CN107967204A - Line pushes method, system and the terminal device surveyed - Google Patents
Line pushes method, system and the terminal device surveyed Download PDFInfo
- Publication number
- CN107967204A CN107967204A CN201711164547.5A CN201711164547A CN107967204A CN 107967204 A CN107967204 A CN 107967204A CN 201711164547 A CN201711164547 A CN 201711164547A CN 107967204 A CN107967204 A CN 107967204A
- Authority
- CN
- China
- Prior art keywords
- data
- line
- request
- under
- performance
- 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
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
- G06F11/34—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
- G06F11/3457—Performance evaluation by simulation
Abstract
The present invention proposes that a kind of line pushes method, system and the terminal device surveyed.The described method includes acquisition user's required parameter from line upper module, and copy under line；The user's required parameter obtained to duplication, carries out data processing and signature analysis；Based on intelligent algorithm, machine learning is carried out to the data after data processing and signature analysis, forms Policy model；Based on Policy model, with reference to the data stored, different test scenes is formed, and plays back test scene, obtains under different test scenes performance data under accurate line.The present invention come corresponding system performance on the line of prediction, is improved by performance data under line and utilizes under line test to the accuracy of inline system Performance Evaluation.
Description
Technical field
The present invention relates to flow and the pressure test field of data, more particularly, to line push the method surveyed, system and
Terminal device.
Background technology
With internet product form, the continuous development of framework, and the raising of business complexity, internet product system
Performance issue, it is caused influence it is also increasing.This not only seriously damages Product Experience, can also cause serious society's warp
Ji loss influences.For example social class website can cause partial service unavailable because system performance problems；The website meeting of booking class
Because system performance problems, cause a large number of users can not complete booking etc..Although various product can all carry out respectively before reaching the standard grade
The performance test of type and rank, but due to the performance data that performance test is drawn under line, it is applied to each scene of true environment
When, accuracy is not high, still results in performance issue on serious line.Can performance test accurately correspond to feelings on line under line
Condition, depending in performance test under line, whether the emulation to test scene true and each offline lower performance data of scene with line
Whether the correspondence of performance data is accurate credible.
Currently, under various lines in performance test methods, in order to obtain under line performance data often through simple structure case
Part carrys out emulation testing scene, the situation of the Traffic Anomaly of this method real traffic scene various to magnanimity on line or burst
Simulate less so that it is inherently inaccurate that the data drawn are tested under line.And obtain under line after data, often by artificial warp
Test the difference based on environment on environment under line and line, using under line in data assessment outlet environment performance data, in product shelf
In the case that reciprocal effect is more complicated between structure and different system, the error bigger of this estimation, data under inaccurate line
Plus the correspondence error of bigger, cause the more inaccurate of performance data on line.Although there are some using environment on line or
The test method of environment in a set of same line is built under online completely, but due to the influence to online service and operability
Difference, leads to not extensive use, and as the increasingly complexity of product architecture and resource environment becomes more difficult.
Therefore, a system accurately how is online assessed in real traffic or various exceptions during lower test
Actual performance under scene is into a problem.
The content of the invention
The embodiment of the present invention provides a kind of line and pushes method, system and the terminal device surveyed, existing at least to solve or alleviate
There is one or more of technology technical problem, provide at a kind of beneficial selection.
In a first aspect, the method surveyed is pushed an embodiment of the present invention provides a kind of line.
In the first embodiment of first aspect, line pushes the method surveyed to be included the present invention：Obtained from line upper module
The request data at family is taken, and is copied under line；The request data obtained to duplication, carries out data processing and signature analysis；Base
In intelligent algorithm, machine learning is carried out to the data after data processing and signature analysis, forms Policy model；Based on plan
Slightly model, with reference to request data, forms test scene under different lines；Lower environment plays back checkout area under each line online
Scape, obtains performance data under the line corresponding to it.
The present invention carries out machine learning, analysis environments in second of embodiment of first aspect, according to environmental characteristic
Relation between feature and performance data, forming properties data correction model；Based on correction model, under the line of test scene under line
Performance data is modified, to obtain and performance data on the corresponding line of environment on line.
The first implementation or second of implementation of first aspect with reference to first aspect, obtain user and ask ginseng
Count and copy under line, including：The Hook Function provided using system netfilter frames, directly replicates user's from IP layers
Under required parameter to line；The request data package replicated is handled, only returns and maintains content needed for tcp transmission.
The first implementation or second of implementation of first aspect with reference to first aspect, the use obtained to duplication
The request data at family, carries out data processing and signature analysis, including：Obtained in the request data obtained from duplication and be used for engineering
The single request feature and combination request feature of habit；Wherein, the performance to line upper module that single request feature includes once asking has
The feature of influence, combination request feature includes repeatedly interaction or the performance to line upper module repeatedly asked in a period has an impact
Distribution characteristics.
