What are different characteristics of performance testing?

Performance means a lot more than actually just testing the performance of a software system or application. It covers a wide range of concepts of software engineering and functionalities.

In performance testing, a software system is not merely tested on the basis of its functionalities, specifications and requirements but, it is also tested on the basis of the software system’s or application’s final performance characteristics which are measurable.

- Performance testing is both quantitative and qualitative kind of testing.
- In the field of software engineering, performance testing is typically done to determine the effectiveness and speed of a software system, hardware system, computer or device etc.

- Being a quantitative process, performance testing involves some lab tests like measurement of response time and MIPS (short form for “millions of instructions per second”) at which a software system performs.

- It also involves tests for testing the qualitative assets of a system like scalability, reliability and inter- operability.

- Often performance testing and stress testing are performed conjunction-ally.

- It’s a general kind of testing done to determine the behavior of a system whether hardware or software in the terms of stability and responsiveness when the system is provided with a significant workload.

- It is also carried out to measure, validate, verify and investigate the qualitative attributes of the system like resilience and resource usage.

- Performance testing is a sub category under performance engineering.
- It’s a kind of testing which aims to incorporate performance into the architecture and design of software or a hardware system.
- It’s basically done before the actual coding of the program.

Performance testing consists of many sub categories of testing. Few have been discussed in details below:

1.Stress testing:
This testing is done to determine the limits of the capacity of the software application. Basically this is done to check the robustness of the application software. Robustness is checked against heavy loads i.e., to say above the maximum limit.

2. Load testing:
This is simplest of all the testings. This testing is usually done to check the behavior of the application software or program under different amounts of load. Load can either be several users using the same application or the difficulty level or length of the task. Time is set for task completion. The response timing is recorded simultaneously. This test can also be used to test the databases and network servers.

3. Spike testing:
This testing is carried out by spiking the particular and observing the behavior of the concerned application software under each case that whether it is able to take the load or it fails.

Endurance testing:
As the name suggests the test determines if the application software can sustain a specific load for a certain time. This test also checks out for memory leaks which can lead to application damage. Care is taken for performance degradation. Throughput is checked in the beginning, at the end and at several points of time between the tests. This is done to see if the application continues to behave properly under sustained use or crashes down.

5.Isolation testing:
This test is basically done to check for the faulty part of the program or the application software.

6.Configuration testing:
This testing tests the configuration of the application software application. It also checks for the effects of changes in configuration on the software application and its performance.

Before carrying out performance testing some performance goals must be set since performance testing helps in many ways like:

- Tells us whether the application software meets the performance criteria or not.
- It can compare the performance of two application soft wares.
- It can find faulty parts of the program.

Validation Phase - System Testing - Regression Testing

The process of regression testing is simple i.e the test cases that have been prepared can be used and the expected results are also known. If the process is not automated it can be very time consuming and tedious operation. Some of the tools that are available for regression testing are:
- Record and Playback tools - In this, the previously executed scripts can be re-run to verify whether the same set of results are obtained. e.g. Rational Robot.

The end goals of regression testing are :
- to ensure that the unchanged system segments function properly.
- to ensure that the previously prepared manual procedures remain correct after the changes have been made to the application system.
- to verify that the data dictionary of data elements that have been changed is correct.

Most of the time the testing team is asked to check the last minute changes in the code just before making a release to the client, in this situation the testing team needs to check only the affected areas.
In short, regression testing should get the input from the development team about the nature and amount of change in the fix so that the testing team can first check the fix and then the affected areas.
Regression testing is the testing in which maximum automation can be done. The reason being the same set of test cases will be run on different builds multiple times. But then, the extent of automation depends on whether the test cases will remain applicable over the time. In case, the automated test cases do not remain applicable for some amount of time then test engineers will end up in wasting time to automate and don not get enough out of automation.

Validation phase - System Testing - Performance Testing - Capacity Planning - Load testing

Load testing is much used industry term for the effort of performance testing. Here, load means the number of users or the traffic for the system. Load testing is defined as the testing to determine whether the system is capable of handling anticipated number of users or not.
In load testing, the virtual users are simulated to exhibit the real user behavior as much as possible. Even the user think time such as how users will take time to think before inputting data will also be emulated. It is carried out to justify whether the system is performing well for the specified limit of load.

Goals of Load Testing:
- Testing for anticipated user base.
- Validates whether the system is capable of handling load under specified limit.

The objective of load testing is to check whether the system can perform well for specified load. The system may be capable of accommodating more than say 1000 concurrent users. But, validating that is not under the scope of load testing. No attempt is made to determine how many more concurrent users the system is capable of servicing.
The inference drawn from load testing is :
- Whether the system is available?
- If yes, is the available system stable?