cumbersome from redundancy
It is freed in work fine crushing, improves work and development efficiency.
The present invention has the advantage that：For the application developer of the overwhelming majority, it is no longer necessary to special Mac environment, you can
Convenient, fast and accurate applied performance analysis and problem diagnosis are carried out, unnecessary development cost is reduced；It no longer needs to use
Built-in Profiler tools go manually checking, analyzing indices frame by frame, the optimization time of application are reduced, so as to more
Early, it more timely puts goods on the market；It does not needing to be looked for according to a variety of different environment, replacing different analyzing and diagnosing tools,
Save time and the cost of optimizing phase.Further the stages such as the exploitation of application, test and optimization analysis are refined, from
And unnecessary work can be assigned to appropriate personnel, the development efficiency of application is improved, the stabilization before application is reached the standard grade is improved
Property；Simultaneously with the continuous increase of data scale, more data minings and analysis can be carried out based on these data informations, from
More macroscopical angle is come good and bad degree and priority degree, the market trend etc. of analyzing application.
Description of the drawings：
The following further describes the present invention with reference to the drawings：
Fig. 1：The package functional block diagram of the present invention
Fig. 2：The off-line tools model structural design figure of the present invention
Fig. 3：The off-line tools implementing procedure figure of the present invention
Fig. 4：Flow chart is embodied in the off-line tools of the present invention
Fig. 5：The class formation design drawing based on true environment of the present invention
Specific implementation mode：
Below in conjunction with the accompanying drawings and specific implementation mode the present invention will be described in detail：
As shown in Figures 1 to 5, a kind of detection of application performance bottleneck and diagnostic method, the method based on clustering utilize
Data collection under the true environment of mobile terminal is with real-time exhibition module, offline client analysis with display module and based on backstage
Statistical analysis module three parts are realized on line.
Data collection and real-time exhibition module, this module provides under a whole set of real-time true environment data collection and
Exhibition scheme can collect properties data target information when operation, and i.e. while true environment runs and applies
When, easily show user's result：By application give out a contract for a project when, the data collection plug-in unit of integration packaging, you can application pacify
When shipping row, all kinds of different data are collected according to the demand of user, are shown.The implementation steps of the module are as follows：
The first step selectes each function to be detected at detection instrument interface；
Second step, according to each performance data Track modules of the unlatching of the function parallelization of selection, including：Screen capture module, Mono moulds
Block, resource module, hardware module etc.；
Third walks, and each data Track modules can call the API of corresponding function in Unity automatically, and acquisition in real time to be detected
Data information；
4th step just generates a two-dimentional buffer pool, by detection data and detection frame if enabling data real-time display function
Number is corresponding to be added in buffer pool, and calling interface refresh message, using on the data real-time update screen in buffer pool
Data are shown；
5th step is stored in local function if enabled, and each function module can be by current data, current data and present frame
The runtime datas such as several type, facility information, the system information of correspondence, current data are serialized and compress archive
It is saved in local file, while uploading to client development environment, for repetition, comparative analysis.
Offline client is analyzed and display module, and this module provides a whole set of off line data analysis tools, can be right
Data carry out off-line analysis when the application operation being collected into true environment, without sticking to fixed exploitation debugging enironment,
To simplify the work of research staff, also allow for finding to hide problem and bottleneck positioning：Pass through be collected into first part
Data upload under the arbitrary development environment configured with data analysis tool when application operation, you can the user provided according to tool
Operation interface is analyzed, checks results of performance analysis on demand.The specific implementation step of the module is as follows：
The first step, initialising subscriber operation interface.By initializing visual user base operation interface, response is based on user
The point selection operation of certain demand pops up corresponding dialog interface；The realization of user base operation interface is provided based on Unity3D engines
Programmable extension interface, inherit EditorWindow base class, realize that OnGUI methods form；
Second step, resource manager instance resource.Interface initialization after the completion of, by resource file management system carry out file,
The management of data, by loading and parse certain test result of user's selection, then by the data file under corresponding resource catalogue
Resource checksum is carried out using the hash algorithm of StrangeCRC, is unziped it using FastZip algorithms after verification is errorless, finally
The data set for the object model that all kinds of fundamental types are formed in memory is instantiated, is resolved in a manner of unserializing；
Third walks, statistical analysis and Classifying Sum.Statistical is carried out to the data set instantiated to the types of objects model of memory
The method of analysis and Classifying Sum, statistical analysis includes mean value, extreme value, normal distribution, trend analysis scheduling algorithm, thus by object mould
Type data set carries out the classification and extraction of the significant datas such as performance indicator；
4th step, shows analysis result in graphical form.The conclusion that analysis is obtained is finally direct in the form of Visual Chart
Feedback is shown to user interface and is presented to the user, and wherein Visual Chart includes that block diagram, line chart, radar map and table etc. are several
Kind common form.
