Discussing our plan

We’ve seen how to use Google Cloud Run as a CaaS solution in a previous lesson. Let’s see whether we can do something similar in AWS. For that, the best candidate is probably ECS combined with Fargate.

Amazon Elastic Container Service (ECS) is a scheduler and orchestrator of containers in a cluster. Unlike Kubernetes, ECS is proprietary and closed source. We don’t know what is running the containers. Depending on the importance we give to understanding the underlying services, that might or might not matter. Just like other container orchestrators, it has a control plane and agents that take instructions from it. In turn, those agents are in charge of running containers, and quite a few other auxiliary tasks.

An ECS service is in charge of fault tolerance, high availability, and scaling. It’s similar to AWS Auto Scaling groups but focused on ECS tasks. It defines the number of tasks that should run across the cluster. It also decides where they should be running (e.g., in which availability zones), and it associates them with load balancers. On top of all that, it can scale tasks automatically based on metrics (e.g., CPU and memory utilization). We could say that it performs similar operations as Kubernetes Scheduler, with Pods, Services, and HorizontalPodAutoscalers.

Note: Given that AWS has Elastic Kubernetes Service (EKS), you might be wondering why we’re providing examples in ECS. The reason is simple. We’re exploring CaaS solutions which abstract underlying technology (e.g., ECS, EKS, etc.). At the same time, AWS is still pushing for ECS and considers it a preferable way to run containers, so we’re just going with the flow and ...

Before We Get Started

Infrastructure as Code (IaC)

Google Kubernetes Engine (GKE) Clusters with Terraform

Manage Infrastructure using Terraform on Google Cloud Platform

AWS Elastic Kubernetes Service (EKS) Clusters with Terraform

Azure Kubernetes Service (AKS) Clusters with Terraform

Packaging, Deploying, and Managing Applications

Deploying a Web Application over Kubernetes

Using Helm as a Package Manager for Kubernetes

Hands-On: Using Helm as a Package Manager for Kubernetes on AWS

Exploring Serverless Computing

Using Managed Functions as a Service(FaaS)

Hands On: Using Managed Functions as a Service(FaaS) on AWS

Serverless as the Future

Using Managed Containers as a Service(CaaS)

Hands On: Using Managed Containers as a Service(CaaS) on AWS

Using Self-Managed Containers as a Service (CaaS)

Hands-On: Using Self-Managed Containers As A Service on AWS

Using Centralized Logging

Hands-On: Using Centralised Logging

Deploying Applications Using GitOps Principles

Hands-On: Deploying Applications Using GitOps Principles

Applying Progressive Delivery

Hands-On: Applying Progressive Delivery

Conclusion

Appendix

Getting Started with Amazon Elastic Container Service (ECS)

Discussing our plan