Software integration, confirmation and system test method
introduction
Software Test is divided into white-box testing in the test case design method.
Black-Box Testing. Software test is divided into unit test according to test or test strategy, software test (Unit)
Testing, integrated test (Integration Testing), confirmation test (Validation Testing) and
System Testing]. In the previous document, we have been testing the test cases in white boxes and black boxes.
The design method has been explained in detail. At the same time, the unit test is also explained and trained. This article will explain from the test strategy
Integrated tests in software testing, confirmation tests and system test strategies.
Software test policy method
Test is a series of planned systematic activities. In order to implement test activities, many test strategy methods have been proposed.
A software test policy not only includes low level testing (Low Level Testing) and to include high-level testing [high
Level testing. Low-level testing is to verify the correctness of the original code. High-level testing is to confirm the main system function
Foot user demand.
Verification and Validation (VERFICATION AND VALIDATION)
In general, software testing is verifying and confirming verification and validation (V & V]. Verification finger
The certificate software correctly implements a series of activities for a specific function. Confirmation means that the software produced can be traced back to user needs.
A series of activities. Boehm's explanation of V & V is:
Veification: "Are We Building The Product Right?"
Validation: "Are We Building The Right Product?"
The definition of V & V contains many activities, that is, software quality assurance SQA. Figure 2-1 shows these activities that implement software quality.
Software engineering methods provide the basis for quality establishment. Analysis, design and encoding methods provide unified technical and predictable
The result is to improve quality. Regular viewing and review helps to ensure the quality of the products of the software engineering. Metrics and controls should be
In each part of the software configuration. Standards and processes help ensure the consistency of development. A regular SQA process is enhanced
Overall quality. The test is the last measure to ensure quality. But you can't see the test as a security net. Quality is running through soft
Each stage of the process process. Therefore, although testing plays a very important role in V & V, many other activities are also necessary.
of. In order to improve the quality of the software, you should pay attention to every activity in V & V.
Software integration, confirmation, and system testing methods www.51testing.com
Bo as Feng Software Make BWF Software
Software quality
Software engineering
method
Standard and
step
Regular view
versus
Review
measure
test
SQA
SCM
Figure 2-1 Activities to implement software quality
Software test strategy
The software engineering process can be seen as a spiral shape. The software test strategy can also be seen as a spiral. Figure 2-2 shows the test strategy
Fig.
From Figure 2-2
See the software test policy
Sketch: starting from the code order
The unit test of the segment is then extended to the integrated test of the design phase, and the confirmation test is extended to the requirements of the demand analysis.
After the system test in the system engineering phase. From the perspective of the system process, the test strategy has four steps: unit test, integrated test
Try, confirm test and system testing. Figure 2-3 shows the software testing step.
Software integration, confirmation, and system testing methods www.51testing.com
Bo as Feng Software Make BWF Software
unit test
Test direction
Code
Integration Testing
High-order test
High-Order Test
Integration Test Design
demand
Figure 2-3 Software Test Steps
As can be seen from Figure 2-3, the initial test is concentrated on each module, and since each module completes a functional unit, the test is called a unit test. Unit test mainly applies a white box test method. Next is the module assembly and integration to make up
The whole package. The integrated test focuses on confirming and procedures, integrated testing mainly uses black box test methods, with
White box test method. After the software is integrated, you need to complete a series of high-order tests (confirmation and system testing). Confirmation criteria must be measured
test. Confirm that the test provides the final guarantee for all functions and performance requirements. Confirmation test only uses a black box test case
Method.
Test standard
In software testing, people often ask such problems: "When is the test completed? - How to know the test is sufficient?
Minute? "Unfortunately, there is no certain answer to this issue, but there are some statistics and experience answers or refers to
guide. In order to determine when the test is sufficient, the software engineer needs more stringent standards. MUSA and
Ackerman recommends answering these questions based on statistical guidelines. Apply statistical model and software reliability theory, software failure /
The model (un found in the test) can be established in the form of execution time function. For example, a software fault model is called pair
Norphous work model. Knowledge and models in software reliability and software testing intensity will be established in later related documents
Introduction.
Unit testing
The unit test is a test based on the program module for proper verification. In this regard, we have already prepared it.
Data and multi-stage training. Therefore, it will not be in terms of this.
