Since autumn 2018, I have professionally used the infrastructure of AWS Amazon Web Services in productive use, after having previously looked after virtual servers at Hetzner. AWS web content is delivered through EC2 instances. The special feature of EC2 is the way this management platform provides virtual servers. This gives every administrator or devop numerous advantages.
Basically, in AWS you have more (virtual) servers than are actually productive for end users (eg test servers). Unlike hardware in the data center, AWS does not have a “conventional” server contract in the sense that you record a specific term in a contract.
AWS works like this:
- As an AWS customer you have a free basic quota of services.
- Leaving the free AWS basic contingent, costs are incurred (“on demand”).
This results in:
- There are no costs if you shut down the server services to zero or you are in the context of free services.
- The contract is not renewed automatically, but it is valid as long as you have an AWS user account.
- You can cancel the AWS contract by pocketing AWS’s user account.
- Billing at AWS is monthly, depending on the intensity of your own use.
This is what the AWS EC2 virtual server management interface looks like:
Each line in the AWS EC2 Server Dashboard shows a virtual server that exists in the cloud and is reachable via an IP address or the domain name pointing to it. You can delete, duplicate or relocate any of these servers within a few minutes. It is thus possible for AWS to move a server from the AWS Availability Zone “eu-central-1b” (Frankfurt) to any other Availability Zone worldwide … but with such a move away from Frankfurt/EU there would be a data privacy problem.
Here is an overview of the currently available AWS Availability Zones:
I particularly like about AWS EC2:
- Elastic IPs: move real-time IP addresses between servers.
- Create server snapshots in live mode.
- Duplicate snapshots to other instances.
- Scripting snap snapshots and automatic snapshots.