2.1. Testing in the Contest of a Software Development Lifescycle (SDLC)

An SDLC model is an abstract, high-level representation of the software development process. An SDLC model defines how different development phases and types of activities performed within this process relate to each other, both logically and chronologically. Examples of SDLC models include: sequential development models (e.g., waterfall model, V-model), iterative development models (e.g., spiral model, prototyping), and incremental development models (e.g., Unified Process).

Some activities within software development processes can also be described by more detailed software development methods and Agile practices. Examples include: acceptance test-driven development (ATDD), behavior-driven development (BDD), domain-driven design (DDD), extreme programming (XP), feature-driven development (FDD), Kanban, Lean IT, Scrum, and test-driven development (TDD).

2.1.1. Impact of the Software Development Lifecycle on Testing

Testing must be adapted to the SDLC to succeed. The choice of the SDLC impacts on the:

• Scope and timing of test activities (e.g., test levels and test types)
• Level of detail of test documentation
• Choice of test techniques and test approach
• Extent of test automation
• Role and responsibilities of a tester

In sequential development models, in the initial phases testers typically participate in requirement reviews, test analysis, and test design. The executable code is usually created in the later phases, so typically dynamic testing cannot by performed early in the SDLC.

In some iterative development models and incremental development models, it is assumed that each iteration delivers a working prototype or product increment. This implies that in each iteration both static testingand dynamic testing may be performed at all test levels. Frequent delivery of increments requires fast feedback and extensive regression testing.

Agile software development assumes that change may occur throughout the project. Therefore, lightweight work product documentation and extensive test automation to make regression testing easier are favored in agile projects. Also most of the manual testing tends to be done using experience-based test techniques (see Section 4.4) that do not require extensive prior test analysis and design.

2.1.2. Software Development Lifecycle and Good Testing Practices

Good testing practices, independent of the chosen SDLC model, include the following:

• For every software development activity, there is a corresponding test activity, so that all development activities are subject to quality control
• Different test levels (see chapter 2.2.1) have specific and different test objectives, which allows for testing to be appropriately comprehensive while avoiding redundancy
• Test analysis and design for a given test level begins during the corresponding development phase of the SDLC, so that testing can adhere to the principle of early testing (see section 1.3)
• Testers are involved in reviewing work products as soon as drafts of these work products are available, so that this earlier testing and defect detection can support shift left (see section 2.1.5).

2.1.3. Testing as a Driver for Software Development

TDD, ATDD and BDD are similar development approaches, where tests are defined as a means of directing development. Each of these approaches implements the principle of early testing (see section 1.3) and follows shift left (see section 2.1.5), since the tests are defined before the code is written. They support an iterative development model. These approaches are characterized as follows:

Test-Driven Development (TDD):

• Directs the coding through test cases (instead of extensive software design) (Beck 2003)
• Tests are written first, then the code is written to satisfy the tests, and then the tests and code are refactored

Acceptance Test-Driven Development (ATDD) (see section 4.5.3 ):

• Derives tests from acceptance criteria as part of the system design process (Gärtner 2011)
• Tests are written before the part of the application is developed to satisfy the tests

Behavior-Driven Development (BDD):

• Expresses the desired behavior of an application with test cases written in a simple form of natural language, which is easy to understand by stakeholders — usually using the Given/When/Then format (Chelimsky 2010)
• Test cases should then automatically be translated into executable tests

For all the above approaches, tests may persist as automated tests to ensure the code quality in the future adaptions / refactoring.

2.1.4. DevOps and Testing

DevOps is an organizational approach aiming to create synergy by getting development (including testing) and operations to work together to achieve a set of common goals. DevOps requires a cultural shift within an organization to bridge the gaps between development (including testing) and operations while treating their functions with equal value. DevOps promotes team autonomy, fast feedback, integrated toolchains, and technical practices like continuous integration (CI) and continuous delivery (CD). This enables the teams to build, test and release high-quality code faster through a DevOps delivery pipeline (Kim 2016).

From the testing perspective, some of the benefits of DevOps are as follows:

• Fast feedback on the code quality, and whether changes adversely affect existing code
• CI promotes shift left in testing (see section 2.1.5) by encouraging developers to submit high quality code accompanied by component tests and static analysis
• Automated processes are promoted like CI/CD that facilitates establishing stable test environments
• The visibility on non-functional quality characteristics increases (e.g., performance efficiency, reliability)
• Automation through a delivery pipeline reduces the need for repetitive manual testing
• The risk of regression is minimized due to the scale and range of automated regression tests

DevOps is not without its risks and challenges, which include:

• The DevOps delivery pipeline must be defined and established
• CI / CD tools must be introduced and maintained
• Test automation requires additional resources and may be difficult to establish and maintain

Although DevOps comes with a high level of automated testing, manual testing — especially from the user’s perspective — will still be needed.

2.1.5. Shift Left

The principle of early testing (see section 1.3) is sometimes referred to as shift left because it is an approach where testing is performed earlier in the SDLC. Shift left basically suggests that testing should be done earlier (e.g., not waiting for code to be implemented or for components to be integrated), but it does not mean that testing later in the SDLC should be neglected.

There are some good practices that illustrate how to achieve a “shift left” in testing, which include:

• Reviewing the specification from the perspective of testers. These review activities on specifications often find potential defects, such as ambiguities, incompleteness, and inconsistencies
• Writing test cases before the code is written and have the code run in a test harness during code implementation
• Using CI and even better CD as it comes with fast feedback and automated component tests to accompany source code when it is submitted to the code repository
• Completing static analysis of source code prior to dynamic testing, or as part of an automated process
• Performing non-functional testing starting at the component test level, where possible. This is a form of shift left as these non-functional test types tend to be performed later in the SDLC when a complete system and a representative test environment are available

Shift left might result in extra training, effort and/or costs earlier in the process but is expected to save efforts and/or costs later in the process.

For shift left it is important that stakeholders are convinced and bought into the concept.

2.1.6. Retrospectives and Process Improvement

Retrospectives are often held at the end of a project or an iteration, at a release milestone, or can be held when needed. The timing and organization of the retrospectives depend on the particular SDLC model being followed. In these meetings the participants (not only testers, but also e.g., developers, architects, product owner, business analysts) discuss:

• What was successful, and should be retained?
• What was not successful and could be improved?
• How to incorporate the improvements and retain the successes in the future?

The results should be recorded and are normally part of the test completion report (see section 5.3.2). Retrospectives are critical for the successful implementation of continuous improvement, and it is important that any recommended improvements are followed up.

Typical benefits for testing include:

• Increased test effectiveness / efficiency (e.g., by implementing suggestions for process improvement)
• Increased quality of testware (e.g., by jointly reviewing the test processes)
• Team bonding and learning (e.g., as a result of the opportunity to raise issues and propose improvement points)
• Improved quality of the test basis (e.g., as deficiencies in the extent and quality of the requirements could be addressed and solved)
• Better cooperation between development and testing (e.g., as collaboration is reviewed and optimized regularly)

---

<< Previous Section | Next Section >>