IaaS definition
Infrastructure as a Service (IaaS) is a cloud computing model in which a provider rents out virtualized computing infrastructure, such as virtual machines, storage, networks and load balancers, on demand. Customers install and manage their own operating systems, middleware and applications, while the provider runs the physical hardware and data centers.
How does IaaS work?
The provider runs physical servers with a hypervisor that carves them into virtual machines. Customers choose an instance type with a set amount of CPU, memory and sometimes GPUs, attach block storage for disks, store files in object storage and connect everything through virtual private networks with firewall rules. All of it can be created and destroyed through an API, which is what makes automation and infrastructure as code possible.
Pricing comes in several forms. On-demand capacity is billed by the second or hour with no commitment. Reserved instances and savings plans discount steady workloads in exchange for a one- or three-year commitment. Spot or preemptible instances sell spare capacity at a deep discount but can be reclaimed at short notice, which suits batch jobs and fault-tolerant workloads.
Examples of IaaS
- Compute: Amazon EC2, Azure Virtual Machines, Google Compute Engine.
- Block and object storage: Amazon EBS and S3, Azure Disks and Blob Storage, Google Persistent Disk and Cloud Storage.
- Networking: virtual private clouds, load balancers, VPN gateways and DNS.
- Simpler providers: DigitalOcean Droplets, Akamai Cloud (formerly Linode), Hetzner Cloud and Vultr.
- Enterprise options: Oracle Cloud Infrastructure and IBM Cloud.
- GPU and accelerator instances for AI, from the major providers and specialist GPU clouds.
- Bare metal servers for workloads that cannot run virtualized.
IaaS vs PaaS vs SaaS
With IaaS you manage the operating system and everything above it: patches, runtimes, scaling scripts, backups and security hardening. With PaaS the provider also manages the operating system and runtime, and you deploy code. With SaaS you only use the finished application. IaaS offers the most control and flexibility, and also the most operational work, which is why many teams combine IaaS for special workloads with managed services elsewhere.
Responsibility for security shifts the same way. On IaaS, the provider secures the hardware and virtualization layer, while the customer is responsible for operating system patches, firewall rules, identity and access management, and encryption of data. Cloud security posture tools help check these settings continuously.
Common IaaS use cases
- Lift-and-shift migration of existing servers from an on-premises data center.
- Development and test environments created and destroyed on demand.
- GPU instances for AI model training and inference.
- High-performance computing and large batch processing.
- Disaster recovery sites that run only when needed.
- Legacy applications that need a specific operating system or configuration.
- Self-managed databases or software that needs full operating system access.
- Temporary capacity for load tests and data migrations.
Controlling IaaS costs and risks
The most common IaaS mistakes are oversized instances, forgotten resources and servers left running around the clock in non-production environments. Right-size instances using monitoring data, schedule development environments to shut down at night, use reserved pricing for steady workloads and tag every resource with an owner. Nexzem sets up IaaS environments with infrastructure as code, tagging and budget alerts from day one, so clients can see exactly what each workload costs.