A fusion of Makefile and Docker

#perl #docker #devops #tutorial

In our previous article we have created a Dockerfile for our Mojolicious application. In there I mentioned that we have to run several docker command and when we map it across different environment we have to take care of different permutations and combinations.
What if we can automate it more. What if we can abstract the complexity and as a user I just have to run minimum command to get it working.
Here comes the Makefile. We will be using this to make our process more easier. For windows you can use gmake.exe utility instead of make. It comes with Strawberry Perl.
Lets create a Make file in mojo_react_app/Makefile

DOCKER_TAG ?= mojo_react_app:development
NO_CACHE ?= false
PULL ?= false
BUILD_TYPE ?= development
CONTAINER_NAME ?= mojo_react_app
HTTPS_OUT_PORT ?= 443
HTTPS_IN_PORT ?= 443
HTTP_OUT_PORT ?= 80
HTTP_IN_PORT ?= 80
DOCKER_REPO ?= <docker_repo_url>
ADDRESS ?= 0.0.0.0
DOCKER_CFG = <path to docker config on machine>

.DEFAULT_GOAL := help

build:
    # Build the dockerfile
    docker build --pull=${PULL} --no-cache=$(NO_CACHE) -t $(DOCKER_TAG) .

create:
    # Create the container and copy the certificates
    echo 'Creating container for ${ADDRESS}'
    docker create --name ${CONTAINER_NAME}_${BUILD_TYPE} -t \
        -p ${ADDRESS}:$(HTTPS_OUT_PORT):$(HTTPS_IN_PORT) -p ${ADDRESS}:$(HTTP_OUT_PORT):$(HTTP_IN_PORT) --env mode=${BUILD_TYPE} \
        ${DOCKER_TAG}
    docker cp ${DOCKER_CFG}/mojo_react_app/apache-certificate/apache_certificate_${BUILD_TYPE}.crt ${CONTAINER_NAME}_${BUILD_TYPE}:/etc/ssl/certs/server.crt
    docker cp ${DOCKER_CFG}/mojo_react_app/apache-certificate/apache_certificate_${BUILD_TYPE}.key ${CONTAINER_NAME}_${BUILD_TYPE}:/etc/ssl/private/server.key

start:
    # Start a container
    docker start ${CONTAINER_NAME}_${BUILD_TYPE}

run:
    # Create and start the container
    make create -e ADDRESS=${ADDRESS} HTTPS_OUT_PORT=${HTTPS_OUT_PORT} HTTPS_IN_PORT=${HTTPS_IN_PORT} BUILD_TYPE=${BUILD_TYPE} DOCKER_TAG=${DOCKER_TAG}
    make start -e BUILD_TYPE=${BUILD_TYPE}
    make show

stop:
    # Stop a running container
    docker stop ${CONTAINER_NAME}_${BUILD_TYPE};

clean_container:
    # remove previous container
    docker rm -f ${CONTAINER_NAME}_${BUILD_TYPE} 2>/dev/null \
    && echo 'Container for "${CONTAINER_NAME}_${BUILD_TYPE}" removed.' \
    || echo 'Container for "${CONTAINER_NAME}_${BUILD_TYPE}" already removed or not found.'

clean_image:
    # remove created image
    docker rmi ${DOCKER_TAG} 2>/dev/null \
    && echo 'Image(s) for "${DOCKER_TAG}" removed.' \
    || echo 'Image(s) for "${DOCKER_TAG}" already removed or not found.'

dev: 
    make build -e PULL=${PULL} NO_CACHE=${NO_CACHE}
    make clean_container
    make create
    make start
    make show

stag:
    $(eval override BUILD_TYPE=staging)
    $(eval override DOCKER_TAG=${CONTAINER_NAME}:${BUILD_TYPE})
    # This will return only the IP address associated with the domain name ans assign it to ADDRESS
    $(eval override ADDRESS=$(shell dig +short <your staging URL e.g. mojo-react-app-staging.com>))

    make build -e PULL=true NO_CACHE=true DOCKER_TAG=${DOCKER_TAG}
    make clean_container -e BUILD_TYPE=${BUILD_TYPE}
    make create -e ADDRESS=${ADDRESS} BUILD_TYPE=${BUILD_TYPE} DOCKER_TAG=${DOCKER_TAG}
    make start -e BUILD_TYPE=${BUILD_TYPE}
    make show