Statistical analysis module on line based on backstage provides automated analysis and diagnosis scheme based on unsupervised learning
A whole set of backstage, reduce the participation and work of research staff, improve analysis and location efficiency：It is logged in and will be developed by backstage
Good is submitted to backstage using packet, that is, has professional and carry out careful test to the application, to right under black box pattern
Using careful data collection is carried out, eventually automatically by information such as the data performance indexs being collected into, with existing types of applications
Performance data information carry out comprehensive analysis, and provide detailed analytical conclusions and report.It can allow developer will more with this
More energy, which are put into, to be solved the problems, such as to navigate to, without being placed in the process of orientation problem, developer is cumbersome from redundancy
It is freed in work fine crushing, improves work and development efficiency.
It is emphasized that：It the above is only presently preferred embodiments of the present invention, not the present invention made in any form
Limitation, it is every according to the technical essence of the invention to any simple modification, equivalent change and modification made by above example,
Still fall within the range of technical solution of the present invention.
Claims (3)
1. a kind of detection of application performance bottleneck and diagnostic method based on clustering, it is characterised in that：The method utilizes shifting
Data collection under the true environment of moved end and real-time exhibition module, offline client analysis and display module and the line based on backstage
Upper statistical analysis module three parts realization,
The data collection and real-time exhibition module, provide the data collection under a whole set of real-time true environment and displaying side
Case collects properties data target information when operation, and user is showed to tie while true environment runs and applies
Fruit, by the way that when application is given out a contract for a project, the data collection plug-in unit of integration packaging is right according to the demand of user in application installation and operation
All kinds of different data are collected, show；
The offline client analysis and display module, provide a whole set of off line data analysis tool, in true environment
Data carry out off-line analysis when the application operation being collected into, and are transported by the application for arriving data collection and real-time exhibition module collection
Data upload under the arbitrary development environment configured with data analysis tool when row, according to tool provide user interface into
Row is analyzed, checks results of performance analysis on demand；
Statistical analysis module on the line based on backstage provides automated analysis and diagnosis scheme based on unsupervised learning
A whole set of backstage, logged in by backstage and what will be develop be submitted to backstage using packet, to this using the careful test of progress,
To carry out data collection to application under black box pattern, the data performance indication information being finally collected into is all kinds of with having
The performance data information of application carries out comprehensive analysis, and provides detailed analytical conclusions and report.
2. a kind of detection of application performance bottleneck and diagnostic method based on clustering according to claim 1, feature
It is：The data collection and the implementation steps of real-time exhibition module are as follows：
The first step selectes each function to be detected at detection instrument interface；
Second step, according to each performance data Track modules of the unlatching of the function parallelization of selection, including：Screen capture module, Mono moulds
Block, resource module, hardware module；
Third walks, and each data Track modules can call the API of corresponding function in Unity automatically, and acquisition in real time to be detected
Data information；
4th step just generates a two-dimentional buffer pool, by detection data and detection frame if enabling data real-time display function
Number is corresponding to be added in buffer pool, and calling interface refresh message, using on the data real-time update screen in buffer pool
Data are shown；
5th step is stored in local function if enabled, and each function module can be by current data, current data and present frame
The runtime datas such as several type, facility information, the system information of correspondence, current data are serialized and compress archive
It is saved in local file, while uploading to client development environment, for repetition, comparative analysis.
