What the cloud actually changed
It is not merely someone else's computer. The real change was turning capital expenditure into an API call, and that has consequences most teams discover through their bill.
“The cloud is just someone else’s computer” is a good joke and a bad model. Renting a server was possible for decades before AWS, and plenty of businesses ran that way happily.
What actually changed was three specific things. Each is a real improvement, each has a consequence you will meet, and the expensive surprises come from taking the improvement and missing the consequence.
Capacity became an API call
Before: you forecast demand, bought servers, waited weeks, installed them. After: you request a server and have it in ninety seconds.
The consequence is that capacity decisions moved from a procurement process with approval gates to a developer with a credential. That is a genuine productivity gain and it is also why cloud bills surprise people — the control that used to be a purchase order is now a config file nobody reviews.
The second-order effect shows up on the invoice. Creation used to be the constrained step, so nobody had to think about deletion. Now creation is free and deletion is nobody’s job. An environment spun up for a two-week test runs for eleven months because it belongs to someone who changed teams and nobody is sure what breaks if it goes.
That is not carelessness. It is what happens when a gate is removed and nothing replaces it. If you want the speed without the sediment, expiry needs an owner: a tag saying who and until when, and a job that chases anything past its date.
Fixed cost became variable cost
A bought server costs the same whether it runs at 5 percent or 95 percent. A cloud server costs per hour, so idle capacity costs money continuously.
This is good if your load varies and you actually scale down. It is worse than owning hardware if your load is steady and high, which is why some companies have moved workloads back off the cloud and were right to.
The trap is paying variable-cost prices for a fixed-cost usage pattern. A database running at constant load around the clock on on-demand pricing is the most common expensive mistake in the category.
On-demand pricing is a premium for the right to change your mind by the hour. That right is worth a lot when load is spiky and nothing when it is flat. Every provider publishes a committed rate that is materially cheaper in exchange for a term. It goes undone because the bill arrives somewhere other than where it was created.
That is the opening. The rest of the lesson — the remaining sections, the exercise and the check-your-understanding questions — is behind the gate below.
This lesson requires full access.
Create an account to explore the published free lessons. This lesson requires full access; signing up does not purchase access or enroll you in Academy.