Copied directly from the link:

Servers or Serverless - Which is better from a cost perspective?

It really depends upon the use case!

From a financial perspective, it’s a rent vs buy decision.

If a system is used sporadically, then serverless is often cheaper.

If a system is used constantly – serverless can be 10 X more expensive!

You can think of it in the same manner as buying a car vs taking an Uber.

For example – if you live in a city and take ten $25 dollar Uber trips per month the cost is $250

The cost of buying a car, fuel, plus insurance and parking could be $1000 - $2000 per month in most cities.

In the above case – it’s cheaper for this individual to use Uber.

But what happens when a person needs six - $20 Uber trips per day? Now that’s $120 dollars per day or $3600 per month.

In this case, it’s far cheaper to buy a car than to take an Uber.

As a bonus to this second scenario, in the end the person still has the car and can continue to drive the car or sell the asset when the payments are finished.

This is essentially the same calculation of server vs serverless on the cloud.

So, which is better from a cost perspective server or serverless?

It depends upon how the systems are used.

If the systems are used infrequently then serverless is often cheaper. If the systems are used consistently then serverless is much more expensive.

Now you have a fundamental understanding of servers and serverless and their costs.

  • @butternutsOP
    link
    21 year ago

    For me I really like serverless when I’m trying to build a proof of concept or demo for my personal projects. A lot of the services are on the free tier so I’ve built many APIs with user auth, websockets, and ton of other features for nothing in price.