Why is using the Cloud referred to as going ‘serverless’ when the companies offering Cloud services are STILL using servers?
In: Technology
Serverless just means u don’t handle the servers yourself. The cloud provider manages everything, so u can focus on coding.
Because the person renting the server doesn’t have to worry about all the nitty gritty of owning a server. I don’t worry about the power supply going bad, or selling it when I don’t need it, or the cost of buying it. It costs me $x/hr to be online, and that’s it.
I’ll also add that many of the server less structures can also deal with all the headache of worrying about the operating system too, so you just worry about the application, and ignore the rest.
Imagine you have a task of moving from point A to point B.
You can get your own car and drive there, or you can call an Uber.
In the second case – the car, maintenance, parking, gas, etc are not your concern because you don’t own the car, you just get a service moving between points. The car is still there, but you don’t care about any of that.
The same goes to “serverless”. You don’t manage servers, hardware, software, backups, etc. You just push your code to them, and they handle it.
Serverless = you don’t have to think about servers. Yes the cloud provider is using servers, but instead of the usual way (even with virtual servers) of writing software in terms of running it on a number of individual servers and passing data between servers, you write the software as if it was just running on one big machine and the cloud provider automatically splits it across however many actual servers it requires.
[There is NO CLOUD, just other people’s computers](https://fsfe.org/contribute/promopics/thereisnocloud-bluecolor-preview.png)
The answer is mostly “Marketing for C-level”. At the core of most Cloud services is still a server, but at the core this is just an extension of the whole “Managed infrastructure” movement, where companies decided they don’t want a dedicated team to run & optimize physical hardware but would rather just have to worry about the software running on it, trading some CAPEX for more OPEX.
If you didn’t own a car and instead used Ubers to get around, you could still call yourself “carless” despite the fact that you relied on other people’s cars.
You don’t have to worry about the fuel, insurance, registration, license, tyres, servicing etc. You leave that to the people providing the car for you.
It is a bit like claiming you are “car-less” but rely on taxis to take you everywhere. Yup, technically you don’t own a car but you’re expecting someone else to own a car and you just pay them for transportation.
In any case “servers” is a really broad term.
It’s kind of like food delivery. When you had a server, it’s like you cooked every meal by yourself, you had to buy ingredients, cool them, eventually you have a meal. Cloud is like delivery, you just order and everything is taken care of and you get your meal all prepared. You don’t have to think of cooking or buying ingredients.
There is a concept called “abstraction”
What this means is complex stuff gets simplified. For example if I say 2+2=4 you can ask “but why” and then we get into some complicated maths to prove it. All you really ned to know is 2+2=4.
Same with cloud. In the old days, you used to have a server (a physical machine) that would have your website (or Whatever) bit you would have to go and restart it if it fucked up. Or you would have to do upgrades. Or of you got a lot of traffic you’d have to buy more RAM and install it, or get a bigger hard drive.
The cloud is basically someone (like google or Amazon or Microsoft… There are many) who have bought a whole bunch of servers, and written a whole bunch of clever software to manage all this.
So you can put up your website and if it has 1 visitor or 1 million visitors you don’t care because the clever software can handle all this stuff and spread it around multiple servers. You don’t even know it’s happening. There is a lot of complicated stuff happening behind the scenes.
So “cloud” just means all the stuff that you used to do (servers) is now part of a vast network of servers and infrastructure that is coordinated by some.imcredibly complex software that allows you to not care about servers, just like you don’t care who or how your burger was prepared and cooked before the delivery guy gave it to you.
Serverless means your code is not being hosted on any server until you’re trying to execute it. Then the provider would send your function to one of their available servers to execute. When you stop executing it for a short while, it will get removed from the server. Now your function is serverless again.
State of serverless means you’re not being billed when the function is not executing.
I think there is a fundamental misconception underlying this question, maybe because of terrible messaging by hype companies misusing a buzzword. Not all cloud computing is “serverless”. In fact, only a tiny portion of cloud computing is “serverless”.
Most cloud computing involves customers being aware of the nodes (e.g. AWS EC2 instances) and storage that they are renting, and paying for by the hour. Customers don’t manage the server hardware, but their VMs are “servers” that they still provision and manage.
“Serverless” is a small niche, provided by products like AWS Lambda. You don’t provision a server or RAM or anything from their cloud. You provide a “function”, some bit of code, and you tell AWS Lambda: “Every time someone hits this URL, run my function/code, giving my code the URL parameters”. So you don’t have any of your code, or any VM, running in their cloud minute-by-minute. You only have code running in their cloud if-and-when someone hits your URL.
You pay per function invocation, based on how much CPU and RAM each function invocation takes.
Latest Answers