Building Secure and Reliable Systems

Chapter 7. Design for a Changing Landscape

By Maya Kaczorowski, John Lunney, and Deniz Pecel

with Jen Barnason, Peter Duff, and Emily Stark

Your ability to adapt while fulfilling the service level objectives promised to your users depends on the robustness and flexibility of your service’s reliability capabilities. Tools and approaches like frequent builds, releases with automated testing, and containers and microservices will enable you to adapt to short-, medium-, and long-term changes, as well as unexpected complications that arise as you run a service. This chapter also presents multiple examples of how Google has changed and adapted its systems over the years, and the lessons we learned along the way.

While most design decisions—particularly those relating to architecture—are easiest and cheapest to implement in the system design phase, many of the best practices in this chapter can be implemented during later stages of the system lifecycle.

“Change is the only constant” is a maxim¹ that certainly holds true for software: as the number (and variety) of devices we use increases every year, so too does the number of library and application vulnerabilities. Any device or application is potentially susceptible to remote exploit, data leakage, botnet takeover, or other headline-grabbing scenarios.

At the same time, the security and privacy expectations of users and regulators continue to rise, requiring stricter controls like enterprise-specific access restrictions and authentication ...