prod:
    $(eval override BUILD_TYPE=production)
    $(eval override DOCKER_TAG = ${CONTAINER_NAME}:${BUILD_TYPE})
    # This will return only the IP address associated with the domain name ans assign it to ADDRESS
    $(eval override ADDRESS=$(shell dig +short <your production URL e.g. mojo-react-app.com))

    make build -e PULL=true NO_CACHE=true DOCKER_TAG=${DOCKER_TAG}
    make clean_container -e BUILD_TYPE=${BUILD_TYPE}
    make create -e ADDRESS=${ADDRESS} BUILD_TYPE=${BUILD_TYPE} DOCKER_TAG=${DOCKER_TAG}
    make start -e BUILD_TYPE=${BUILD_TYPE}
    make show

show:
    # show running containers
    docker ps | grep ${CONTAINER_NAME}

rebuild:
    # rebuilt the dockerfile
    make clean_container -e BUILD_TYPE=${BUILD_TYPE}
    make build -e PULL=true NO_CACHE=true DOCKER_TAG=${DOCKER_TAG}

up:
    # Run container on port
    make build -e PULL=true NO_CACHE=true DOCKER_TAG=${DOCKER_TAG}
    make run -e HTTPS_OUT_PORT=${HTTPS_OUT_PORT} HTTPS_IN_PORT=${HTTPS_IN_PORT} BUILD_TYPE=${BUILD_TYPE} DOCKER_TAG=${DOCKER_TAG}

login:
    # run as a service and attach to it
    docker exec -it ${CONTAINER_NAME}_${BUILD_TYPE} bash

release:
    make build -e PULL=true NO_CACHE=true DOCKER_TAG=${DOCKER_TAG}
    make push -e VERSION=${VERSION}

push:
    docker push $(DOCKER_REPO)/$(CONTAINER_NAME):$(VERSION)

pull:
    docker pull $(DOCKER_REPO)/$(CONTAINER_NAME):$(VERSION)

# Docker tagging
tag:
    ## Generate container tags for the `{version}`
    @echo 'create tag $(VERSION)'
    docker tag $(DOCKER_TAG) $(DOCKER_REPO)/$(CONTAINER_NAME):$(VERSION)

help:
    @echo ''
    @echo 'Usage: make [TARGET] [EXTRA_ARGUMENTS]'
    @echo 'Targets:'
    @echo '  build      build docker --image--'
    @echo '  rebuild    rebuild docker --image--'
    @echo '  dev        run docker --container-- in development mode => $(DOCKER_TAG)'
    @echo '  stag       run docker --container-- in staging mode'
    @echo '  prod       run docker --container-- in production mode'
    @echo '  login      run as service and login --container--'
    @echo '  clean_image        remove docker --image-- '
    @echo ''
    @echo 'Extra arguments:'
    @echo 'CONTAINER_NAME=:     make clean_container -e CONTAINER_NAME=my_app (no need to provide this param, it will be set by default)'
    @echo 'BUILD_TYPE=:     make clean_container -e CONTAINER_NAME=my_app BUILD_TYPE=staging (whether the build type is 'development', 'staging' or 'production')'
    @echo 'DOCKER_TAG=:     make build -e DOCKER_TAG=my_app:staging'
    @echo 'HTTPS_OUT_PORT=:     make create -e HTTPS_IN_PORT=8080 (port from which the request will come- outside world)'
    @echo 'HTTPS_IN_PORT=:      make create -e HTTPS_IN_PORT=443 (port to which the request will be forwarded)'

I know its overwhelming. Lets go through each one of them.

Initial 12 lines are the default parameters. In case you will not provide any parameter it will take those arguments. By default everything is in development mode. You will have to overwrite it for different modes.