Integration testing (Integration Testing)
Integrated testing, also called assembly tests or joint tests. On the basis of unit testing, all modules are required as design requirements)
According to the structural diagram] Assembly, it is assembled into a subsystem or system, and integrated tests. Practice shows that although some modules can work alone
Do not guarantee that it can work properly. The program is very good in some local reflection.
It can be exposed to the implementation of the function. That is, you should consider the following questions:
(1] When connecting each module, the data that traverses the module interface will be lost;
Software integration, confirmation, and system testing methods www.51testing.com
Bo as Feng Software Make BWF Software
(2] The individual function is combined to achieve the parent function required for expected requirements;
(3) Whether the function of a module will adversely affect the functionality of another module;
(4) There is a problem with the global data structure;
(5) The error accumulated in a single module, whether it will enlarge, thereby achieving unacceptable levels.
Therefore, after the unit test, it is necessary to integrate testing, discover and exclude the above problems that may occur in the module connection.
The software subsystem or system that eventually constitutes the requirements. The subsystem, integrated test is also called component testing.
Any reasonable organization of integration tests, that is, what way is selected to assemble the module to form a runoff system, directly affect
To the form of the module test case, the type of test tool used, the module number, and the order of the test, generate test cases and debugging
cost. Typically, there are two different assembly methods: disposable assembly and value-added assembly.
Disposable assembly (BIG BANG)
Disposable assembly is a non-value assembly method (non-incremental integration), also called
Body assembly. According to this assembly, first, each module is tested, and then all modules are assembled together.
The test, the software system that eventually obtained. For example, there is a system structure, as shown in Figure 4-1 (a). Its unit test and assembly sequence are shown in Figure 4-1 (b).
A
B C D
E f
D1 D2 D3 D4 D5
D b C
S1 S2
E f
A a a
B C D
F e
S3 S4 S5
(a)
(b)
Figure 4-1 Dispensing
In the figure, the modules D1, D2, D3, D4, D5 are the driving modules, S1, S2, S3, S4, S5 established when each module is tested, and is
The pile module established by the unit test. This disposable assembly is attempted to test in the module unit with the assistance of the auxiliary module.
On the basis, connect the module to the test. However, due to the inevitable module intergeneration in the program, the global data
Structure, etc., so the possibility of successful trial operation is not very big. The result found that there was a mistake, but it could not be found.
because. It will encounter difficulties and let go.
Proliferation assembly (Incremental Integration)
Software integration, confirmation, and system testing methods www.51testing.com
Bo as Feng Software Make BWF Software
Value-added assembly is also known as inclusive assembly. The first is a module unit test for one module, and then these modules
Assembled into a larger system, the connection side is tested while the assembly process is to find problems generated during the connection. Finally proliferation
Assembled as a required software system.
Top-down integration
This assembly is to assemble the modules according to the system program structure, along the top of the control hierarchy. The steps are as follows:
(1) With the main module as the measurement module and drive module, all subordinate modules directly under the main module use pile module to the main module
carry out testing.
(2) Replace the phase with the actual module with a strategy of depth-first (as shown in DEPTH-FIRST) or BREPTH-FIRST.
The pile module, replace the pile instead of their direct subordinate module, and assembles the new subsystem with the test module or subsystem.
(3) Return test (ie, re-executing all the test or partial tests previously done), excluding the assembly process
Musical possibilities.
(4) Judgment whether all modules have been assembled into the system? Yes, the test ends, otherwise go to (2) to execute.
A a a a
S1 S2 S3
S4
S2 S2 S3 S3 B B B
E e
C S3
Test A Add to B Add to E-added C
A
B C D
E S5
A
B C D
E f
Join D Add to F
Assembly in depth direction
Figure 4-2 Since the top downward proliferation (assembled in depth)
The top-down proliferation method verifies the main control and judgment points earlier during the test. Reasonable in one function
In the program module structure, it is often encountered in a higher level. If this is mainly controlled
Troune, it is important to find it as soon as possible to reduce future rework, so this is necessary. If you are assembled in depth direction,
You can first implement and verify a complete software feature, you can assemble and test, check and overcome the branches of logic entries.
Errors and defects of hidden, verify the correctness of its function, provides a guarantee for the assembly and testing of major machining branches.
In addition, feasibility is proven earlier, and it is also possible to bring success for developers and users.
There is a logic order problem in the assembly and testing of the top of the top. The lower layer is required in order to fully test the higher level
This type of problem occurs when processing information. In the top-down assembly phase, it is necessary to replace the lower layer module with a pile module.