3. a kind of detection of application performance bottleneck and diagnostic method based on clustering according to claim 1, feature
It is：The offline client analysis is as follows with the implementation steps of display module：
The first step, initialising subscriber operation interface, by initializing visual user base operation interface, response is based on user
The point selection operation of certain demand pops up corresponding dialog interface；The realization of user base operation interface is provided based on Unity3D engines
Programmable extension interface, inherit EditorWindow base class, realize that OnGUI methods form；
Second step, resource manager instance resource, interface initialization after the completion of, by resource file management system carry out file,
The management of data, by loading and parse certain test result of user's selection, then by the data file under corresponding resource catalogue
Resource checksum is carried out using the hash algorithm of StrangeCRC, is unziped it using FastZip algorithms after verification is errorless, finally
The data set for the object model that all kinds of fundamental types are formed in memory is instantiated, is resolved in a manner of unserializing；
Third walks, statistical analysis and Classifying Sum, and statistical is carried out to the data set instantiated to the types of objects model of memory
The method of analysis and Classifying Sum, statistical analysis includes mean value, extreme value, normal distribution, trend analysis algorithm, thus by object model
Data set carries out the classification and extraction of performance indicator significant data；
4th step, shows analysis result in graphical form, and the conclusion that analysis is obtained is finally direct in the form of Visual Chart
Feedback is shown to user interface and is presented to the user, and wherein Visual Chart includes the normal of block diagram, line chart, radar map and table
See form.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810319392.6A CN108563568A (en) | 2018-04-11 | 2018-04-11 | A kind of detection of application performance bottleneck and diagnostic method based on clustering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810319392.6A CN108563568A (en) | 2018-04-11 | 2018-04-11 | A kind of detection of application performance bottleneck and diagnostic method based on clustering |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108563568A true CN108563568A (en) | 2018-09-21 |
Family
ID=63534501
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810319392.6A Pending CN108563568A (en) | 2018-04-11 | 2018-04-11 | A kind of detection of application performance bottleneck and diagnostic method based on clustering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108563568A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113626073A (en) * | 2021-08-06 | 2021-11-09 | 航天中认软件测评科技（北京）有限责任公司 | Software adaptation optimization method based on knowledge base |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105786864A (en) * | 2014-12-24 | 2016-07-20 | 国家电网公司 | Offline analysis method for massive data |
| US20160278927A1 (en) * | 2010-04-15 | 2016-09-29 | Zimmer, Inc. | Methods of ordering and manufacturing orthopedic components |
-
2018
- 2018-04-11 CN CN201810319392.6A patent/CN108563568A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160278927A1 (en) * | 2010-04-15 | 2016-09-29 | Zimmer, Inc. | Methods of ordering and manufacturing orthopedic components |
| CN105786864A (en) * | 2014-12-24 | 2016-07-20 | 国家电网公司 | Offline analysis method for massive data |
Non-Patent Citations (1)
| Title |
|---|
| UNITY官方: "UPA性能分析工具使用详解", 《UPA性能分析工具使用详解 HTTPS://WETEST.QQ.COM/LAB/VIEW/358.HTML》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113626073A (en) * | 2021-08-06 | 2021-11-09 | 航天中认软件测评科技（北京）有限责任公司 | Software adaptation optimization method based on knowledge base |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110852347B (en) | Fire detection method using improved YOLO v3 | |
| CN106331361B (en) | A kind of protection processing method that mobile terminal falls, apparatus and system | |
| CN106202514A (en) | Accident based on Agent is across the search method of media information and system | |
| CN107679189A (en) | A kind of point of interest update method, device, server and medium | |
| CN109451532A (en) | A kind of check method and device of base station location | |
| CN106355115B (en) | A kind of drop information processing method and processing device | |
| CN110379036A (en) | Intelligent substation patrol recognition methods, system, device and storage medium | |
| CN106886517A (en) | Business site selecting method, device and system | |
| CN108563568A (en) | A kind of detection of application performance bottleneck and diagnostic method based on clustering | |
| CN104657248A (en) | Java thread stack analysis method and system | |
| CN115203194A (en) | Metadata information generation method, related device, equipment and storage medium | |
| CN114238062B (en) | Board card burning device performance analysis method, device, equipment and readable storage medium | |
| CN109302305A (en) | A kind of clustering performance test method based on industrial control system | |
| CN116166889B (en) | Hotel product screening method, device, equipment and storage medium | |
| CN115801845B (en) | Industrial Internet data acquisition method and related equipment based on edge calculation | |
| CN113191455B (en) | Edge computing box election method and device, electronic equipment and medium | |
| CN116822804B (en) | Digital asset management analysis method, device and medium | |
| CN115078952B (en) | IGBT driving fault detection method and system | |
| CN116308158B (en) | National asset supervision and management system | |
| CN116484230B (en) | Method for identifying abnormal business data and training method of AI digital person | |
| CN110888775B (en) | Method, device and equipment for monitoring data warehouse by utilizing data balance | |
| CN115809832A (en) | Intelligent comprehensive benefit analysis method, device and equipment for distributed photovoltaic project | |
| Chen et al. | Study and Design an Equipment-fault-diagnosis System Based on 3D Draft Software | |
| CN116502120A (en) | Data screening method, device and system | |
| CN117010218A (en) | Map rendering method, device, equipment and medium of power grid equipment |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180921 |
|
| WD01 | Invention patent application deemed withdrawn after publication |