Intelligent algorithm is preferably based on, machine learning, shape are carried out to the data after data processing and signature analysis
Into Policy model, specifically include：By acquired single request feature and combination request feature, the algorithm based on machine learning,
The characteristics of analyzing the request data of user and the regularity of distribution of request data, to form Policy model.
Second of implementation with reference to first aspect, the environmental characteristic include at least one below：Hardware resource, net
System event intervention in network resource, fabric topology complexity, program language, resource isolation mode, and abnormal scene, it is automatic dry
In advance, the time of manual intervention, flow scheduling strategy.
Second aspect, the system surveyed is pushed an embodiment of the present invention provides a kind of line.
In the first embodiment of second aspect, line pushes the system surveyed to be included the present invention：Data capture unit, matches somebody with somebody
Put the request data for obtaining user from line upper module and replicate the data；Data processing unit, is configured to receive number
The request data replicated according to acquiring unit, and carry out data processing and signature analysis；Model computing unit, is configured to receive
Data after data processing unit carries out data processing and signature analysis, and machine learning is carried out to the data, form strategy
Model；Data readback unit, is configured to according to Policy model, with reference to request data, forms test scene under different lines, and
Lower environment plays back test scene under each line online, obtains performance data under the line corresponding to it.
With reference to the first implementation of second aspect, data capture unit is set on the server, utilizes system
The Hook Function that netfilter frames provide, and be configured to from the IP layers of request data for directly acquiring user and replicate the number
According to.
With reference to the first implementation of second aspect, the data processing of data processing unit includes：Obtained from duplication
The single request feature and combination request feature for machine learning are obtained in request data；Wherein, single request feature is included once
The influential feature of the performance on line upper module of request, combination request feature include repeatedly asking in repeatedly interaction or a period
The influential distribution characteristics of the performance on line upper module.
With reference to the first implementation of second aspect, model computing unit, is configured specifically for receiving by data processing
Single request feature and combination request feature, the homing method based on machine learning acquired in unit, analyze the number of request of user
According to the characteristics of, to form Policy model.
With reference to second of implementation of second aspect, model computing unit, is further configured to according to environmental characteristic
Machine learning is carried out, relation between analysis environments feature and data, forms data correction model.
Preferably, the environmental characteristic includes at least one below：Hardware resource, Internet resources, fabric topology complexity,
System event intervention in program language, resource isolation mode, and abnormal scene, automatic to intervene, the time of manual intervention, flow
Scheduling strategy.
With reference to second of implementation of second aspect, the system also includes data correction unit, is configured to receive
Performance data under the line of test scene under correction model and line, and according to correction model to performance number under the line of test scene under line
According to being modified, to obtain and performance data on the corresponding line of environment on line.
The third aspect, the terminal device surveyed, including one or more processing are pushed an embodiment of the present invention provides a kind of line
Device；Storage device, for storing one or more programs；Communication interface, is configured to be led between memory and processor
Letter；When one or more of programs are performed by one or more of processors so that one or more of processors
Realize the method as described in any implementation in above-mentioned first aspect.
A technical solution in above-mentioned technical proposal has the following advantages that or beneficial effect：By by data duplication on line
To under line, data processing and signature analysis are carried out, based on intelligent algorithm, Policy model is formed and forms difference with reference to data
Test scene under line, and Performance Evaluation under line is improved come forecasting system performance by test scene under request data and line
Accuracy.
Above-mentioned general introduction is merely to illustrate that the purpose of book, it is not intended to is limited in any way.Except foregoing description
Schematical aspect, outside embodiment and feature, it is further by reference to attached drawing and the following detailed description, the present invention
Aspect, embodiment and feature would is that what is be readily apparent that.
Brief description of the drawings
In the accompanying drawings, unless specified otherwise herein, otherwise represent the same or similar through the identical reference numeral of multiple attached drawings
Component or element.What these attached drawings were not necessarily to scale.It should be understood that these attached drawings depict only according to the present invention
Some disclosed embodiments, and should not serve to limit the scope of the present invention.
Fig. 1 shows the flow chart that the method 100 surveyed is pushed according to the line of the application one embodiment；
Fig. 2 shows the flow chart that the method 200 surveyed is pushed according to the line of the application another embodiment；
Fig. 3 shows the structure diagram that the system 300 surveyed is pushed according to the line of the application one embodiment；
Fig. 4 shows the structure diagram that the system 400 surveyed is pushed according to the line of the application another embodiment；
Fig. 5 shows the schematic diagram of the terminal device according to the application.
Embodiment
Hereinafter, some exemplary embodiments are simply just described.As one skilled in the art will recognize that
Like that, without departing from the spirit or scope of the present invention, described embodiment can be changed by various different modes.
Therefore, attached drawing and description are considered essentially illustrative rather than restrictive.
Refering to what is shown in Fig. 1, the line of one embodiment of the application first aspect pushes the method 100 surveyed, by by line
Data duplication is under line, and to index of correlation therein, for example, single request feature and combination request feature, carry out equipment analysis
With study, to obtain relevant model, and the model is based on, with reference to request data, to construct line test scene.
