AWS ECS CI/CD Demo with Code*

git clone <this repo>
cp * <your dir>
git push <your repo>

$(aws ecr get-login --no-include-email --region us-west-2)

docker build -t 211752457047.dkr.ecr.us-west-2.amazonaws.com/apidemo80:1 .
docker push 211752457047.dkr.ecr.us-west-2.amazonaws.com/apidemo80:1

In Add container step, make sure you specify the image & image tag, NOT ':latest', it Named your container, the name will be used in CodeBuild

Choose either Fargate or EC2

Before setup ECS service, create an ALB & target group without target group targets Setup a proper health check intervals and other timing parameters, otherwise the service update process could be very long.

Specify Task definition defined in previous step For Fargate, dynamic port mapping for ALB is not available, container port & service port should be the same, make sure you set the correct container port in DockerFIle.

curl <ALB entry>

#update api.py to another version
#build a new images with new tag
docker build -t 211752457047.dkr.ecr.us-west-2.amazonaws.com/apidemo80:2 .
docker push 211752457047.dkr.ecr.us-west-2.amazonaws.com/apidemo80:2

update task definition with new container definition click update service and point to another task definition

Choose github and follow the instructions

Choose CodeBuild

For buildspec file, set use the buildspec.yml in the repo root dir. Pay attention to the variables in buildspec.yml Configure Advanced options Setup environment variables AWS_ACCOUNT_ID e.g. 211752457047 AWS_DEFAULT_REGION e.g. us-west-2 IMAGE_REPO_NAME e.g. apidemo80 CONTAINER_NAME e.g. apidemo80(you defined it in Task definition)

By CodeBuild create a role for you automatically, however this role does not have permission to access ECR. Add AmazonEC2ContainerRegistryFullAccess policy to this role manually.

Choose Amazon ECS For image file, use imagedefinitions.json, it is defined in the buildspec.yml, in this line, printf '[{"name":"$CONTAINER_NAME","imageUri":"%s"}]' $REPOSITORY_URI:$IMAGE_TAG > imagedefinitions.json

It would run the pipeline automatically right after you create the pipeline. If encounter any error, refer to the error message and details for troubleshooting. After that, update the version attribute in api.py then push