This is really a good question, that many potential prospects are asking us. For most of the time the answer is easy: we can point them to hacks to the likes of Equifax, British Airways, Carphone Warehouse and so on. But what if someone believes they’re already protected?
We are already checking our code!
Yes, some of us already have a SAST solution in place, like SonarCube or Fortify, and we think we are fully protected. Some of us instead use opensource alternatives, like Spotbugs or PMD, and we also feel okay. These tools will continuously scan your developers’ source code and make sure that it’s not including any pattern that could be exploited by a malicious hacker.
On top of that, we have already secure coding in place: we use peer reviews and pair programming, so that all our code is severely scrutinized. We use a set of agreed secure coding practices, we follow detailed checklists, and we have developers trained continuously on these matters. And yes, that’s good!
But not everything is written from scratch…
When developers code, they do not write everything from scratch. Have you ever been surprised at how little time it took a piece to be delivered? Maybe, just once. And you ask how did that happen? Your developers, rightly so, dipped into the big pool of opensource and pulled a component that was able to accelerate their development, providing ready to use building blocks so that they could concentrate only on the business logic.
Those opensource components are so widely used, that in your final packaged product, up to 80% of the code is made of those components. Only a mere 20% is the carefully crafted code by your developers, where your internal practices and existing tools make a difference. But who’s checking the security of those opensource components? Well, it so happens that many people across the world checks them. This allows the components to evolve quickly. When a problem is found, either a bug or a security problem, it’s rapidly fixed and a new version of the component is released. When this becomes a security problem, often the issue is publicly reported in specific mailing lists or newsrooms dedicated to security.
How do you know an opensource component needs to be updated?
The core problem is, how do you know about all this? How does your team become aware that a certain component they use is now vulnerable, and that they need to update to a new version? How do they make sure that the code they do not write, which could account for 80% of your product, is also secure? Do you maintain a team the does these things? A team that painfully scrutinises multiple news sources, makes sure that your application does not contain a vulnerable or out of date component, alerting your developers accordingly?
This is where SCA (Software Composition Analysis) like Meterian enters the scene. These products are a natural complement to tools like SonarQube or PMD, as they check the code your team did not write. Typically running along with your build system, they will prevent your product to go live if a vulnerable component is discovered and will provide your developers the information they need to fix the problem immediately (and, in some occasions automatically). You will be able to secure 80% of the code your development teams produce without investing a fortune, without further head count, and with a minimal integration effort.
So, if you do not have a solution for your opensource components in place, it won’t hurt taking a look at Meterian: it’s an all round affordable system that will give you peace of mind about the code that developers do not write. And, by the way, it also checks component licences, but this is for another blog post 🙂