What is serverless architecture?

Asked

Viewed 301 times

16

Could someone briefly explain what a Serverless architecture is? After reading a few pages here, I was in doubt. Of Wikipedia, basically sums up what’s out there:

[...] A cloud computing Execution model in which the cloud Provider dynamically manages the allocation of machine Resources

But this doesn’t seem to be much different from AWS services. Or we can also understand that it seems to be a form of Amazon Lambda, where it stores functions that run in Nodejs (and other languages) and return results of CRUD operation. After all, is it like an API? I’m a bit confused.

Update

After writing this post I continued with my research and found this video very interesting and which corroborates what has been said so far.

2 answers

14


Amazon itself made its definitions of this architecture here (my emphasis)

build and run applications without [having to] think about on servers

It is not that it is without servers, it is that the focus on it will be less.

Amazon Lambda is in vogue when we talk about serverless and AWS because you write code without worrying about which server, container or virtual machine it will run.

The big gain is in the development of the application, since the focus will be on it and not on server configuration and these Devops blah blah blah blah. Worth a read on noops culture. From the AWS website, same page I quoted earlier:

decrease [server management] overhead enables developers to have the energy and time to spend developing good products that scale and are reliable

There is no way to apply everything using this architecture, there are cases (maybe most), that the best would be to manage the server, the scalability and the availability of the application.

In the definition you brought from Wikipedia, I can emphasize

is a cloud computing Execution model in which the cloud Provider dynamically manages the allocation of machine Resources

It’s Azure, AWS or Google Cloud Platform that would manage the server on issues of scalability, resource allocation, location and so on, while you and your team would focus on application development.

It’s an interesting development paradigm, but obviously there are controversies. It’s good to be in control too. Question everything!

"A human being undoubtedly does not evolve. Questioning is the first step in opening ourselves to the new." - Roberto Shinyashiki

  • I guess we’re forced to do stateless like that, huh? When you manage the application server, there are people who are doing something that turns out to be stateful.

11

One more marketing and wrong name. I like the idea of not having a server, but it is not the case of this term.

It’s a way where you "buy" services and not structure. You don’t have to know anything about the server, where it is, capacity, anything. You only use and pay for what you use individually, it can even be the function execution you developed and climbed in the cloud. You don’t have to know any details, you are only service user.

AWS, Azure and Google Cloud Platform are increasingly strengthening this idea, not least because it is probably more financially advantageous for them. Not everything, of course, but many of these and other cloud providers' services are based on the ill-called serverless.

On the API, it depends on the service. The AWS Lambda is not quite an API in the sense that you might be thinking, because most people understand API the wrong way. It even has an API to handle the service, but what you consume is not an API, at least provided by them. It may be an API, but you developed and put it there. But there are other services that are their Apis.

  • 1

    Tbm agree with your reply.

Browser other questions tagged

You are not signed in. Login or sign up in order to post.