Building Multi-Tenant SaaS Architectures

Chapter 10. EKS (Kubernetes) SaaS: Architecture Patterns and Strategies

Up to this point, most of the topics we’ve covered have been focused on building a foundation of core concepts that would apply to any multi-tenant SaaS architecture. This information should have equipped you with a clear view of the multi-tenant landscape. Now it’s time to shift gears a bit and move from best practices strategies to looking more at how these concepts are influenced by the realities of different technology stacks. More specifically, in this chapter I’ll be focusing on how multi-tenant principles are landed within a Kubernetes environment. For our scope, I’ll actually be looking at Kubernetes through the lens of the Amazon Elastic Kubernetes Service (EKS). Much of what’s here will be applicable to any Kubernetes environment. However, there are areas where the managed nature of EKS influences our options.

I’ll start this chapter off by reviewing some of the key areas where I see good alignment between the EKS stack and SaaS architecture principles. The goal is to better illuminate some of the key factors that have teams selecting EKS as their preferred technology for building and developing their SaaS solutions. With that foundation in place, we’ll shift to exploring the different EKS deployment patterns that SaaS environments use to address their tiering, noisy neighbor, and isolation needs. We’ll look at a range of possibilities here, highlighting the different constructs you can use to define ...