Everyone in tech uses cloud terminology. Almost nobody can define it under pressure. Here are the 10 terms that trip up learners, interviewers, and even working engineers — with the plain-English definitions that actually stick.
1. IaaS, PaaS, and SaaS
These three are always listed together, which makes them feel interchangeable. They're not.
- IaaS (Infrastructure as a Service) — You rent the raw hardware: servers, storage, networking. You manage everything above it. Example: AWS EC2.
- PaaS (Platform as a Service) — You rent the hardware and the operating system layer. You manage only your application code. Example: Heroku, Google App Engine.
- SaaS (Software as a Service) — You rent the finished software. You manage nothing. Example: Gmail, Salesforce.
The pattern: each step up the stack, you own less and the provider manages more.
2. Serverless
Serverless does not mean "no servers." It means you don't manage the servers. The cloud provider provisions, scales, and tears down compute on demand. You write a function, the provider runs it when called, and you pay only for execution time. AWS Lambda is the canonical example.
The confusion: calling it "serverless" when servers clearly exist. Better mental model: server-invisible.
3. Scalability vs. Elasticity
These are often used as synonyms. They're related but distinct.
- Scalability is the ability to handle increased load — either by scaling up (bigger machines) or scaling out (more machines).
- Elasticity is the automatic scaling up and back down in response to real-time demand.
A system can be scalable without being elastic (you manually add servers). Elastic systems are scalable by definition, but add the automation layer.
4. Multi-Cloud vs. Hybrid Cloud
- Multi-cloud — Using multiple public cloud providers (AWS + Azure + GCP). Usually for redundancy, cost optimization, or avoiding vendor lock-in.
- Hybrid cloud — A mix of private infrastructure (on-premises data center) and public cloud, integrated to work as one environment.
The key distinction: multi-cloud is public + public. Hybrid cloud is private + public.
5. Region vs. Availability Zone
- Region — A geographic area with multiple data centers. Examples: us-east-1 (Northern Virginia), eu-west-1 (Ireland).
- Availability Zone (AZ) — An isolated data center within a region. Each region has 2–6 AZs, physically separate but connected by low-latency links.
Deploying across multiple AZs protects against a single data center failing. Deploying across multiple regions protects against an entire geographic area going down.
6. Containers vs. Virtual Machines
Both run isolated workloads. The difference is what they isolate.
- Virtual Machines (VMs) — Virtualize an entire machine including the operating system. Each VM has its own OS kernel. Heavy but fully isolated.
- Containers — Virtualize the application layer only. They share the host OS kernel. Lightweight, fast to start, portable.
Docker is the most common container runtime. Containers are smaller and faster than VMs; VMs offer stronger isolation.
7. Kubernetes (K8s)
Kubernetes is a container orchestration platform. It automates the deployment, scaling, and management of containerized applications across a cluster of machines. If Docker answers "how do I package and run one container," Kubernetes answers "how do I manage thousands of containers across hundreds of machines."
Common confusion: people treat Docker and Kubernetes as competitors. They're complementary — Docker builds and runs containers, Kubernetes manages them at scale.
8. Object Storage vs. Block Storage
- Block storage — Data stored in fixed-size blocks, like a traditional hard drive. Fast, low-latency, used for databases and OS volumes. Example: AWS EBS.
- Object storage — Data stored as objects with metadata and a unique ID. Infinitely scalable, cheap, designed for unstructured data like images, videos, and backups. Example: AWS S3.
The rule of thumb: databases and OS disks use block storage. Files, media, and backups use object storage.
9. CDN (Content Delivery Network)
A CDN is a network of servers distributed globally that cache and serve static content (images, CSS, JavaScript, videos) from the location closest to the user. Instead of every user hitting your origin server in Virginia, a user in Tokyo gets content from a CDN node in Tokyo.
Result: faster load times, reduced origin server load, and better availability under traffic spikes.
10. VPC (Virtual Private Cloud)
A VPC is a logically isolated section of a public cloud where you can launch resources in a virtual network that you define. You control the IP address range, subnets, routing tables, and network gateways. It's essentially your private data center inside the public cloud, with full network control.
Without a VPC, your cloud resources are in a shared network. With a VPC, you define who can talk to what.
The pattern across all 10: Most cloud confusion comes from names that obscure what they actually do. "Serverless" has servers. "Containers" don't contain like boxes. Once you strip the marketing language and focus on what each thing manages, the distinctions become clear and memorable. The fastest way to lock them in is retrieval practice — testing yourself until you can define each term without looking.