AWS Amazon EC2 instances - Free Tier?

Question

Amazon Web Services

First, statistics. A clean CentOS 7 was installed, nothing was installed additionally, the VPS was not used in any way this time, statistics on the consumption of free allocated resources:
5.04-17-00

spoiler

6.04.-16-00

spoiler

07.04.-17-00

spoiler

08.04-17-00

spoiler

09.04-17-00

spoiler

Further, instead of CentOS 7, Ubuntu Server 18.04 LTS was installed. Also, nothing was added and VPS was not used in any way. Statistics:
12.04-18.00

spoiler

13.04-16-00

spoiler

14.04-19-30

spoiler

Free limits allocated per month:

spoiler

spoiler

As you can see, not only are the resources themselves consumed, on a clean OS, and even if I don’t use it, they are also spent above the limit. Naturally, I wrote on this issue to the support with a detailed description of the situation and the attachment of all screenshots, here is the answer:

spoiler

I understand your concern about using AWS EC2, and I will be happy to help you.

I see that your account currently has one instance of t2.micro linux in a “running” state under US West (Oregon) region.

As you can refer to the AWS Pricing page, you can see that AWS takes into account EC2 usage on demand based on hours the instance was running rather than CPU usage of the running instance.

You can learn more about how an EC2 instance counts usage here: https://aws.amazon.com/premiumsupport/knowledge-ce...

This means that when you run an instance from your AWS account and when it is in a "running" state, it is completely allocated for your use, and it doesn't matter if CPU usage is 100 or idle.

He assures me that only consumed hours are taken into account, within the allocated 750 hours per month. At the same time, in the screenshots of the "Top Free Tier Services by Usage" statistics I provided, there are limits for at least 3 items, including "30 GB of Amazon Elastic Block Storage in any combination of General Purpose (SSD) or Magnetic". Actually how to be in this situation?

Answer 1

Everything is correct. As long as you have a running resource, you pay for it. For EBS while the disk exists (in general, regardless of its employment), and for EC2 for the entire time of the running resource. Even if there are no uses. Dawns are produced regardless of the number of instances and disks - only for the actual amount of resources used.