Building Security Into the Software Development Lifecycle

It’s difficult, if not impossible, to build web application software that’s 100 percent secure. Even with a top-notch development team, mistakes are an inevitable reality of writing software. Some of these software defects can impact security: they create vulnerabilities that, if left undetected, can cause havoc months or even years later.

One way to approach this problem is to make security a priority at every stage in the software development lifecycle (SDLC). It’s important for everyone involved—developers, architects, executive stakeholders, business owners, and quality assurance testers, along with security analysts—to understand and be on board with secure design, development, testing, and deployment processes.

The consequences of failing to implement security across the SDLC are clear, and they can be damaging and costly. For example, your organization could suffer:

• Financial loss due to regulatory fines or outright theft
• Reputational damage due to hijacking or defacing of corporate websites due to easily-exploited weaknesses emerging from insecure design
• Information loss through orphaned user accounts belonging to former employees, enabling unauthorized access to corporate systems

Test Throughout the SDLC, but Don’t Stop There

It’s often suggested that the solution to securing application software is to implement a rigorous program of testing during the development cycle. There’s no doubt this is an important step, but testing by itself isn’t enough. Starting with the requirements gathering phase, it makes sense to map out a standard framework for securing the app, keeping in mind how the app needs to function across its whole lifecycle. Set specific security criteria that must be met to ensure the project is ready to move into the coding phase.

Threat Modeling: How Robust Is Your App?

By threat modeling, or putting your software through its paces to assess its robustness under various attack scenarios, you’ll gather critical data to inform the rest of your development and deployment process. Comparing various ways to address vulnerabilities and threats lets you assess the best possible solutions—and might illustrate flaws in the software that weren’t apparent, or were missed, in earlier iterations.

Move Beyond Language- and Platform-Specific Problems

Developers clearly need to be well-versed in potential security flaws that are specific to given development languages and platforms, but this isn’t all they need to know. If developers have a good understanding of how web application security vulnerabilities are created, including technical flaws and issues with business logic, they’ll be mindful of these issues during the coding phase. Secure code reviews to ensure software meets previously established coding standards help keep projects on the right track.

Keep Security in Mind During Quality Assurance

If code lacks embedded security, its quality can’t be assured, regardless of other technical considerations. The scope of QA testing should be defined prior to the project’s start, along with metrics linking performance, functionality, and security.

Ensure a Smooth Deployment

Once your app moves into production, it’s important to look at the big picture. Take platform security into consideration, ensuring that the platform your app operates on is free from exploitable flaws. (Ways to do this include deploying intrusion detection systems & firewalls and turning off unneeded services.)

Finally, it makes sense to take advantage of the wealth of free tools available to help your organization build robust, secure development processes. OWASP offers a “Secure SDLC Cheat Sheet” based on its OWASP Software Assurance Maturity Model—it’s worth examining this in detail. It includes specific, step-by-step guidance for improving the security of your organization’s development processes, along with links to other valuable OWASP resources.