Stress Test: In this case we increase the load of the system to check if it can sustain such load and what is the bottleneck point. Also when system fails for such heavy load can it recover gracefully.
Load Test: In the tested system we will continuously increase the pressure until the performance index reached its limit, response time exceeds a predetermined target, or a resource has already reached saturation point. Load test can find system processing limit, to provide a basis for system tuning.
Large data volume test: In this case a large volume of data is used to test the system. This volume will be recorded in the requirement by the client, mainly the daily data volume transaction for the system. For example a system will process 700 transactions per day. Configuration test: Test to determine what is the optimal level of recourses needed by the system to run.
Reliability Test: You can increase CPU resources to maintain a 70% -90% utilization of the pressure. Then continue pressure on the system to run 8 hours, and do the analysis of the stability of the system based on the results. Concurrency Testing: Run same transactions simultaneously to find some algorithm design problems.
Performance testing using concurrent users:
1. Performance testing is mainly to find software problems and hardware bottlenecks.
2. At the requirement phase people should find all the performance requirements which are also called non functional requirements.
3. For each performance requirements we should have one or two test cases.
4. Is expected to test the performance of indicators are usually single-user oriented, if it involves concurrent users, then we need to design test cases accordingly.
Network Performance Testing
The main focus of this testing to determine the number of users and network bandwidth relationship. It also shows how the bandwidth, load and port change affects the user’s response time. The best way is to adjust the performance of soft and hard combination.
Large data volume testing
Mainly aimed at the database systems have special requirements for testing, divided into three types:
1. Real-time large amounts of data: real-time simulation of a user working with large amounts of data, the main purpose was to test the user operations have a greater amount of data and the system can run stable.
2. Limit state test: The main focus is to test the system for a period of time where the system accumulate a certain amount of data and can normally run the business
3. The combination of the first two: Test the system has accumulated large amounts of data, some real-time transactions have a greater amount of data and can work steadily.