Performance Test Model

A. Performance testing of expected indicators;

B. Performance testing of concurrent users;

C. Performance testing of fatigue strength and large data volumes;

D. server performance testing;

E. Network performance test.

In a specific test design, performance test cases are often combined with test tools to combine the use case design of server and network performance testing with the first three types. For example, LoadRunner can perform the data acquisition of the latter two types of tests while carrying out the stress test. As a result, the next two parts of the test case are only designed to be overall.

Expand to have the following eight items:

1. Performance testing of expected indicators

The system will put forward some performance indexes during the requirement analysis and design stage, and the testing related to these indexes is the first task of performance testing. This model tests the performance of some pre-determined performance indicators as expected indicators.

These metrics mainly refer to " system can support up to Four concurrent users "," system response time not longer than Ten seconds " such as these in the product specification and other documents specified in the very clear content. For such pre-committed performance requirements, the test team should first test the validation.

2. Independent Business Performance Test

Independent business actually refers to some of the core business modules corresponding to the business, these modules often have more complex functions, the use of more frequent, belonging to the core business characteristics. This kind of special, functionally independent business module is always the focus of performance testing. Therefore, not only to test such modules and performance-related algorithms, but also to test the response of such modules to concurrent users.

The core business modules can be identified during the requirements design phase and test their performance individually during the integration or system testing phase. If it is a system-class software or a special application domain software, usually from the unit testing phase began to test, and in the subsequent integration testing, system testing, acceptance testing further, to ensure that the core business module performance stability. When to start testing the core module is primarily determined by the performance testing strategy, and the reader can refer to the 1.2.2 section , " performance test case model " .

3. Combined Business Performance test

Usually all users do not use one or several core business modules, and each functional module of an application system may be used. So the performance test is to simulate the " same " Operation for multiple users ( " same " here means that many users use the same function), It also simulates the " different " operation of multiple users ( " different " here). Refers to a number of users at the same time to the different functions of one or more modules to operate, multi-business performance test combined. The combined business test is the most realistic test for the user, and the core content of the performance test. The combined concurrency of each template is typically modeled according to the user's actual usage ratio.

Combined testing is often combined with performance testing of the server (operating system,Web Server, database server) because the composite business test is the most reflective test of user usage. Through the tool simulation user operation, but also through the monitoring function of testing tools to collect server counter information, and then comprehensively analyze the bottleneck of the system, to improve the system to provide a favorable basis.

4, fatigue strength performance test

Fatigue strength testing is the test of long-running system with a certain load pressure under the condition of stable operation of the system. Its main purpose is to determine the performance of the system when it handles large volumes of traffic for a long time. The fatigue strength test can basically determine whether the system is stable after a period of time.

5. Large Data Volume performance test

Large data volume testing is usually done for some system storage, transmission, statistical query and other services to carry out large data volume testing. The performance of a major test run with a large amount of data or a large amount of historical data is typically tested for some particular core business or some commonly used combination business. Since the large data volume test is generally carried out in the production environment, it is isolated and put together with the fatigue strength test in the later stages of the entire performance test. Big data volume testing can be understood as a core business or portfolio business test under specific conditions.

6. Network Performance test

Network performance testing is primarily intended to accurately demonstrate how bandwidth, latency, load, and port changes affect user response time. In the actual software project, the relationship between the number of users of the application system and the network bandwidth is mainly tested. Network performance testing generally has a special tool, this book does not add detail. The task of network testing is usually done by the system integrator.

7. Server performance test (operating system,Web Server, database server)

The server performance test mainly is to the database, theWeb Server, the operating system test, the goal is through the performance test to find each kind of server bottleneck, provides the correlation basis for the system expansion, the optimization.

8. Some special tests

It mainly refers to some special Web performance tests, such as configuration test, memory leak test, and so on . This type of performance test or / in combination with the previous test, or in some special cases independently, this book focuses on the previous situation. In the latter case, due to the large input often through the unique tools, can not be included in the scope of performance testing

Performance Test Model