Software integration, confirmation, and system testing methods www.51testing.com
Bo as Feng Software Make BWF Software
Therefore, the writing of the pile module may differ from the following options as shown in Figure 4-3 depending on the situation.
Pile Module STUB Pile Module STUB Pile Module STUB Pile Module Stuba B C D
Display tracking information display passed information from a table (or outside
Department files) Returns a value
Conduct a table query
In accordance with input parameters
Returns the output parameter indicates the transferred data message
Figure 4-3 Several choices of pile modules
In order to accurately implement the test, the function and reasonable interface of the submodule should be made correct and effectively to the pile module, not
Can be a dumb block that only contains the return statement or only the module has called information, and does not perform any functional dumb block. If you can't make the pile module
The useful information correctly can be transmitted up, and the following solutions can be used.
(1) After the multi-test is postponed until the pile module is replaced by the actual module;
(2) Further development of pile modules that can simulate actual module function;
(3) Assembly and test software from bottom to;
Since the bottom upward proliferation
This assembly is to assemble and test from the bottom of the program module structure. Because the module is from bottom to
Assembly, for a given level module, its sub-module (including all sub-modules of the sub-module) has been assembled and tested
Become, so no longer need a pile module. The information obtained from the sub-module can be run directly from the sub-module during the test of the module.
. The steps in the next to the bottom are as follows:
(1) Control the parallel test of the most underlying module by the drive module; or the bottom layer module can also be combined into a particular software.
The functional cluster is controlled by the drive module.
(2) Instead of the drive module with the actual module, assemble the straight sub-module it has tested into a subsystem.
(3) To provide a drive module for a subsystem, perform a new test.
(4) Determine if the main module is assembled. Yes, the test ends, otherwise (2) is executed.
Software integration, confirmation, and system testing methods www.51testing.com
Bo as Feng Software Make BWF Software
The system structure shown in Fig. 4-1 (a) is an example, and the order of the bottom up to the upward assembly and test is described below with the following Figures 4-4.
D1 D2 D3 D4 D5
E e
C C f D D B B E
F F
A
Figure 4-4 Since the bottom upward value-added assembly method
When assembling and tested up, it is necessary to prepare a corresponding drive module for the measurement module or subsystem. Common kinds
The type driver module is shown in Figure 4-5:
Driver Driver Driver Driver Driver Driver Driver Driver
A B C D
Call the slave module
From table (or external file)
Transfer parameters
Display parameters and driver B,
C function
Figure 4-5 Several choices of driving modules
Distance to send parameter information
With the upward movement of the assembly level, the drive module will be greatly reduced. If the top two layer modules of the program module structure
Making and testing with the top downwards, it can significantly reduce the number of drive modules, and can greatly reduce several system groups
The work you need to do.
Mixed proliferation test
It has advantages and disadvantages in the way in the top-down proliferation and the above-bottom proliferation. Generally speaking, the advantage of one way is that another
Disadvantages of the way.
The disadvantage of the top-down proliferation method is to establish a pile module. To make the pile module to simulate the function of the actual sub-module is very difficult
Difficult, because the pile module needs to return to the actual sub-module function that it is replaced after receiving the information sent by the module.
The information sent, which will increase the complexity of the establishment of the pile module, and lead to additional additional tests. At the same time, involving complex algorithms
The module that is real input / output is generally under the bottom, and they are the most prone to problems, and they will encounter this in the later stage of assembly and testing.
Some modules, once found problems, leading to excessive regression test, and the advantages of the top-down proliferation method can be found in the Lord
Problems to control.
The disadvantage of the above-bottom proliferation method is that the program has not been able to exist as an entity until the last module plus it.
Only one entity ". That is to say, in the process of assembling and testing from the bottom up, the main control until finally
. However, the advantage of this method is that it is not necessary to establish a driving module, and it is generally easier than the establishment of a pile module.
Software integration, confirmation and system testing method www.51testing.com Bo is a peak software to make BWF Software
The module to complex algorithms and real input / output is first to be assembled and tested, and some of the most prone to problems can be solved early in early
Decided. In addition, the parallel test of multiple modules can be implemented in a manner, and the test efficiency can be improved. Therefore, usually
The assembly and testing is combined in two ways. Three common integrated proliferation methods are briefly introduced below.
(1) Down-down proliferation test of the distraction: Its basic idea is to strengthen the new algorithm module for input / output modules and introduce
Test, and assemble it up by an upstore to a relatively complete and relatively independent subsystem, and then start from the main module
Long proliferation test.
