Threat Modeling: Building reliable software by breaking it.

June 18, 2021

Threat modeling is the process by which engineers are able to represent in a structured way all the possible threats that could affect the security of an application.

Security breaches are a constant worry when developing and maintaining an application since malicious users become more and more advanced with time by either attacking at vulnerable hours and/or leaving behind a traceable path to attack again with no possible detection.

When developing for threat modeling, engineers need to think through all potential threats that surround their application and come up with responses for these security breaches. Understand that threats can be caused by either internal users committing simple mistakes or external users with performing malicious attacks. Some threats can also be completely unrelated to security and are more on the “natural” side of things such as a mishap with the facility where servers are based on.

By performing threat modeling processes we are trying to optimize the security of our application by identifying key vulnerabilities, however, the question that remains is: When should we approach this? The answer is simple: Start early and implement often. Including threat modeling early on the project allows engineers to refine their processes (so that modeling is simpler later on) and also gives the opportunity to examine vulnerabilities, study what’s at stake and decide what risks are worth taking.

The average time between the start of a security breach and its detection averages 230 days, this is a long time that allows attackers to basically do whatever they want with applications. By implementing often we increase the chances of finding these threats.

One of the biggest advantages of threat modeling is that it’s scalable, it allows us to continuously refine our processes basing off of what we’ve done before.

At Rootstack, we understand the importance of having secure and stable software. We implement the best practices for threat modeling with our DevOps services.