BUILDING.md

Document outlining the new build process for the analyzere-python repository

Dockerfile.base contains the base level dependencies for running the tests against the python client.

Dockerfile is the image that inherits from the base image and runs the actual tests, so it doesn't have to rebuild the image each time.

For creating a new baseline image, you will edit the Dockerfile.base, when you are happy with your changes, you can then retag it as the latest and push it to ECR. In this example we will just tag it latest, but you could push a versioned tag up to ECR if you like.

make your changes

vi Dockerfile.base

login to AWS

$(aws ecr get-login --no-include-email --region us-east-1)

build the image

docker build -t analyzere-python-base:latest -f Dockerfile.base .

tag the image

docker tag analyzere-python-base:latest 612007926530.dkr.ecr.us-east-1.amazonaws.com/analyzere-python-base:latest

push the image to ECR

docker push 612007926530.dkr.ecr.us-east-1.amazonaws.com/analyzere-python-base:latest

Now you can use the regular Dockerfile, which will use your new base when you execute it, without having to rebuild the prior layers.

Alternatively you can use build/rebuild_base to automatically build / tag / push your new base image changes