(2) Since the bottom-up-top-top proliferation test: it first upsheet up to the root node module
Assembly and testing, and then doing a self-descending assembly and test on a subsystem containing write operation.
(3) Return test: This way takes the self-turning manner to test the modified module and its submodules, and then this
Subject to the subsystem, then test it up by end to check whether the subsystem is adapted to the interface of the upper class module.
When assembly tests, testers should determine key modules to test these key modules early. Key module should at least
One of the following features: (1) satisfying some software requirements; (2) at a higher level (high-level control) in the module structure of the program
Module); (3) is more complicated, more prone to errors; (4) has a clear defined performance requirement.
Test key module functionality when making regression tests.
Organization and implementation of integrated testing
Integrated testing is a formal test process, must be carefully planned and coordinated with the completion time of the unit test. In formula
When testing the plan, you should consider the following factors:
1) What kind of system assembly method is used for assembly testing;
2) The order of connecting each module during assembly testing;
3) Whether the module code preparation and test progress are consistent with the order of assembly testing
4) Do you need a special hardware device during the test;
After solving the above problems, you can list the preparation, test schedule of each module, indicating that each module unit is finished.
Date, the date of the first integration test, the date of the integrated test, and the required test case and the desired measurement
Test results.
When the hardware device required for the software test is missing, it should be checked whether the delivery date of the hardware is consistent with the integrated test plan.
For example, if the test requires a digitizer and a plotter, the corresponding test should be arranged when these equipment can be put into use, and need
Use the installation and delivery of the hardware for a period of time to leave a time margin. In addition, you need to consider testing in the test plan.
Preparation of software (drive module, pile module, test example generation procedure, etc.).
Integrated test completion
How to determine the integrated test process complete, you can check as follows:
1) Successfully implement all integration tests specified in the test plan;
Software integration, confirmation, and system testing methods www.51testing.com
Bo as Feng Software Make BWF Software
2) Fixed the mistakes found;
3) The test results passed the special group review.
The integrated test should be carried out by a special test team, and the test team consists of experienced system designers and programmers. whole
Test activities should be carried out in the event of reviewers.
After completing the predetermined assembly test work, the test team should be responsible for finishing, analyzing, analyzing, and forming a test report.
. In the test report, the actual test results are recorded, and the problems found in the test, solve these problems and solve them.
The result of the test again. In addition, there should be some problems that are currently unable to solve, and some issues that managers and developers should also need.
Provide test review and final decisions to make handling opinions.
The documents that integrated tests need to be submitted are: integrated test plan, integrated test specification, integrated test analysis report.
Validation Testing
Confirmation test is also called efficient testing. Its task is to verify that the functionality and performance of the software and whether its characteristics are consistent with the requirements of the user. The function and performance requirements of the software have been clearly stated in the software requirements specifications. Describe in software requirements specifications
All user visible software properties, which is called valid guidelines, which contain information is the basis for software confirmation test.
foundation. After the integrated test is complete, the dispersion development module is coupled to constitute a complete program. The interface between each module
The various problems have been eliminated. So the test work enters the final phase - Validation Testing. What is confirmation
Try, the statement, the easiest and most stringent explanation is whether the software developed can operate according to the requirements of the customer.
If you meet this requirement, the software that is considered to be evolved is qualified. Therefore, some software development departments call the confirmation test as qualified.
Qualification Testing. The customer's request herein usually refers to the software features determined in the software specification manual and
Technical indicators, or specifically specified for the test specified in the test.
The work you need to do in the confirmation test phase is shown in Figure 5-1. First of all, validity testing and software configuration review,
The acceptance test and installation test are then performed, and after the expert identification, it can be delivered software.
Select tester
Construct test case
Actual operation test
Software plan
User documentation
Development documentation
Source program text
Support environment
testing report
Software configuration
effective
Sexuality
test
software
Configure
Review
management
mechanism
ruling
expert
Identification
meeting
Pay a user
Operation and Maintenance
Figure 5-1 Procedure for confirming the test
Software integration, confirmation, and system testing methods www.51testing.com
Bo as Feng Software Make BWF Software
Confirmation of the test
How to determine the software developed is successful? In order to confirm its function, performance, and whether the restrictions have reached
What should I do if I should test? In the demand specification, principled provisions may be made, but it takes more details in the test phase.
