Chapter 16 - Serverless Overview

Question 1

Containers

Answer 1

Immutable resources

• I fill the container up with all of my code, all of my dependencies, all of my packages, and then I can pick that container up and move it anywhere that I want.
• focus on using containers with microservices, with small, flexible applications.
• If it’s talking about that monolithic architecture, where we just need one giant server to run everything, then you probably want to avoid containers for that application.
• If possible, it’s best to use a managed architecture.
• Containers are flexible
• We can put anything we want or just about anything that we want into a container
• Think of a container like those microservices. They’re small, they’re quick, they’re fast, they’re portable, they’re immutable.
• Generally we should favor using a container over just deploying our application straight to EC2 instance

Question 2

Lambda

Answer 2

• Lambda loves roles
• Anytime you see a situation in which you’re creating a Lambda function, ensure that you are selecting the appropriate role to attach to it.
• We always, always, always want to attach a role to Lambda to make our permissions just a little bit easier to manage.

Question 3

Kubernetes

Answer 3

• Amazon Elastic Kubernetes Service (Amazon EKS) gives you the flexibility to start, run, and scale Kubernetes applications in the AWS Cloud or on-premises.
• Amazon EKS helps you provide highly available and secure clusters and automates key tasks such as patching, node provisioning, and updates.
• Customers such as Intel, Snap, Intuit, GoDaddy, and Autodesk trust EKS to run their most sensitive and mission-critical applications.
• EKS runs upstream Kubernetes and is certified Kubernetes conformant for a predictable experience.
• You can easily migrate any standard Kubernetes application to EKS without needing to refactor your code.
• EKS makes it easy to standardize operations across every environment.
• You can run fully managed EKS clusters on AWS.
• You can have an open source, proven distribution of Kubernetes wherever you want for consistent operations with Amazon EKS Distro.
• You can host and operate your Kubernetes clusters on-premises and at the edge with AWS Outposts and AWS Wavelength, and have a consistent cluster management experience with Amazon EKS Anywhere (coming in 2021).

Question 4

Fargate

Answer 4

• Allows me to standardize with containerization.
• If I create a container here in AWS, I can run it in another cloud provider.
• I can run it on-premise.
• I can run it anywhere I want.
• It allows me to have more consistent workloads versus Lambda
• Fargate can take that containerized architecture and run it for maybe a little bit longer.
• It doesn’t have that time window limit that Lambda has.

Question 5

When do we use Lambda versus Fargate versus EC2?

Answer 5

Lambda

• Lambda is for light weight functions.
• They can run very quickly and they generally
• Need to be easily integrated into our AWS architecture.

Fargate

• Is for when I have containers that don’t need to run all the time
• Fargate is a serverless tool.
• Allows us to run containers without hosts
• Doesn’t work by itself
• Think about that use case of Fargate of running containers without servers and the ECS or EKS portion is not applicable to what the question is getting at.

EC2

• Is for when I have containers that need to run 24/7 and I’m really concerned about cost.

We want to use EC2 when we think about costs and long running content and we want to think about Fargate
when we just want to run that container for a little bit of time and ease of use is the most important piece

When it comes to Fargate versus Lambda, Fargate excels for containers that need to run for a bit longer but we still don’t want servers and Lambda excels at that short and simple code.