re:Invent Guide: Containers


Container Hero

This guide is for re:Invent attendees who are looking for sessions focused on container technology. Back in 2013, when I was at one of the first Docker meetups, I realized how powerful Docker could be and hoped that it would grow. In 2014, I presented at DockerCon as a use-case speaker. Since then, I’ve been building tools like ufo using Docker and Amazon Elastic Container Service (Amazon ECS) as my small part in helping the ecosystem grow. Fast-forward six years and that ecosystem is now huge. Within the AWS world only there’s Amazon Elastic Container Registry (Amazon ECR), Amazon ECS, Amazon Elastic Kubernetes Service (Amazon EKS), AWS Fargate, AWS App Mesh, and more. Containers continue to be one of the fastest growing technologies. With this guide, I hope to help you learn about containers and why they are so important to modern infrastructure.

Recommended Sessions



Build your microservices application on AWS Fargate ›

This workshop uses a cool tool called mu to build a full CodePipeline that deploys a few demo applications to ECS. mu is written in go by Casey Lee and Paul Duvall, who are also AWS Heroes.


Improving observability of your containers ›

This talk is focused on monitoring and observability. It seems more focused on AWS specific observability tools. It mentions CloudWatch Container Insights, which recently went GA and may remove the need to build most host and container-level monitoring homegrown tools.


CI/CD pipeline integration using AWS native tools ›

Have seen many approaches that people have taken to building and deploying Docker containers and getting them deployed onto ECS. This session focuses on the native AWS tools. It's helpful to see how it can be done with only AWS tools. I like to compare the tools and see when they make sense to use vs other tools.


Container logging: Best practices for consistency ›

This session mentions how to get logs when you are running both ECS and EKS. Sounds like this is more than just centralizing logs to CloudWatch. There’s another log storage solution involved here. If you are using both ECS and EKS, it would be useful to standardize how to view logs.


Chaos Experiments for Container Applications with Service Mesh ›

This chalk talk looks interesting. It'll talk about injecting failures into your containers, introducing container faults, http delays, http abort faults, request timeouts, and more. It discusses using a service mesh as a part of it. I'm wondering if the service mesh is App Mesh in this case. Am interested in seeing the approach.


Deep dive into firecracker-containerd ›

Firecracker is AWS microVM technology. Lambda runs on top of firecracker, but I am interested to know more about about firecracker-containers. It sounds like it will make containers portable on top of the firecracker microVM.

AWS App Mesh


Getting started on AWS App Mesh ›

AWS App Mesh is the AWS way to achieve things like canary deployments. At my previous company, we tried a few different approaches for canary deployments—also known as percentage deployments or blue-green deployments. We used Amazon Route53, middleware, and even HAProxy to control the switch. App Mesh aims to generalize that and make it simpler. In this session, you’ll likely also get an introduction to the terms service mesh, virtual service, virtual node, virtual router, and routes.


Using AWS App Mesh to monitor and control your first containerized app ›

I have been fortunate enough to be at Tony Pujals’s workshops at the AWS SF Loft. He’s got a positive, natural energy that makes his workshops even more enjoyable. He does an excellent job of introducing things for newer folks and he’s been working with AWS AppMesh since it was in beta. This session is worth your time if you want to a gentle introduction to AppMesh.


AWS App Mesh integration with an Amazon ECS microservice application ›

This session specifically runs through a demo of getting AWS App Mesh to work with Amazon ECS. So, if you’re already using Amazon ECS and are considering App Mesh for things like canary deployments, this session is the place to be for an excellent introduction to it.


AWS App Mesh under the hood ›

I find that understanding a little more about how things work under the hood makes engineers better. This session covers some details on how App Mesh works, including scaling, debugging, and logging. Learning how to do this directly from the engineers who built App Mesh debug could prove extremely valuable.


AWS App Mesh deep diveChalk Talk ›

This is for folks who are already familiar with the concept of a service mesh. Check out this chalk talk to learn how to tune AWS App Mesh route settings.



