Question

4. We have introduced the error, fault, and failure, please explain the relationship among this three terminology, and write a source code to mark where is the error happened, how the fault appear, and what is the failure caused, and also explain it. (10 points) 5. When do we stop to do testing? What is the stop reasons? (10 points) 6. What is the advantages and disadvantages of Test driven testing and model based testing? (10 points)

Answer 1

Fault is a wrong or mistaken step, process or Data definition in a computed program which causes the program to perform in an unintended or unanticipated manner. It can be caused due to the mistake from programmer's end or due to some constraints in the environment. It can be as simple as a NULL pointer to simple omission of a = condition. Faults aren't always harmful, there might be a scenario that fault never gets exposed. Faults if they get exposed will lead to an error. Here due to the error, program will not give the desired output.
In absence of any corrective action, it will lead to Failure. Our software will not work according to our needs and might lead to the system failure.
Fault-> Error -> Failure

In this piece of Code, Withdrawal should not be allowed if Balance is lesser or equal to 500
if (Balance<500) 
    {
    return false; 
    }
    else WithDraw();

There is a fault ,the missing = operator in the comparison.It should have been Balance <=500.

If at some point there is a balance of 500 and the check is done, the error will come out. The system will be in incorrect state. It should not have allowed withdrawal, but instead it did. If the withdrawal is allowed, the system is failing to work as it should and user sees the failure.

5) Testing is a complex process and if done properly it will never end. But in modern scenario where software releases are under tight deadlines, we can't give infinite amount of time for testing and thus it needs to end at some point of time.

• Deadlines: The major factor affecting testing is deadlines. Once the deadline of software release or testing approaches, the testing has to finish.
• Test cases completed with certain percentage passed: There will be test cases designed to check out different combinations. Once the test cases get executed and a satisfactory percentage of accuracy is acheived the testing ends
• Test budget depleted: Testing is a resource intensive task and might be affected by budget constraints
• Bug rate falls below a certain level: Based on the kind of accuracy rate desired for that particular scenario, if the bug rate falls below certain threshold, testing ends
• Beta or alpha testing period ends