A friend of mine recently called me and asked me about Serverless computing. He's about to take a job at VMware and he told me that he's going to be working on Serverless technology, when he begins work there. He was curious if I knew anything about Serverless tech and wanted me to give him a brief 101.
I did not know anything about it, so I looked deeper and whatever I found from my primary research tells me that Serverless is going to change the DevOps world. No, scratch that. It's going to change the way software engineering is done.
Our production costs are going to come down by a huge huge factor
So, what's Serverless?
Serverless computing allows you to build and run applications and services without thinking about servers.
Doesn't make a lot of sense, does it? Like most fancy text book definitions. Let me try to break it down for other Serverless rookies like me.
Right now, we have our APIs running 24*7 and they serve requests. How about a scenario where the server is going to be active only for the duration for which it receives the request? And hey, here's a bonus. You don't have to worry about scaling them.
That's the whole point of serverless, you don't have to think about servers even though they exist.
A real life example:
Say, you have an app like Facebook. Think of user login. There will be a route in the backend that logs the user in.
In a Serverless situation, the backend on receiving the request activates the code that's supposed to handle the logging in functionality, logs the user in and goes back to sleep.
It's going to active only for say a fraction of a second, opposed to what we have now: APIs being active all the time (and not to mention, we get billed for the same).
I have always believed that those who adapt faster, innovate faster. During the course of this week, I am going to be trying out the various solutions available out there by building sample applications and get my hands dirty with Serverless.
AWS Lambda, Google Functions and Azure Functions are the choices I have to play around with. If all goes well and we're convinced enough, I am going to try and build our next microservice at Hashnode in a serverless fashion.
I will be back with my findings next Monday.