Further description is further specifically described in Test Specification. For example, when developing a test plan, it is clear that it is clear.
What are the test should be tested, and the necessary test cases are given. In addition to considering functionality, other aspects
Requirements. For example, whether portability, compatibility, maintainability, human-machine interface, and development of document materials are required.
After confirming the test, conclusion evaluation should be made for the developed software. This is nothing more than two cases:
(1) The inspected software function, performance, and other requirements have met the requirements of the demand specifications, and thus can be accepted. recognize
It is a qualified software. (2) After the inspection discovery is quite deviated from the demand instructions to obtain a list of defects. for
In the second case, it is often difficult to correct the problems found before the delivery period. This requires development departments and customers to cooperate
Business, find a solution.
Effective test (black box test)
The effectiveness test is a method of using a black box test under analog environment (which may be the environment developed), verification
Whether the test piece meets the needs listed in the demand specifications. To this end, we need to set the test plan first, and the specified species must be tested.
Class, also need to develop a set of test procedures to describe specific test cases. By implementing a predetermined test plan and test step,
Whether the characteristics of the software are consistent with the needs, ensuring that all software features can be met, all software performance requirements can be
All documents are all correct and easy to use. At the same time, for other software requirements, such as portability, compatibility, automatic
Recovery, maintainability, etc., also test, confirm that it is satisfied. After the test cases of all software tests,
Some test results can be divided into two categories:
1) The test results are consistent with the expected results. This shows that this part of the software or performance characteristics is consistent with the requirements specifications.
So you have to submit a question report for it.
2) The test results are inconsistent with the expected results. This shows that this part of the software or performance characteristics and demand specifications are different.
To, there is therefore submit a question report for it.
Software configuration review
The software configuration review is an important part of confirming the test process. The purpose is to ensure that all the components of the software configuration are complete, each
The quality of the aspects meets the requirements necessary, and the details of the maintenance phase are required, and the classified catalog has been arranged.
In addition to the content and requirements specified in the contract, outside the worker review software configuration, in confirming the process of the test, should strictly follow
The use steps specified in the user manual and the operating manual to check the integrity and correctness of these document data. Must be careful
Record the omissions and errors found and appropriately supplement and correctly.
α test and beta test
After the software delivery is used, the user will actually use the program, which is unpredictable for developers. Because of the user
In the process of use, the error, abnormal data portfolio of how to use, and it seems to be due to some users.
Clear but it is difficult to understand the output, etc.
Software integration, confirmation, and system testing methods www.51testing.com
Bo as Feng Software Make BWF Software
When the software is developed for a specific user, you need a series of acceptance to make the user verify that all needs have been
Satisfy. These tests are based on the user, not the system developer. The acceptance test can be a simple
Formal "test run". It can also be a set of complex organized test activities. In fact, the acceptance test may hold
Several weeks to a few months.
But if the software is a product developed for multiple users, it is not cut to perform formal acceptance tests one by one by one.
Actual. Many software product producers use a test method called A test and B test to find that only the final use
The user can find the mistake.
α test is a test conducted by a user in the development environment, or the user of the development mechanism is simulated actual operation.
Test in the environment. The software is used in a natural setting state. The developer sits next to the user, keeping a mistake at any time.
The problems in the situation and use. This is a test in the controlled environment, the purpose of the α test is the class price software product
FluRPS (ie functional, localization, usability, reliability, performance, and support). Especially pay attention to the interface and characteristics of the product. a test
The person who first saw the product except the product developers, and their proposed features and modifications were particularly valuable.
A Test can start from the end of the software product coding, or start after the module (subsystem) test is completed, or it can be confirmed
The product reaches a certain stability and reliable procedure during the test. The relevant manual (draft) should be prepared in advance.
The beta test is tested by multiple users of the software in the actual use environment of one or more users. These users are
The company signed an external customer who supports product pre-contracted contracts, and they ask for the product and willing to return to misunderstand error messages
Interest to the developer. Unlike the α test, developers are usually not in the test site. Therefore, the beta test is uncontrollable in the developer.
Software on-site applications in the environment. In beta beta test, all issues encountered by users, including real and subjective
Determination, regularly report to the developer, developers have made modifications after the comprehensive user's report, and the software product is delivered to all
Body user is used. The beta test mainly measures the FLURPS of the product. Focus on product support, including documentation, customer training and support
Hold product production capacity. The beta test can be started only when the alpha test reaches a certain reliability. Since it is in the entire test
The final stage, I can't expect the main problem at this time. At the same time, all the text of the product should also be completely fixed at this stage.
draft.
