Infrastructure as a Service (IaaS): Renting the Data Center

WHY IT EXISTS Companies needed to escape the high cost, capital expense, and slow pace of managing their own physical data centers. IaaS offers the flexibility of on-demand computing resources without the long-term commitment of buying and maintaining server hardware, letting teams scale up or down as needed.

THE MENTAL MODEL Think of IaaS as leasing a vacant commercial building. The landlord provides the physical space, electricity, and plumbing (the servers, network, storage), but you are responsible for everything inside. You bring your own ovens, staff, and recipes (the OS, applications, and data) and manage it all. You have full control over your operations but don't own the building.

HOW IT WORKS A cloud vendor owns and operates massive data centers full of servers, storage arrays, and networking gear. Through a web console or high-level APIs, you request virtual resources like servers (virtual machines), block storage, and network configurations. The provider provisions these for you, and you pay for what you use. You then access these virtual machines to install your operating system, software, and applications, just as you would with a physical server.

WHEN TO USE IT Use IaaS when you need maximum control and flexibility over the computing environment. It's a good fit for migrating existing on-premise applications to the cloud without re-architecting them (a "lift and shift"), running legacy software that requires a specific OS version, or for teams with deep infrastructure expertise who want to build a custom platform from the ground up.

WHEN NOT TO USE IT Avoid IaaS if your goal is simply to run application code without managing servers. If you don't want to handle OS patching, system administration, and scaling infrastructure yourself, a Platform as a Service (PaaS) or Serverless model is a much better fit, as they abstract away the underlying infrastructure management.

ONE CANONICAL EXAMPLE Provisioning a virtual machine like an Amazon EC2 instance or a Google Compute Engine VM. You choose the CPU, RAM, and storage, select an operating system image (like Ubuntu or Windows Server), and get remote access. From there, you are responsible for installing a web server, a database, and your application code.