Line pushes the method 100 surveyed, and specifically includes following steps S101-S104 '.
S101, obtains the request data of user from line upper module, and copies under line.In S101 steps, do not influencing
On line on the premise of user's request, the flow that will be formed in line upper module by the request data of user, packing is copied under line.
Preferably, can utilize system netfilter frames provide Hook Function, from IP layer directly duplication user request data to
Under line, realize the decoupling with upper layer transport protocol, make acquisition methods that there is certain versatility.Also, the request to having replicated
Data are handled, i.e. the copy package of request data that processing replicates so that server not to the data in copy package into
Row calculates analysis and feedback, so as to will not affect that the feedback that the input to user is asked.Also, to the request data replicated
Bag is handled, and is only returned to content needed for maintaining tcp transmission, is avoided impacting online service.
S102, the request data obtained to duplication, carries out data processing and signature analysis.Wherein, data processing and feature
Analysis specifically includes：Obtain the single request feature and combination request feature for machine learning.
Wherein, single request feature includes the feature once asked, and specifically, a line upper module receives the request of processing
There may be polytype, the required parameter, request type in once asking take relative to request processing, network, machine resources
Deng consumption be all it is closely related, we it is this once ask in the influential feature of line upper module performance is known as it is single please
Seek feature.For example, inquiry request, renewal request etc. are single request feature.
Combination request feature includes repeatedly interaction or the distribution characteristics repeatedly asked in a period, specifically, once complete
Whole user's request may relate to repeatedly interacting between server, they have time sequencing, be asked since single is different
Ask the consumption to performance different, so the combination of different requests has a great influence the performance of line upper module, we are this one
The influential distribution characteristics of the performance on line upper module repeatedly asked in time phase is known as combining request feature.For example, one
A complete purchase operation may successively be related to multiple inquiry between server, exiting and continue, prop up several times
Pay and acknowledgement of orders etc. is just finally completed.
Data after data processing and signature analysis based on intelligent algorithm, are carried out machine learning, are formed by S103
Policy model.In one embodiment, specifically include：By acquired single request feature and combination request feature, based on machine
The regularity of distribution of the algorithm of device study, the characteristics of analyzing the request data of user and request data, forms Policy model.S103 is walked
In rapid, after acquired single request feature and combination request feature, the features such as machine, deployment topologies can be combined, carry out relevant mode
The machine learning of type, searches out the species and the regularity of distribution of single request feature and combination request feature, is eventually converted into various fields
Policy model under scape.
S104, based on Policy model, with reference to request data, forms test scene under different lines.Simulating various scenes
Pressure test when, except needing the Policy model set up in advance, it is also necessary to the request data having.Wherein, request data can
To be that the data for obtaining and storing are replicated from line upper mold request data in the block or in S101 steps.
S104 ', test scene under each line of lower environment playback, obtains performance data under the line corresponding to it online.
According to Policy model, request data is combined, is finally played back under the line corresponding to Policy model in test scene, so that
Obtain performance data under the line under the line under test scene.
In conclusion line pushes the method 100 surveyed, by under data duplication on real line to line, and on real line
Data are analyzed and learnt, and obtain relevant Policy model, can be more by the Policy model and request data of truthful data
Add and really emulate all kinds of test scenes.This method 100 can be effectively used in testing service platform, met under user's line
Performance test accuracy lifting demand, and under more convenient line performance test progress, improving performance test accuracy with
Timeliness.
Refering to what is shown in Fig. 2, the line of another embodiment of the application first aspect pushes the method 200 surveyed, by by line
Upper data duplication carries out equipment analysis and study under line to index of correlation therein, for example, single request feature and combination please
Feature, and the environmental index such as hardware resource, environment deployment are asked, so as to obtain relevant Policy model and correction model.It is based on
Policy model constructs test scene under line, obtains performance data under line, performance data under line is carried out further according to correction model
Correct, obtain performance data on the corresponding line of environment on line.
Line pushes the method 200 surveyed, and specifically includes following steps S101-S106.
S101, obtains the request data of user from line upper module, and copies under line.
S102, the request data of the user obtained to duplication, carries out data processing and signature analysis.
Data after data processing and signature analysis based on intelligent algorithm, are carried out machine learning, are formed by S103
Policy model.
S104, based on Policy model, with reference to request data, forms test scene under different lines.
S104 ', test scene under each line of lower environment playback, obtains performance data under the line corresponding to it online.
Above-mentioned S101 is corresponding to S104 ' steps with the S101 in method 100 and hereinbefore right to S104 ' steps
It is illustrated, and is repeated no more herein.
