Free Essay Sample - Failure Analysis

Software engineering is a sub-field of engineering that deals with the systematic application of an engineering approach to software development. In the field of software engineering, there is a high likelihood of failures. However, to properly evaluate the causes of failures and to impede future failures from happening, Failure Analysis (FA), an integration of unique analytical tools and methods, is always required. The paper seeks to offer a general background and overview of the failure of software in the area of software engineering. It will initially concentrate on the dominant types of failures and their likely causes. Later, two types of testing methods; Non-Destructive Testing (NDT) and Destructive Testing (DT) will be examined. Methods and practices that make an impact on the failures will be discussed separately.

Is your time best spent reading someone else’s essay? Get a 100% original essay FROM A CERTIFIED WRITER!

Order now

1. Types of Failures1.1Transient

The transient software failures happen only for particular input values. Most of the time, these values evolve to be out of the range that the system is capable of handling, which results in system crashing and, therefore, failure. Therefore, whereas testing, the software is tested for both invalid and valid inputs to review the norm and reliability of the software for the conditions to be eliminated to the maximum degree.

1.2 Permanent

This form of software failure happens for the input values. In such instances, a function that has faults is invoked via the internal calling, and then the error causes the failure. Some instances of it are the infinite loops that the functions encounter, which causes the system to tolerate the surpass load, which then surpasses the limit leading to a whole system failure.

1.3 Recoverable

The recoverable system failures occur when the software fails to react to the command of the user. In such scenarios, the system often gets hanged; it does not react, or sometimes the screen might go blank after a period, the system attempts to solve the issue, and the software again begins to function.

1.4 Unrecoverable

The unrecoverable software is often the highest-ranked in terms of the damage inflicted on the hardware. This software does not recover itself whenever it fails. Most of the time, they destroy the components of the hardware, or the system requires to be restarted, which causes an anticipated norm to the other running applications.

1.5 Cosmic

The cosmic categorization of software failures is lowly ranked in terms of the damage. They rarely cause any issue to the other software or hardware. Cosmic software failures might just cause some minor irritations, such as getting the mouse stuck at a venue for a period. Arguably, there are no incorrect outcomes that are produced in this form of software failure.

2. Methods of Failure Analysis

2.1 Preventative

2.1.1 NDT

Security testing is one of the preventive software testing methods. It is a technique that involves recognizing the security flaws using functional and unit tests. Some of the security problems assessed are attempting common passwords buffer overflows, incorrect data structure, invalid inputs, denial of security attacks, and blank inputs. Arguably, security testing is vital in case the software will talk to the internet or run as root.

2.1.1 DT

Static testing is a destructive technique as acts as a preventive software testing method. Many corporations have ways of the code review process to search for code to guarantee that it satisfies the coding standards, including the error traps and sufficient comments. It assists teach the novice programmers the anticipations and serves as a deterrent against the senior programmers getting sloppy when working. Some corporations use many forms of code reviews with distinct objectives. These forms of code reviews are often named as walk-through, informal review, peer review, management review, and test design review.

2.2 Forensic

NDT

The positive testing is an alternative for failure testing methods and should be implemented for non-destructive software testing. Whereas carrying out the testing, the tester should consider the fact that the goal of the non-destructive test is to confirm that the application will work out fine on providing input data. Therefore, the aim is to verify the norm of the application for the positive set of the data. The best practice is to review if the software is doing the role that it is designed for.

2.2.2 DT

Currently, regression testing is the gold standard in forensic failure testing methods. It is a destructive technique that involves testing the software to confirm that the latest program or code change has not adversely impacted the current features. Regression testing is a part or full selection of the already performed test cases which are re-performed to guarantee existing functionalities work well (Khatibsyarbini et al., 2018). The testing is performed to ensure that the new code changes do not have side impacts on the coexisting functionalities. Regression testing guarantees that the old code still works when the recent code changes are performed. It is often performed after verification of changes. For the release that is consuming too much time to complete, regression tests should be included in the everyday test cycle.

References

Khatibsyarbini, M., Isa, M. A., Jawawi, D. N., & Tumeng, R. (2018). Test case prioritization approaches in regression testing: A systematic literature review. Information and Software Technology, 93, 74-93.