Software performance testing and reliability testing

Performance test:
1. Software performance testing consists of three objectives: ① found defects, ② performance tuning, ③ capability testing and planning

2. The main indicators of software performance are: Response time, system response time and application delay time, throughput, number of concurrent users, resource utilization

3. The response time of the system usually refers to the average response time of all functions of the system or the maximum response time of all functions.

4. For applications without concurrency, throughput is inversely proportional to response time

5. Throughput can not be infinitely increased, the key reason is that resource utilization can not be infinitely improved

6. Resource utilization can be a great help for performance tuning

7. In a stress test, the software is usually in the range of the performance degradation curve: slight performance drop zone

8. The role of the core technical force in software performance testing is the design staff

9. When the performance degradation curve is usually analyzed, it will be divided into several intervals: flat performance area, slight performance drop zone, and steep performance drop zone. The performance flat area is the normal state of the software running, so people tend to expect the interval to be as long as possible; slightly degraded performance area

is the software to withstand the high load of the buffer, the interval is also better; the performance drop zone is not the normal operating range of the software, this phase of response time can be dramatically increased to the user can not tolerate, the throughput will drop sharply or even lower than the throughput of a single user, but the interval

It can be useful for analyzing performance bottlenecks, which often means that the starting point of a rapidly declining performance zone (also known as performance inflection points) is where performance bottlenecks occur, and further analysis of resource utilization allows you to find the cause of performance bottlenecks.

10. Fail-back testing refers to the testing of the system's ability to recover from failure, and many systems are capable of having some fault tolerance that can still be used by the user in the event of some failure.

11. There are three measures to protect against physical damage: (1) All equipment should be kept away from disasters (e.g. floods, fires, thunder, fire, war, etc.). (2) The main equipment needs to be away from human destruction, such as computers and printers to be placed in the room with a secure door lock. (3) The main equipment to

Shielding electromagnetic interference.

Reliability Test:
1. At present, the main methods of software reliability analysis include failure mode impact analysis, severity analysis, fault tree analysis, event tree analysis, and potential line analysis method.

2. Reliability testing: Software reliability testing under relatively large business pressures

3. The software reliability testing process consists of five steps: Determining the reliability target, defining the software running profile, designing the test case, implementing the reliability test, and analyzing the test results.

4. The purpose of software reliability prediction is to predict the failure and effectiveness of software in the running time according to the failure condition revealed by the software in reliability and test.

5. Software reliability testing aims to predict the reliability of software in real-world operation through controlled software testing process

6. The test case set of the reliability test should fully conform to the definition of the running section

7. Test cases used by software reliability tests must meet statement overrides

8. Software reliability testing is typically used for software with reliability requirements

9. Software reliability indicators are average no failure time

10. In software reliability testing, the core of people's focus is software failure

11. Software Reliability analysis methods include: Failure mode impact analysis, rigorous analysis, fault tree analysis, event tree analysis and potential line analysis.

12. Software reliability has two meanings: ① under the stipulated conditions, the software does not cause the probability of the system failure during the stipulated time. ② the ability to perform the required functions under the stated conditions within the specified time period. And it has nothing to do with the software operator.

Software performance testing and reliability testing