Online in lower test process, although true environment can be simulated to greatest extent by carrying out playback by Policy model
Scene, but since the limitation of machine resources (environmental characteristic) causes with actual production environment there are larger difference, in order to make line
Under test scene can accurately correspond to environment on line, combining environmental feature is also needed in method 200 to test scene
It is modified.First, correction model is set up, is comprised the following steps that：
S105, carries out machine learning according to environmental characteristic and performance data, is closed between analysis environments feature and performance data
System, forms data correction model.Wherein, environmental characteristic includes at least one below：Hardware resource, Internet resources, fabric topology
System event intervention in complexity, program language, resource isolation mode, and abnormal scene, it is automatic to intervene, manual intervention when
Between, flow scheduling strategy.
The correcting action of correction model is on test scene, i.e. in following steps：
S106, after test scene under obtaining line, based on correction model, to performance data under the line of test scene under line into
Row is corrected, corresponding with environment on line to obtain.For example, be 10qps corresponding to the performance data under environmental testing scene under line,
And due to the influence of environmental characteristic, 100qps can be supported by corresponding under scene on line.At this point it is possible to by environmental characteristic
The correction model for learning and being formed, is modified the performance data 10qps under test scene under line, and obtain and performance on line
Data can support and predict scene on the line that 100qps is close.
In conclusion line pushes the method 200 surveyed, by under data duplication on real line to line, and on real line
Data are analyzed and learnt, and obtain relevant Policy model, can be more by the Policy model and request data of truthful data
Add and really emulate all kinds of test scenes.Further, can be under relatively simple line by the correction model of environmental characteristic
In environment, more really reflect performance data on the line on the line of complexity under environment, and its corresponding prediction scene.This
The progress of performance test under 200 more convenient line of method, and the further accuracy of improving performance test.
Refering to what is shown in Fig. 3, one embodiment of the second aspect of the application, there is provided under the corresponding line of Fig. 1 the methods
Press examining system 300.Line pushes examining system 300 and specifically includes data capture unit, data processing unit, model computing unit sum number
According to playback unit.
Data capture unit, is configured to obtain the request data of user from line upper module, and the data are packed again
Make under line.Wherein, data capture unit can utilize the Hook Function provided in netfilter frames, directly be obtained from IP layers
Take the request data at family and replicate the data.It is thus possible to realize the decoupling with upper layer transport protocol, there are acquisition methods
Certain versatility.Also, the request data to having replicated is handled, i.e. the duplication for the request data that processing replicates
Bag so that server does not carry out copy package to calculate analysis and feedback, so as to will not affect that the anti-of the input request to user
Feedback.Also, the request data package to having replicated is handled, content needed for maintaining tcp transmission is only returned, is avoided to online service
Impact.
Data processing unit, is configured to receive the request data that data capture unit is replicated, and carries out data processing
And signature analysis.Data processing unit can include obtaining single request feature for machine learning to data processing and analysis
Feature is asked with combination.Wherein, single request feature and combination request feature in the S102 steps of the above method 100 it is stated that,
Repeated no more at this.
Model computing unit, is configured to receive the number after data processing unit carries out data processing and signature analysis
According to, and machine learning is carried out to the data, form Policy model.Specifically, model computing unit is received by data processing unit
Acquired single request feature and combination request feature, by the algorithm of machine learning, the characteristics of analyzing the request data of user
With the regularity of distribution of request data, search out single request feature and the species and the regularity of distribution of feature are asked in combination, to form plan
Slightly model.
Data readback unit, is configured to according to Policy model, with reference to request data, forms checkout area under different lines
Scape, and descend environment to play back test scene under each line online, obtain performance data under the line corresponding to it.It is various simulating
During the pressure test of scene, except needing the Policy model set up in advance, it is also necessary to the request data having.Wherein, number of request
According to the number that can be acquired data from line upper module or be replicated to obtain by data capture unit and stored
According to.According to Policy model, request data is combined, is finally played back in the scene corresponding to Policy model, so as to obtain
The test scene of the data.
In conclusion line pushes the system 300 surveyed, by under data duplication on real line to line, and on real line
Data are analyzed and learnt, and obtain relevant Policy model, can be more by the Policy model and request data of truthful data
Add and really emulate all kinds of test scenes.The system 300 can be effectively used in testing service platform, met under user's line
Performance test accuracy lifting demand, and under more convenient line performance test progress, improving performance test accuracy with
Timeliness.
Refering to what is shown in Fig. 4, another embodiment of the second aspect of the application, there is provided the corresponding line of Fig. 2 the methods
Push examining system 400.Line pushes examining system 400 and specifically includes data capture unit, data processing unit, model computing unit,
Data readback unit and data amending unit.Wherein, herein in data capture unit and data processing unit and system 300
Data capture unit and data processing unit are corresponding, and have hereinbefore been described, are repeated no more at this.
Model computing unit, is configured to receive the number by after data processing unit progress data processing and signature analysis
According to, and machine learning is carried out to the data, form Policy model；Also, further it is configured to according to environmental characteristic and performance
Data carry out machine learning, and relation between analysis environments feature and performance data, forms data correction model.
