Shulou(Shulou.com)06/03 Report--

1.4.1 performance testing

The performance of the system is a big concept, covering a wide range of areas. The performance of the software system includes execution efficiency, resource consumption, system stability, security, compatibility, reliability, scalability and so on. Performance testing is a kind of test implemented and executed to describe and evaluate the performance-related characteristics of the test object. Performance testing mainly uses automated testing tools to simulate a variety of normal, peak and abnormal load conditions to test the performance indicators of the system. Performance testing, load testing and stress testing are usually referred to as performance testing.

1.4.2 load test

Load testing is to test the change of system performance by gradually increasing the system load, and finally determine the maximum load that the system can bear on the premise of meeting the system performance index. In short, load testing is a step-by-step way to determine the processing capacity of the system and the thresholds it can bear. For example, thresholds for metrics such as "response time less than 10 seconds" and "server average CPU utilization less than 85%" are obtained by gradually pressing.

1.4.3 stress test

Stress testing is to obtain the maximum service level that the system can provide by gradually increasing the system load, testing the changes of the system performance, and finally determining under what load conditions the system performance is in a state of failure. Stress testing is to gradually increase the load, so that some resources of the system reach saturation or even failure.

1.4.4 configuration Test

The main purpose of configuration testing is to find the optimal allocation principle of each resource of the system by testing the software and hardware configuration of the tested software. Configuration testing can make full use of the limited software and hardware resources, give full play to the best processing capacity of the system, and can be combined with other performance testing types, so as to provide an important basis for system tuning.

1.4.5 concurrent testing

Concurrent testing is to test whether there are deadlocks or other performance problems when multiple users access the same application, the same module or data records at the same time, so almost all performance tests involve some concurrent tests. Because of the strict time requirements of concurrent testing, the simulation of concurrent users is usually with the help of tools, using multi-thread or multi-process to simulate the concurrency operations of multiple virtual users. In the subsequent introduction of the LoadRunner tool, there is a concept of rendezvous point, which is used to simulate concurrency. You can set the rendezvous point in VuGen and set its corresponding strategy in Controller to simulate the scenario of use case design.

1.4.6 capacity test

Capacity testing is to construct different orders of magnitude of records in the database under certain software and hardware conditions, and to obtain different levels of performance indicators in the case of a certain number of virtual users by running one or more business scenarios, so as to obtain the maximum session capacity, maximum capacity and so on that the database can handle. The system can handle the maximum number of users online at the same time, usually related to the database.

1.4.7 Reliability testing

Reliability testing is run for a period of time to check the stability of the system under certain business pressure (for example, the utilization rate of CPU resources is 70%-90%). Because of the long running time, it is usually possible to test whether the system has memory leaks and other problems.

In the actual performance testing process, users may often encounter system performance testing requirements that require 7 × 24 hours and run stably. For this kind of system with high stability requirements, reliability testing is particularly important. But usually a reliability test can not be performed for one year, so in most cases, the reliability test is performed for a period of time. For example, 24 hours, 3 × 24 hours or 7 × 24 hours to simulate long-time operation, through the relevant monitoring and results of long-time operation to determine whether to meet the demand, the mean time between failures (MTBF) is an important index to measure reliability.

1.4.8 failed test

For systems with redundant backups and load balancing, failure tests are used to verify whether users can continue to use the system if the system fails locally, and how much it will affect users, such as several machines for load balancing. the pressure that the system can withstand when one or more machines collapse.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.