DOCKER_TAG     =>  Name of your app with the env name (e.g. mojo_react_app:development)
NO_CACHE       =>  Whether to start the build from scratch (true|false). Default - false
PULL           =>  Whether to pull a newer version of the image (true|false). Default - false
BUILD_TYPE     =>  Build environment (development|staging|production). Default - development
CONTAINER_NAME =>  Name of the container (e.g. mojo_react_app)
HTTPS_OUT_PORT =>  HTTPS host port visible to outside world while publishing the container(443)
HTTPS_IN_PORT  =>  HTTPS container port on which traffic will be coming inside the container(443)
HTTP_OUT_PORT  =>  HTTP host port visible to outside world while publishing the container(80)
HTTP_IN_PORT   =>  HTTP container port on which traffic will be coming inside the container(80)
DOCKER_REPO    =>  URL of the docker repo where you want to push or pull image.
ADDRESS        =>  IP address of the container from where the outside word can access it. Default - 0.0.0.0
DOCKER_CFG     =?  Path to docker config files on host machine. This will contain the apache certificate and will be private to you.

build will build your dockerfile
create will create the container and copy the apache certificates
start will start the container
run will create and start the container. It will internally call the create and start command.
stop will stop the running container.
clean_container will remove/delete the remove previous container if exist.
clean_image will remove/delete created image if exist.
show will show the info about running container.
rebuild will rebuilt the dockerfile. It will just call clean_container and build the image again.
up will run the container on given port. It will always pull the new image.
login will login to the container. Using this you can see the traffic logs and what is going on inside the container.
release push pull tag - These commands you will be using when you want to create tag and push it to docker repo url(somewhere where you can access it and you don't have to create the image again). This is handful in Jenkins deployments, otherwise in normal dev work its not that important and will not be used much.
help will give you the info about different param and what they do.
Now 3 important commands which we will be using more often is -

    make dev
    make stag
    make prod

make dev

dev is written in such a way that you don't have to worry about your container name and param. Just update the default value with your project specific value one time and your dev environment will be ready in no time.
It will build the container, create it and run it with single command.

make dev

Run this command.
Here are the screenshot at different point of time for this command. For the first time this command will take few minutes but subsequent run will be completed within few seconds.

I have highlighted different section for better understanding.
You can access your app on https://localhost/

make stag

This command you will be running when you want to deploy your application in staging environment.
If you look closely we are overriding the default config for staging in initial few lines.

    $(eval override BUILD_TYPE=staging)
    $(eval override DOCKER_TAG=${CONTAINER_NAME}:${BUILD_TYPE})
    # This will return only the IP address associated with the domain name ans assign it to ADDRESS
    $(eval override ADDRESS=$(shell dig +short <your staging URL e.g. mojo-react-app-staging.com>))

The 4th lines is where you have to make your changes as per your requirement. We are overriding the default value of address(0.0.0.0) with your domain name of staging environment. We are using the dig utility to generate the IP address from that domain name and assigning it to ADDRESS.
Just run

make stag

and you can access your staging environment.
E.g.
https://mojo-react-app-staging.com/
or
https://staging-ip-add/

make prod

This command you will be running when you want to deploy your application in production environment.
Again, you can see we are overriding the default config for production in initial few lines.

    $(eval override BUILD_TYPE=production)
    $(eval override DOCKER_TAG = ${CONTAINER_NAME}:${BUILD_TYPE})
    # This will return only the IP address associated with the domain name ans assign it to ADDRESS
    $(eval override ADDRESS=$(shell dig +short <your production URL e.g. mojo-react-app.com))

Here again in 4th lines is where you have to make your changes as per your requirement. Update it to your prod environment hostname. We are using the dig utility to generate the IP address from that domain name and assigning it to ADDRESS.
Just run

make prod

and you can access your staging environment.
E.g.
https://mojo-react-app.com/
or
https://production-ip-add/

Stop the container

make stop

By default it will stop the dev container. Pass different param for stag or prod env.

Login to container

make login

Again it will login to dev one by default. You can pass param to override that behavior.

Conclusion

This Makefile is just a wrapper around different docker commands.
Especially during development you can enable live reload and your changes in host will be available in docker container also . More info here.
You can use it for any Dockerfile. Here I am using a Mojolicious application. But the same Makefile can be used for a Django application or any other.

I hope this is useful for you and make your life more smoother.
The Makefile is also available on github

Perl onion logo taken from here
Mojolicious logo taken from here
Docker logo taken from here
GNU logo taken from here