In one embodiment, specifically, model computing unit receives special as single request acquired in data processing unit
Seek peace combination request feature, by the algorithm of machine learning, the characteristics of analyzing the request data of user, search out single request feature
The species and the regularity of distribution of feature are asked with combination, to form Policy model.Meanwhile model computing unit can also be according to hardware
System event is done in resource, Internet resources, fabric topology complexity, program language, resource isolation mode, and abnormal scene
In advance, automatic to intervene, the environmental characteristic such as time of manual intervention, flow scheduling strategy carries out machine learning, analysis environments feature with
Relation between data, forms data correction model.
Data readback unit, is configured to according to Policy model, bonding wire upper mold request data in the block or is obtained by data
The data that unit is replicated are taken, are analyzed, form test scene under different lines.Specifically, the pressure of various scenes is being simulated
When power is tested, except needing the Policy model set up in advance, it is also necessary to the request data having.Wherein, request data can be
Acquired data or the data replicated by data capture unit from line upper module.It is right according to Policy model
Request data is combined, and is finally played back in the scene corresponding to Policy model, so as to obtain checkout area under the line of the data
Scape.
Data correction unit, is configured to receive test scene under correction model and line, and according to correction model under line
Performance data is modified under the line of test scene.Specifically, after data correction unit obtains test scene, based on its gained
The correction model arrived, is modified performance data under line, to obtain performance data on line corresponding with environment on line, formed
Different prediction scenes.
In conclusion line pushes the system 400 surveyed, by under data duplication on real line to line, and on real line
Data are analyzed and learnt, and obtain relevant Policy model, can be more by the Policy model and request data of truthful data
Add and really emulate test scene under all kinds of lines.Further, can be relatively simple by the correction model of environmental characteristic
Under line in environment, more really reflect performance data on the line on the line of complexity under environment, and its corresponding prediction field
Scape.The progress of performance test under 400 more convenient line of the system, and the further accuracy and timeliness of improving performance test
Property.
The third aspect of the application provides a kind of line and pushes the terminal device surveyed, as shown in figure 5, including one or more
Processor；Storage device, for storing one or more programs.When one or more of programs are by one or more of places
When managing device execution so that one or more of processors realize the side as described in any in the above method 100 or method 200
Method.
Wherein, memory and processor quantity can be one or more.
The equipment further includes：
Communication interface, is configured to make processor and memory communicate with external equipment.
Memory may include high-speed RAM memory, it is also possible to further include nonvolatile memory (non-volatile
Memory), a for example, at least magnetic disk storage.
If memory, processor and communication interface are independently realized, memory, processor and communication interface can pass through
Bus is connected with each other and completes mutual communication.The bus can be industry standard architecture (ISA, Industry
Standard Architecture) bus, external equipment interconnection (PCI, Peripheral Component) bus or extension work
Industry standard architecture (EISA, Extended Industry Standard Component) bus etc..The bus can be with
It is divided into address bus, data/address bus, controlling bus etc..For ease of representing, only represented in Fig. 5 with a thick line, it is not intended that
Only a bus or a type of bus.
Optionally, in specific implementation, if processor, memory and communication interface integrate on one chip, locate
Reason device, memory and communication interface can complete mutual communication by internal interface.
In the description of this specification, reference term " one embodiment ", " some embodiments ", " example ", " specifically show
The description of example " or " some examples " etc. means specific features, structure, material or the spy for combining the embodiment or example description
Point is contained at least one embodiment of the present invention or example.Moreover, particular features, structures, materials, or characteristics described
It may be combined in any suitable manner in any one or more of the embodiments or examples.In addition, without conflicting with each other, this
The technical staff in field can be by the different embodiments or example described in this specification and different embodiments or exemplary spy
Sign is combined and combines.
In addition, term " first ", " second " are only used for description purpose, and it is not intended that instruction or hint relative importance
Or the implicit quantity for indicating indicated technical characteristic.Thus, " first " is defined, the feature of " second " can be expressed or hidden
Include at least one this feature containing ground.In the description of the present invention, " multiple " are meant that two or more, unless otherwise
It is clearly specific to limit.
Any process or method described otherwise above description in flow chart or herein is construed as, and represents to include
Module, fragment or the portion of the code of the executable instruction of one or more the step of being used for realization specific logical function or process
Point, and the scope of the preferred embodiment of the present invention includes other realization, wherein can not press shown or discuss suitable
Sequence, including according to involved function by it is basic at the same time in the way of or in the opposite order, carry out perform function, this should be of the invention
Embodiment person of ordinary skill in the field understood.