Getting started with Kubernetes on AWS ›

At re:Invent there is an Amazon EKS three-part workshop; this is the first one of the series. It will get you a basic application running on Amazon EKS. The second workshop is CON205-RDeploying applications using Amazon EKS, which covers deployment pipelines, code updates, rollbacks, and load balancers. The third workshop, CON206-RManagement and operations for Amazon EKS, covers logging, monitoring, service mesh, and more. If you get the chance to attend all three workshops, you’ll get a pretty comprehensive view of Amazon EKS.


Deep dive: Observability of Kubernetes applications ›

The rate of change is so fast for many companies, we sometimes don’t spend enough time on important things like monitoring. It’s like buying a car without a speedometer—it’ll be tricky to tell how slow or fast you’re going. Monitoring is key to understanding your application. In a containerized world, monitoring is essential. This talk covers some projects and tools that will be helpful for monitoring.


Live debug microservice running on Kubernetes cluster ›

This Amazon EKS session shows off a cool technique called “pod swapping.” Essentially, you can start a container locally on your machine and swap it with a live Amazon EKS pod. I feel like this feature is easy to get hooked on because it’s so useful.

Amazon ECS


Mythical Mysfits: Build a modern web application on AWS ›

This is an excellent gentle workshop introduction for deploying an application to Amazon ECS. Interestingly, the application uses all serverless components. It uses AWS Fargate, Amazon Simple Storage Service (Amazon S3), and Amazon DynamoDB. Be sure to bring your laptop, and have an AWS account and a GitHub account ready.


Successful deployment of CI/CD using containers and AWS CDK ›

This session focuses on using AWS Cloud Development Kit (AWS CDK) to deploy to Amazon ECS using a CI/CD pipeline. AWS CDK can handle building your Docker image, creating your Amazon ECS task definition, and deploying the Amazon ECS service with only a few lines of code. AWS CDK is another interesting approach that may become more common since it is procedurally based, which software developers are acclimated to.


Infrastructure as code for containers ›

This session covers three different tools to codify and deploy your application to Amazon ECS: AWS CDK, AWS CloudFormation, and Terraform. If you’re researching the different approaches and trying to figure out which one is right for you, this session could be a good resource.

AWS Container Heroes dev chats


From monolith to cloud native in 30 minutes or less ›

Do you have to deal with a mega-monolith? Casey Lee will show you how migrate a monolith by containerizing it and running it on Amazon EKS. He’ll also discuss setting up a continuous-delivery pipeline. He knows a thing or two about AWS CodePipeline. As I mentioned earlier, he’s one of the authors of the mu CodePipeline tool.


Getting started with Amazon EKS ›

Vicky will introduce Amazon Elastic Kubernetes Service (Amazon EKS). She knows a thing or two about teaching, as she’s a computer science professor at Santa Monica College in Los Angeles. She trains many students and runs AWS workshops. This session will be a helpful introduction to Amazon EKS.


Highly available Amazon ECS spot architecture ›

This last session will be presented by yours truly. I’ll talk about how, with the right components, you can run a highly available architecture with spot instances and save 50% to 90% on Amazon Elastic Compute Cloud (Amazon EC2). We’ll cover Amazon Elastic Container Service (Amazon ECS), Spot Instances, Spot Fleet, and Amazon EC2 Auto Scaling.


I’ve found that just physically being at re:Invent, you will learn through osmosis. So, while the conference is about learning, don’t forget to have some fun and meet people. With so many attendees, it may feel overwhelming to meet new folks. I recommend looking up the after parties, reaching out to people, and offering to meet them there to hang out. AWS Heroes will also be around and usually make a point to meet people at re:Invent. If you want to connect with me, I’d be happy to meet up. You can find me on Twitter @tongueroo. I will also be giving a talk on highly available spot architectures with Amazon ECS at the Dev Lounge. Hope to see you there!

Explore other AWS Hero Guides ›

Registration now open

Register now