Building Multi-Tenant SaaS Architectures

Chapter 11. Serverless SaaS: Architecture Patterns and Strategies

The serverless computing model has become quite popular with builders. Consuming compute in an entirely managed model where there are no servers enables SaaS builders and architects to shift their mindset away from chasing elusive scaling and cost optimization strategies. The function-centric nature of serverless compute can also influence how you approach the design and implementation of your multi-tenant SaaS architecture. For these reasons, I thought it made sense to devote a chapter to looking at how multi-tenant strategies are realized within a serverless compute model. The goal here is to dig into the specific nuances and implications that come with building a SaaS environment that delivers its application services via a serverless compute. To make this more concrete, I’ll map these strategies to the AWS Lambda service, which provides the managed compute capabilities that will configure, host, and scale the functions of our environment.

At the outset of this chapter, I will start by outlining the natural alignment between the profile of SaaS environments and the serverless model. We won’t spend too much time here, but I think it’s essential for SaaS architects, builders, operations, and business stakeholders to understand the broader value proposition that serverless represents for SaaS providers. I touched on this briefly in Chapter 8 as part of looking at how serverless can influence storage. Now, though, ...