Expression or logic and/or step described otherwise above herein in flow charts, for example, being considered use
In the order list for the executable instruction for realizing logic function, may be embodied in any computer-readable medium, for
Instruction execution system, device or equipment (such as computer based system including the system of processor or other can be held from instruction
The system of row system, device or equipment instruction fetch and execute instruction) use, or combine these instruction execution systems, device or set
It is standby and use.For the purpose of this specification, " computer-readable medium " can any can be included, store, communicate, propagate or pass
Defeated program is for instruction execution system, device or equipment or the dress used with reference to these instruction execution systems, device or equipment
Put.The more specifically example (non-exhaustive list) of computer-readable medium includes following：Electricity with one or more wiring
Connecting portion (electronic device), portable computer diskette box (magnetic device), random access memory (RAM), read-only storage
(ROM), erasable edit read-only storage (EPROM or flash memory), fiber device, and portable read-only storage
(CDROM).In addition, computer-readable medium can even is that the paper that can print described program on it or other suitable Jie
Matter, because can be for example by carrying out optical scanner to paper or other media, then into edlin, interpretation or if necessary with other
Suitable method is handled electronically to obtain described program, is then stored in computer storage.
It should be appreciated that each several part of the present invention can be realized with hardware, software, firmware or combinations thereof.Above-mentioned
In embodiment, software that multiple steps or method can be performed in memory and by suitable instruction execution system with storage
Or firmware is realized.If, and in another embodiment, can be with well known in the art for example, realized with hardware
Any one of row technology or their combination are realized：With the logic gates for realizing logic function to data-signal
Discrete logic, have suitable combinational logic gate circuit application-specific integrated circuit, programmable gate array (PGA), scene
Programmable gate array (FPGA) etc..
Those skilled in the art are appreciated that to realize all or part of step that above-described embodiment method carries
Suddenly it is that relevant hardware can be instructed to complete by program, the program can be stored in a kind of computer-readable storage medium
In matter, the program upon execution, including one or a combination set of the step of embodiment of the method.
In addition, each functional unit in each embodiment of the present invention can be integrated in a processing module, can also
That unit is individually physically present, can also two or more units be integrated in a module.Above-mentioned integrated mould
Block can both be realized in the form of hardware, can also be realized in the form of software function module.The integrated module is such as
Fruit is realized in the form of software function module and as independent production marketing or in use, can also be stored in a computer
In readable storage medium storing program for executing.The storage medium can be read-only storage, disk or CD etc..
The above description is merely a specific embodiment, but protection scope of the present invention is not limited thereto, any
Those familiar with the art the invention discloses technical scope in, its various change or replacement can be readily occurred in,
These should be covered by the protection scope of the present invention.Therefore, protection scope of the present invention should be with the guarantor of the claim
Protect subject to scope.
Claims (14)
1. a kind of line pushes the method surveyed, it is characterised in that including：
On the premise of user asks without influence, the request data of user is obtained from line upper module, and is copied under line；
The request data obtained to duplication, carries out data processing and signature analysis；
Based on intelligent algorithm, machine learning is carried out to the data after data processing and signature analysis, forms Policy model；
Based on Policy model, with reference to request data, test scene under different lines is formed；
Lower environment plays back test scene under each line online, obtains performance data under the line corresponding to it.
2. according to the method described in claim 1, it is characterized in that, further include：
Machine learning is carried out according to environmental characteristic and performance data, relation between analysis environments feature and performance data, forms number
According to correction model；
Based on correction model, performance data under the line of test scene under line is modified, it is corresponding with environment on line to obtain
Line on performance data.
3. method according to claim 1 or 2, it is characterised in that wherein, obtain the request data of user and copy to line
Under, including：
Using the Hook Function provided in netfilter frames, under request data to the line that user is directly replicated from IP layers；
The request data package replicated is handled, only returns and maintains content needed for tcp transmission.
4. method according to claim 1 or 2, it is characterised in that wherein, the request data of the user obtained to duplication,
Data processing and signature analysis are carried out, including：
The single request feature and combination request feature for machine learning are obtained in the request data obtained from duplication；Wherein, it is single
Request feature includes the influential feature of the performance on line upper module once asked, combination request feature include repeatedly interaction or
The influential distribution characteristics of the performance on line upper module repeatedly asked in one period.
5. according to the method described in claim 4, it is characterized in that, wherein, based on intelligent algorithm, to through data processing and
Data after signature analysis carry out machine learning, form Policy model, including：
Ask feature by acquired list and combine to ask feature, the algorithm based on machine learning, analyze the number of request of user
According to the characteristics of and request data the regularity of distribution, to form Policy model.
6. according to the method described in claim 2, it is characterized in that, wherein, the environmental characteristic includes at least one below：
System in hardware resource, Internet resources, fabric topology complexity, program language, resource isolation mode, and abnormal scene
Event intervention, automatic to intervene, the time of manual intervention, flow scheduling strategy.
7. a kind of line pushes the system surveyed, it is characterised in that including：
Data capture unit, is configured to obtain the request data of user from line upper module and replicates the data；
Data processing unit, is configured to receive the request data that data capture unit is replicated, and carries out data processing and spy
Sign analysis；
Model computing unit, is configured to receive the data after data processing unit carries out data processing and signature analysis, and
Machine learning is carried out to the data, forms Policy model；
Data readback unit, is configured to according to Policy model, with reference to request data, forms test scene under different lines, and
Lower environment plays back test scene under each line online, obtains performance data under the line corresponding to it.
