Introduction to Amazon ECS

Amazon Elastic Container Service (ECS) is a fully managed container orchestration service that lets you run, stop, and manage containers on a cluster. ECS removes the need to install, operate, and scale your own cluster management infrastructure — you describe what you want to run, and ECS handles the rest.

---

What Is Container Orchestration?

Running a single container on your laptop is simple. Running hundreds of containers across multiple servers in production is not. Container orchestration solves the hard problems:

Problem	What Orchestration Provides
Scheduling	Deciding which host should run each container
Scaling	Adding or removing container instances based on demand
Health monitoring	Detecting failed containers and replacing them
Load balancing	Distributing incoming traffic across containers
Service discovery	Allowing containers to find and communicate with each other
Rolling updates	Deploying new versions without downtime
Resource management	Ensuring containers get the CPU and memory they need

Amazon ECS is AWS's native answer to container orchestration.

---

ECS Core Concepts

Understanding ECS requires four key concepts:

graph TD
  Cluster["Cluster"] --> Service["Service"]
  Service --> Task1["Task"]
  Service --> Task2["Task"]

1. Cluster

A cluster is a logical grouping of tasks and services. It is the top-level resource in ECS. A cluster can run tasks on:

• AWS Fargate (serverless — no EC2 instances to manage)
• EC2 instances (you manage the underlying hosts)
• A mix of both (capacity providers)

Think of a cluster as a pool of compute capacity where your containers run.

2. Task Definition

A task definition is a blueprint that describes how to run one or more containers. It is similar to a docker-compose.yml file but in JSON format. A task definition specifies:

• Which container images to use
• How much CPU and memory to allocate
• Port mappings
• Environment variables
• Logging configuration
• IAM roles
• Volumes

We will explore task definitions in detail in the next lesson.

3. Task

A task is a running instance of a task definition. When ECS starts a task, it pulls the container images, configures networking, and starts the containers as described in the task definition.

Tasks can be:

• Standalone tasks — run once and stop (e.g. a batch job, database migration)
• Part of a service — kept running continuously (e.g. a web server)

4. Service

A service maintains a desired number of tasks running simultaneously. If a task fails or stops, the service scheduler launches a replacement. Services also integrate with load balancers to distribute traffic.

---

Launch Types

ECS supports two primary launch types that determine where your containers run:

EC2 Launch Type

You manage a fleet of EC2 instances that are registered to your ECS cluster. The ECS agent runs on each instance and communicates with the ECS control plane.

graph TD
  subgraph Cluster["ECS Cluster"]
    subgraph EC2A["EC2 Instance (ECS Agent)"]
      TA["Task A"]
      TB["Task B"]
    end
    subgraph EC2B["EC2 Instance (ECS Agent)"]
      TC["Task C"]
      TD["Task D"]
    end
  end

When to use EC2 launch type:

• You need access to the underlying host (e.g. GPU workloads, custom AMIs)
• You want to use Reserved Instances or Spot Instances for cost savings
• You have compliance requirements that mandate specific host configurations
• You need to run privileged containers or use specific Linux capabilities

Fargate Launch Type