IaaS, PaaS and SaaS

LPIC1-101

2/21/2026

1. IaaS (Infrastructure as a Service)

The Concept: You get the "Raw Materials." The provider gives you virtual hardware (CPU, RAM, Storage, Networking), but you have to install the OS, manage the security patches, and install the software.

Your responsibility: Everything from the Operating System upward.
Example: Amazon EC2 or Google Compute Engine.

2. PaaS (Platform as a Service)

The Concept: You get a "Workbench." The provider manages the hardware and the Operating System. You just bring your code (Python, Java, PHP).

Your responsibility: Just your application and its data. You don't care about yum update or fstab here.
Example: Heroku, Google App Engine, or AWS Elastic Beanstalk.

3. SaaS (Software as a Service)

The Concept: You get the "Finished Product." You access the software via a web browser. You don't manage anything technical at all.

Your responsibility: Just your own settings and the data you enter.
Example: Gmail, Dropbox, Slack, or Office 365.

Difference between Load Balancing and Elasticity (IaaS)

This is a great "big picture" question. In a Cloud Service Provider (CSP) environment, these two concepts work together like a team, but they solve different problems.

Think of a busy Highway Toll Plaza.

1. Load Balancing (The Traffic Cop)

Load Balancing is the act of distributing incoming traffic across multiple servers. Its goal is to make sure no single server gets overwhelmed while others are sitting idle.

Purpose: High Availability and Efficiency.
How it works: It sits in front of your servers. When a request comes in, the Load Balancer says, "Server A is busy, go to Server B."
Analogy: A toll plaza with 4 booths open. The traffic cop directs cars into the shortest line so everyone gets through faster.

2. Elasticity (The Magic Expanding Wall)

Elasticity is the ability of the system to automatically grow or shrink its resources based on demand.

Purpose: Cost-efficiency and Scalability.
How it works: If your 2 servers are both at 90% CPU, Elasticity (Auto-scaling) triggers the creation of a 3rd and 4th server. When the rush is over, it deletes them so you stop paying for them.
Analogy: If the line at the toll plaza gets too long, a magic wall opens up and creates 4 more toll booths instantly. When the traffic dies down, those booths disappear.

3. How they work together (The "Cloud" Secret)

In a modern IaaS setup (like AWS or OpenStack):

The Elasticity service monitors your servers.
When it sees a spike, it creates a new VM (using cloud-init to set it up automatically!).
The Load Balancer automatically detects the new VM and starts sending it traffic.

LPIC-1 Context: While the exam focuses on the individual Linux server, understanding this "Cloud Logic" is why we learn tools like cloud-init. A server needs to be able to configure itself without a human so that Elasticity can work!

Summary

Load Balancing = Spreading the load you already have.
Elasticity = Changing the amount of infrastructure you need.