8. system according to claim 7, it is characterised in that wherein,
Data capture unit utilizes the Hook Function provided in netfilter frames, and the number of request of user is directly acquired from IP layers
According to and replicate the data.
9. system according to claim 7, it is characterised in that wherein,
The data processing of data processing unit and signature analysis include：
The single request feature and combination request feature for machine learning are obtained in the request data obtained from duplication；
Wherein, single request feature includes the influential feature of the performance on line upper module once asked, combination request feature bag
Include repeatedly interaction or the influential distribution characteristics of the performance on line upper module repeatedly asked in a period.
10. system according to claim 9, it is characterised in that wherein,
Model computing unit, is configured specifically for receiving special as single request feature acquired in data processing unit and combination request
The characteristics of levying, the algorithm based on machine learning, analyzing the request data of user and the regularity of distribution of request data, to form strategy
Model.
11. according to claim 7 to 10 any one of them system, it is characterised in that wherein,
Model computing unit, is further configured to carry out machine learning according to environmental characteristic and performance data, analysis environments are special
Relation between sign and performance data, forms data correction model.
12. system according to claim 11, it is characterised in that the environmental characteristic includes at least one below：
System in hardware resource, Internet resources, fabric topology complexity, program language, resource isolation mode, and abnormal scene
Event intervention, automatic to intervene, the time of manual intervention, flow scheduling strategy.
13. system according to claim 11, it is characterised in that further include,
Data correction unit, is configured to receive under correction model and line performance data under the line of test scene, and according to amendment
Model is modified performance data under line, to obtain and performance data on the corresponding line of environment on line.
14. a kind of line pushes the terminal device surveyed, it is characterised in that the equipment includes：
One or more processors；
Storage device, for storing one or more programs；
Communication interface, is configured to communicate between memory and processor；
When one or more of programs are performed by one or more of processors so that one or more of processors
Realize the method as described in any in claim 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711164547.5A CN107967204B (en) | 2017-11-21 | 2017-11-21 | Method and system for measuring line down pressure and terminal equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711164547.5A CN107967204B (en) | 2017-11-21 | 2017-11-21 | Method and system for measuring line down pressure and terminal equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107967204A true CN107967204A (en) | 2018-04-27 |
| CN107967204B CN107967204B (en) | 2020-12-25 |
Family
ID=61999814
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711164547.5A Active CN107967204B (en) | 2017-11-21 | 2017-11-21 | Method and system for measuring line down pressure and terminal equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107967204B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109684211A (en) * | 2018-12-18 | 2019-04-26 | 北京顺丰同城科技有限公司 | A kind of order dispatch system pressure test method and device based on shop dimension |
| CN110458379A (en) * | 2018-05-08 | 2019-11-15 | 阿里巴巴集团控股有限公司 | Pressure testing system, method, apparatus and electronic equipment |
| CN111239529A (en) * | 2020-03-05 | 2020-06-05 | 西南交通大学 | Excitation test method and system supporting predictive maintenance of electromechanical equipment |
| CN111294245A (en) * | 2020-02-19 | 2020-06-16 | 北京百度网讯科技有限公司 | Offline system quality control method, offline system quality control device and electronic equipment |
| CN111723000A (en) * | 2019-03-22 | 2020-09-29 | 阿里巴巴集团控股有限公司 | Test method, test device, electronic equipment and storage medium |
| WO2021244639A1 (en) * | 2020-06-05 | 2021-12-09 | 第四范式（北京）技术有限公司 | Auxiliary implementation method and apparatus for online prediction using machine learning model |
| CN115033495A (en) * | 2022-08-02 | 2022-09-09 | 荣耀终端有限公司 | Pressure measurement model creating method and electronic equipment |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130124705A1 (en) * | 2011-11-16 | 2013-05-16 | International Business Machines Corporation | Manifold proxy for generating production server load activity on test machine(s) |
| CN103902447A (en) * | 2012-12-27 | 2014-07-02 | 百度在线网络技术（北京）有限公司 | Distributed system testing method and device |
| CN104410542A (en) * | 2014-11-18 | 2015-03-11 | 小米科技有限责任公司 | Method and device for simulation test |
| CN104778123A (en) * | 2015-03-30 | 2015-07-15 | 微梦创科网络科技（中国）有限公司 | Method and device for detecting system performance |
