Mostly serverless

Even EC2 has serverless attributes, and you can leverage them to realize the benefits of serverless in your classic enterprise cloud architectures. John Chapin shares the true story of an enterprise IT organization for which a potent combination of mostly serverless technology and a DevOps mindset have laid the groundwork for a future serverless transformation.

For those of us in the serverless community, technologies like AWS Lambda, API Gateway, and the Serverless Application Model seem like old hat. We’ve built and scaled applications, dealt with the idiosyncrasies, and espoused the value of a serverless approach for the past couple of years. However, for many large enterprises, a serverless transformation (or even a proof of concept) is still years away. But many of these more cautious companies are already dabbling in a serverless world, without even realizing it. John Chapin shares the true story of an enterprise IT organization for which a potent combination of “mostly serverless” technology and a DevOps mindset have laid the groundwork for a future serverless transformation. John explains the latent serverless properties of legacy AWS services like S3, EC2, and Elastic Beanstalk and shows that while all those technologies don’t necessarily meet the strict criteria for truly serverless systems, they can still be used in a “mostly serverless” manner. For example, treating EC2 instances as stateless, ephemeral components means that, from an architectural perspective, they are interchangeable, and you can simply deploy code to them with little or no manual system configuration or tweaking. And, of course, S3 is one of the original serverless services.