Serverless architecture is a fancy new jargon in the architecture world. There is lot of buzz around the tech community, especially with the introduction of AWS Lambda and API gateway. Of course there are lot of debates going on about what we need to consider as Serverless and not. One important thing to keep in mind is that Serverless does not mean that you run your application without any physical server. At the end of the day, your code has to run on some hardware and you cannot avoid that. It is just that whether you maintain the hardware (deployments, upgrades, scaling issues and so on) and software (in some cases) or someone else does it on your behalf. Serverless is often linked to low cost assuming the Lambda pricing, but it is not. Pricing is up to the cloud service provider and it also depends on how much resources you consume. Startups definitely gets some pricing benefits especially with thousands of free seconds (offered by Lambda) and with zero entry cost. But for mid-sized to big organizations, the problem it solves in the application stack matters. It boils down to “One less problem to worry“.
I will soon write another detailed post about my experience with AWS Lambda and what you can do extract more juice out of the Lambda sugarcane.