Techniques For Testing

Techniques For Testing

1. Test Harnesses

   • To streamline software development and possibly to assist with future maintenance.
   • Organisations use test harnesses to automate the testing processes.
   • Consists of predefined software and testing scripts, plus expected test results
   • Can then be operated whenever a change is made to the software
     • Establish quickly whether or not changes been successful
   • Use of test harnesses can avoid the need to perform lengthy and time- consuming manual testing.

2. Performance Testing

   • Performed to determine how fast some aspect of the system performs under a particular workload.
   • It can help demonstrate that a system meets performance criteria
   • It can compare two systems to find which performs better
     • Or the parts that perform badly

3. Volume Testing and Scalability Testing

   • Testing using the anticipated peak volumes of date input
     • OR network traffic necessary to test the performance
       • Of a system under live condition
   • Database + log files are likely to increase in size over time
     • Continued operation with these increase volumes needs to be tested
     • Some systems are used intensively at certain times of the year and volumes may increase dramatically at those times.
       • E.g. Black Friday
   • Scalability - Can it cope with future demands

4. Usability

   • It is a means for measuring how well people can use an ICT system or a part of a system for an intended purpose
   • If it uncovers difficulties, such as people having difficulty understanding instructions, knowing what to do next etc... then developers should improve the design and re-test

5. Prototyping

   • Where prototypes have been used to define (and possibly to sign off) the system, the outputs and appearance of the completed system can be compared with the prototype system.

6. Multi-platform/portability testing

   • Some systems will be used on a range of different hardware and operating system platforms.
   • Whilst theoretically the systems should operate identically, testing is needed on all the possible hardware/software platforms anticipated before live implementation.