| CN105426308A (en) * | 2015-11-11 | 2016-03-23 | 百度在线网络技术（北京）有限公司 | Offline data construction method and device |
| CN106484610A (en) * | 2015-09-02 | 2017-03-08 | 阿里巴巴集团控股有限公司 | A kind of Beta method and apparatus |
-
2017
- 2017-11-21 CN CN201711164547.5A patent/CN107967204B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130124705A1 (en) * | 2011-11-16 | 2013-05-16 | International Business Machines Corporation | Manifold proxy for generating production server load activity on test machine(s) |
| CN103902447A (en) * | 2012-12-27 | 2014-07-02 | 百度在线网络技术（北京）有限公司 | Distributed system testing method and device |
| CN104410542A (en) * | 2014-11-18 | 2015-03-11 | 小米科技有限责任公司 | Method and device for simulation test |
| CN104778123A (en) * | 2015-03-30 | 2015-07-15 | 微梦创科网络科技（中国）有限公司 | Method and device for detecting system performance |
| CN106484610A (en) * | 2015-09-02 | 2017-03-08 | 阿里巴巴集团控股有限公司 | A kind of Beta method and apparatus |
| CN105426308A (en) * | 2015-11-11 | 2016-03-23 | 百度在线网络技术（北京）有限公司 | Offline data construction method and device |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110458379A (en) * | 2018-05-08 | 2019-11-15 | 阿里巴巴集团控股有限公司 | Pressure testing system, method, apparatus and electronic equipment |
| CN110458379B (en) * | 2018-05-08 | 2022-12-27 | 阿里巴巴集团控股有限公司 | Pressure testing system, method and device and electronic equipment |
| CN109684211A (en) * | 2018-12-18 | 2019-04-26 | 北京顺丰同城科技有限公司 | A kind of order dispatch system pressure test method and device based on shop dimension |
| CN109684211B (en) * | 2018-12-18 | 2022-04-08 | 北京顺丰同城科技有限公司 | Order scheduling system pressure testing method and device based on shop dimensions |
| CN111723000A (en) * | 2019-03-22 | 2020-09-29 | 阿里巴巴集团控股有限公司 | Test method, test device, electronic equipment and storage medium |
| CN111294245A (en) * | 2020-02-19 | 2020-06-16 | 北京百度网讯科技有限公司 | Offline system quality control method, offline system quality control device and electronic equipment |
| CN111294245B (en) * | 2020-02-19 | 2022-09-30 | 北京百度网讯科技有限公司 | Offline system quality control method, offline system quality control device and electronic equipment |
| CN111239529A (en) * | 2020-03-05 | 2020-06-05 | 西南交通大学 | Excitation test method and system supporting predictive maintenance of electromechanical equipment |
| WO2021244639A1 (en) * | 2020-06-05 | 2021-12-09 | 第四范式（北京）技术有限公司 | Auxiliary implementation method and apparatus for online prediction using machine learning model |
| CN115033495A (en) * | 2022-08-02 | 2022-09-09 | 荣耀终端有限公司 | Pressure measurement model creating method and electronic equipment |
| CN115033495B (en) * | 2022-08-02 | 2022-12-20 | 荣耀终端有限公司 | Pressure measurement model creating method and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107967204B (en) | 2020-12-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107967204A (en) | Line pushes method, system and the terminal device surveyed | |
| EP1214656B1 (en) | Method for web based software object testing | |
| Abrahart et al. | Investigating the role of saliency analysis with a neural network rainfall-runoff model | |
| CN107274023A (en) | Flow of insuring generation method, insure request processing method and device and electronic equipment | |
| WO2008134143A1 (en) | Resource model training | |
| CN108647141A (en) | Automatic test approach, device, computer-readable medium and electronic equipment | |
| CN112052172B (en) | Rapid test method and device for third-party channel and electronic equipment | |
| Ali et al. | Role of requirement prioritization technique to improve the quality of highly-configurable systems | |
| US20150018060A1 (en) | System and method for decision making in strategic environments | |
| CN108320045A (en) | Student performance prediction technique and device | |
| CN107908548A (en) | A kind of method and apparatus for generating test case | |
| CN107533473A (en) | Efficient wave for emulation generates | |
| CN116910567B (en) | Online training sample construction method and related device for recommended service | |
| Horstkemper et al. | Assessing the suitability of archimate to model industry 4.0 production systems | |
| CN106021631A (en) | Method, server, client, and system for interactive digital integrated circuit simulation verification | |
| CN106055714A (en) | Method for capturing cloud calculating data from RIA (Rich Internet Application) page | |
| CN114610615A (en) | Project test processing method, device, equipment and storage medium | |
| Rachuri et al. | Towards a methodology for analyzing sustainability standards using the Zachman framework | |
| JPH07141166A (en) | Program analyzing method using cluster analysis | |
| CN107645412B (en) | Web service combination multi-target verification method in open environment | |
| CN110458607A (en) | A kind of market clearing price analysis of Influential Factors system | |
| CN106776291B (en) | A kind of system structure complexity calculating method | |
| Bondavalli et al. | Certifications of Critical Systems-The CECRIS Experience | |
| CN110333844A (en) | Calculation formula processing method and processing device | |
| CN109918288A (en) | Use-case test 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 |