Elastic Container Service

Solution Architect Associate

Developer Associate

A highly scalable, fast container management service that makes it easy to run, stop and manage Docker containers on a cluster of EC2 instances
Regional service that you can use in one or more AZs across a new or existing VPC to schedule placement of containers across your cluster
Eliminates the need for you to operate your own cluster management and configuration management systems, or to worry about scaling your management infrastructure
ECS can be used to create consistent deployment and build experiences, as well as manage and scale batch workloads
Amazon Elastic Container Registry (ECR) is a managed Docker registry (their version of Dockerhub)
- Can use IAM to restrict access to repositories of images
Integrates with various load balancer solutions within AWS

Required to run Docker containers in ECS. Can choose between EC2 and Fargate
JSON files that describe one or more containers that form your application
Contains info on what docker image to use, networking modes, ports to map to the host, data volumes, etc.

Allows you to run and maintain a desired count of instances of a task definition
Kind of like Auto-Scaling groups for ECS

A logical group of container instances that you can place tasks on
You get one by default when you first use the ECS service, but you can create multiple clusters in an account
Clusters are region specific
Can use IAM to restrict access to specific clusters
Containers can be a member of only one cluster at a time
Can contain multiple different types of container instance types

Service Scheduler: Ensures that a specific number of tasks are consistently running and reschedules tasks when they fail
Custom Scheduler: You can create your own schedulers to meet your business need, including leveraging 3rd party services

Fargate

A serverless compute engine for containers that works with both ECS and EKS
Allows you to continue using ECS/EKS for scheduling, but AWS manages the underlying infrastructure
It only works with Linux workloads
When using Fargate, you have to define the amount of CPU and Memory to allocate to your container

EC2
- The EC2 option allows you to use the EC2 pricing model, making it more cost effective. It also excels at scenarios with long-running containers.
- With EC2, you can put multiple containers on a single host
Fargate
- With Fargate, you lose all operating system access
- You only pay for the compute time and resource allocation that you setup within Fargate
- Excels at short running tasks that pop up and can go away
- Containers are all isolated