Are applications teams prepared to manage the chaos arising from an ever-growing landscape of heterogeneous deployment types? A recent survey of application and operations professionals sought to better understand how the industry is shifting and what the future of DevOps might look like. Here is what the survey uncovered ...
The Twelve-Factor App is a methodology that offers a 12-step best practice approach for developers to apply when building software-as-a-service apps that are both scalable and maintainable in a DevOps world. As software continues to be written and deployed at a faster rate and in the cloud, development teams are finding there is more room for failure and vulnerabilities. This blog series will discuss how to build a Twelve-Factor app securely.
This first blog in the series looks at Step I, Codebase, in more detail.
Defining Codebase in the Twelve-Factor App
According to 12factor.net, a Twelve-Factor app is always tracked in a version control system, for example Git, Mercurial or Subversion. A copy of the revision tracking database is referred to as a "code repository," which developers often shorten to "code repo" or simply "repo."
A "codebase" is therefore any single repo, or set of repos, that share a common root, in a decentralized revision control system. The Twelve-Factor advocates putting all code into one repository or source control system that is tracked in revision control. So essentially there is only one codebase per app, but there can be many deploys of the app.
Applying Security to Codebase
Having one codebase tracked in revision control, with many deploys is the ideal, but it’s important to remember that your code is only as secure as the systems used to build it. As the guardian of the code and the systems in which it is housed, what steps can you as a developer take to ensure that the repository is properly secure?
1. Lock it down
These days source code is the core of most businesses, therefore it’s imperative to protect it. The best way to do this is by imposing restrictions — limit access to the source code repository to authenticated users only. Believe it or not it’s very common for people to be able to view the entire source code when on the internal network. While this maintains integrity, unfortunately it can put your entire intelligent property at risk.
Limit commits to pull requests as opposed to direct commits. This refers to a source management system model, in which no one can write directly to the source code repository. Pull requests are reviewed by the team responsible for the master code, which then decide whether to merge the new code with the existing base, or not. In this way the codebase is protected through the provision of checks and balances. Ultimately, limiting commits is an enabling requirement that acts as a shield for other activities.
2. No more secrets
Write code as if everything you did was open sourced. To share a practical example of this, there’s a scene in the movie Sneakers (1992) where Robert Redford’s character realizes that "Setec Astronomy," the name of the project that produced a black box, was actually an anagram for "Too Many Secrets." Soon after, Redford’s engineers realized that the black box was actually a piece of hardware capable of decrypting anything and everything — to which Redford then famously (at least for me) said "No More Secrets." In the same way we need to apply the idea of "No More Secrets" to our code. It means developers should write application code with the assumption that it could be disclosed, so ideally the goal will be to have no secrets in the actual application.
Externalize secrets and make their location known via startup arguments. This means putting secrets into configuration files where people can find them. Use the environment to protect the configuration.
3. Review code regularly
Enforce secure coding guidelines via a review of all pull requests. Throughout the software development stage, there are only a few key inflection points i.e. central areas where everything merges, where changes can be made. It’s a good policy to use these central points to enforce your company’s security guidelines. From a coding point of view these guidelines can include:
a. Use organizationally approved cryptographic controls.
b. Make use of HTTP security headers for all responses.
c. Consistent use of contextual encoding via HTML templating engines.
d. All database queries are performed using variable binding.
e. Keeping your dependencies updated.
Successful development teams find it’s equally important is to educate the team regarding solutions to pull requests that fail code reviews. It’s key to understand why some pull requests are rejected, and how the team can work to avoid these issues.
Collaborating in terms of sharing feedback will go a long way towards promoting awareness, and ultimately supporting the coding of applications that can be done more quickly, efficiently and securely. It’s an approach that should in fact be applied to all 12 steps of Twelve Factor.