Skip to content

Latest commit

 

History

History
39 lines (29 loc) · 2.09 KB

aws.md

File metadata and controls

39 lines (29 loc) · 2.09 KB
Raw
meta
name content
description
Fairwinds rok8s Scripts | Deploying to AWS

Deploying to AWS

It's quite straightforward to push Docker images to AWS ECR and deploy to Kubernetes clusters running AWS IAM Authenticator with rok8s-scripts.

AWS Credentials

rok8s-scripts uses the AWS CLI and expects it to have appropriate credentials in place. One of the simplest ways to do that involves setting the following environment variables:

  • AWS_DEFAULT_REGION
  • AWS_ACCESS_KEY_ID
  • AWS_SECRET_ACCESS_KEY

More information about configuring the AWS CLI is available from the official AWS documentation. It's important to note that sensitive credentials like AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY should never be checked into your codebase. Instead, most CI systems, including CircleCI, GitLab CI, and Bitbucket Pipelines, include a way of configuring sensitive credentials for each project.

AWS ECR Docker Auth

With AWS credentials properly configured, you can run the following script to get Docker credentials for pulling from and pushing to an ECR repo.

prepare-awscli

AWS IAM Authenticator

The standard authentication mechanism for Kubernetes clusters running on AWS has quickly become aws-iam-authenticator. The required client binary is included as part of all rok8s-scripts CI Images.

Kubernetes Configuration

Connecting to Kubernetes clusters on AWS works the same as it would anywhere else. A valid kubeconfig file is needed in a base64 encoded format. That can be accomplished by a command like this:

cat valid-kube-config-file.json | base64

That base64 encoded value is expected to be in a KUBECONFIG_DATA environment variable. This value usually contains sensitive information and therefore should be kept out of your codebase and instead loaded as a protected variable in your CI platform. With this value set, the following command will set up your Kubernetes configuration for your pipeline:

prepare-kubectl