travis-and-docker.md

Working with travis & docker

Travis is a CI/CD tool that works with docker. Any time changes are being done, travis detects them, runs test and pushes them to cloud (AWS).

Travis understanding

File: .travis.yml

sudo: required              # sudo access is mostly requried
services: 
    - docker                # installs a copy of docker ready to go.

before_install:
    - docker build -f -t <docker_username>/<cotaniner_name> <Dockerfile.dev> . 
    # runs before deploying, or final processes.
    # builds a container from a Dockerfile

script:                     # commands to run tests/deployment.
    - docker run <container_id> <sample_code>
    # runs docker scripts inside the container.


after_success:
    - docker build -t <docker_id>/<image_name> <path>
    # build container from docker file

    - docker login -u <docker_id>  --password-stdin (take password from STDIN)
    # Log in to docker CLI

    - docker push <docker_id>/<image_name>
    # push images to docker hub
deploy:                     # take stuff and deploy our project
    provider: elasticbeanstalk
    # cloud service provider

    region: 
    # region of the ElasticBeanTalk instance.

    app: 
    # name of the app  <app_name> on the the cloud platform

    env:
    # name of the environment.

    bucket_name:
    # S3 bucket. (Give Elastic BeansTalk access the files from S3)

    bucket_path:
    # path for the project. <app_name> by default.

    on:
    # auto deploy on pushes to what place.
        branch: master 
        # master by default

    access_key_id:
    # ACCESS KEY ID

    secret_access_key:
    # SECRET ACCESS KEY
    

AWS with Travis & Docker

For CI/CD with AWS.

Amazon Elastic Beanstalk handles continues deployment.

For multicontainer projects use:

Dockerrun.aws.json - Container definitions, and build images from docker hub. AWS Task definition reference

Dockerrrun.aws.json:

{
    "AWSEBDockerrunVersion": "(numveric) version of interpretation",
    "containerDefinitions": [
        {
            "name": " name of service, could be anything",
            "image": "<docker_id>/<container_name> // docker image to use",
            "hostname": "// name of host service",
            "essential": "true/false // atleast 1 service in json has to be essential.",
            "portMappings": [
                {
                    "hostPort": "<port_id>",
                    "containerPort": "<port_id>"
                }
            ],
            "links": "allows containers to communicate with each other without need for port mapping.",
            "memory": "numeric // memory allocation for deployment."
        }
    ]
}