Since the main goal of beta test is to test supportability, the beta test should be taken from the personnel issued by the host product.
. Acceptance Testing
After passing the effectiveness test and software configuration review, the system's acceptance test should be started. Acceptance test is
User-based test. Software developers and QA (quality assurance) should also participate. From the user to the design test case, use
The user interface inputs the test data and analyzes the output of the test, which generally uses the actual data in the production to test. At test
In the process, in addition to considering the functionality and performance of the software, it also responds to the replaceability, compatibility, maintenance, and error of the software.
Confirmation of complex function or the like.
The acceptance test experiment is a "Walkthrough".
Confirm the structure of the test
Software integration, confirmation, and system testing methods www.51testing.com
Bo as Feng Software Make BWF Software
There are two cases of confirming the test:
1) The function and performance are consistent with the requirements of the user, and the software can be accepted;
2) There is a gap between function and performance and user requirements.
In the latter case, it is usually related to errors in the software demand analysis phase. At this time, a software defect table needs to be opened.
Or Software Problem Report, through the negotiation with the user, resolve the discovered defects and errors.
Confirm that the documentation to be delivered is: confirm the test analysis report, the final user manual and the operating manual, project development summary
report.
System Testing
The system test is the software that will be tested by confirming the test, as an element based on computer system, with computer hardware,
Peripherals, some other system elements such as support software, data, and personnel are combined, in actual operation (use) environments,
The computer system performs a series of assembly tests and confirmation tests.
The purpose of system testing is to find software and system definitions or contradictions in accordance with the definition of the needs of the system.
The place. The test cases for system testing should be designed according to the requirements analysis instructions and operate in the actual use environment.
Since the software is only an integral part of the computer system, after the software development is complete, it will eventually be in other parts of the system.
The distribution set is running. The system needs to complete the assembly and confirmation test before the operation of the operation, to ensure that the components can not only be alone
It is also subject to test and works fine in the environment where the system is coordinated. The system components mentioned here are removed
Software, it is also possible to include computer hardware and related peripherals, data and its collection and transport mechanisms to master the computer system.
The personnel and operations of the system, which may even include an actuator being calculated. Obviously, the confirmation test of the system has been
It is completely exceeded the scope of software. However, software has a considerable position in the system, how is the quality of the software,
Software test work is to be strong and successfully completed the system test relationship. On the other hand,
The test is actually a comprehensive test for each component in the system. Although every test has a specific goal,
However, all inspections must verify that each of the systems has been properly integrated and can complete the specified function. the following
Summary of several system tests separately, respectively:
Recovery test
The recovery test is to take a variety of artificial intervention to make the software errors without working properly, and the system's recovery can be inspected.
force. If the system itself can recover automatically, check: reinitialize, checkpoint setup mechanism, data recovery
Whether the restart is correct. If this recovery requires human intervention, it should be considered whether the average repair time is limited.
Inside.
Safety test
The purpose of the safety test is to verify that the protective mechanism installed in the system is determined to protect the system so that it is not subject to various
Very disturbed. System security test To set up some test cases, the system is securely confidential, check if the system is
There is a safe confidential vulnerability.
Software integration, confirmation, and system testing methods www.51testing.com
Bo as Feng Software Make BWF Software
Strength test
The ability to test the system is the highest actual limit. When performing an intensity test, let the system run under the conditions of resources, abnormal frequencies, and abnormal bullets. For example, if the normal interruption average frequency is one to two times per second, the intensity test is designed for each
Seconds 10 interruptions. Another example is the normal operation of a system to support 10 terminal parallel, and the intensity test checks 15 terminal parallel work.
Case.
Performance Testing
Performance Test Inspection is installed in the software running performance. This test is often carried out with intensity testing. Remember
Recording performance requires the software (or block) that is required to install the necessary measuring instrument or set for metrics.
Reference: 1. "Computer Software Test Technology" Zheng Renjie, Tsinghua University Press, 1992.
2. "Software Engineering - A PRActioner 's Approach", R. S. Pressman, 1998.
3. "Practical Software Engineering", Zheng Renjie, Yin Kun, Tao Yonglei, Tsinghua University Press, 1997.
Software integration, confirmation, and system testing methods www.51testing.com
Bo as Feng Software Make BWF Software
__
