Creating the pipeline

Getting Started

Steps to Create the Pipeline

First, let’s make sure we are working in the correct directory:

cd ~/environment/aws-atlassian-connect/content/60_pipeline/

Next, from your Cloud9 IDE, let’s launch the CloudFormation template:

aws cloudformation create-stack \
  --stack-name atlassian-connect-cicd --template-body=file://pipeline.yaml \
   --parameters ParameterKey=ProjectName,ParameterValue=atlassian-connect \
  --capabilities CAPABILITY_IAM

Wait for the CloudFormation template to successfully deploy.

until [[ `aws cloudformation describe-stacks \
  --stack-name "cicd" --query "Stacks[0].[StackStatus]" \
  --output text` == "CREATE_COMPLETE" ]]; \
  do  echo "The stack is NOT in a state of CREATE_COMPLETE at `date`"; \
  sleep 30; done && echo "The Stack is built at `date` - Please proceed"

Obtain the clone URL for our repository. We will pass the output of our CLI query to an ENVIRONMENT VARIABLE for later use.

CloneURL=$(aws codecommit get-repository \
  --repository-name atlassian-connect \
  --query 'repositoryMetadata.cloneUrlHttp' \

Cloning your CodeCommit Repo

When your Cloud9 environment was provisioned, it should also have automatically cloned the GitHub repository for this lab. We will want to push the contents of that repo to our CodeCommit repo. From your Cloud9 IDE change directories to your locally clone copy of the lab content:

cd ~/environment/atlassian-connect/

Before we are able to clone our CodeCommit repo, we will need to configure our credentials allow HTTP:

git config --global credential.helper '!aws codecommit credential-helper $@'
git config --global credential.UseHttpPath true

Next, we will add CodeCommit as a remote repo.

git remote add codecommit $CloneURL

We are almost ready to commit. Let’s configure git with our email and name:

git config --global " your_email "
git config --global " your_name "

Now we are finally ready for our initial commit. Type the following:

git status
git add .
git commit -am "initial"
git push --force-with-lease codecommit master

This should create a branch called master in your CodeCommit repo and push the contents of our lab. You should see the following results in your Cloud9 console:

Counting objects: 457, done.
Compressing objects: 100% (283/283), done.
Writing objects: 100% (457/457), 11.56 MiB | 15.70 MiB/s, done.
Total 457 (delta 153), reused 457 (delta 153)
 * [new branch]      master -> master

You can also see the same results by navigating to the CodeCommit console where you will find results similar to these:


Build Process

Remember that our pipeline has been configured to watch for any changes to CodeCommit. When a change is detected it will trigger the pipeline and the build process will commence.

You can also trigger the process by clicking the Release change button from the CodePipeline console


Once triggered, you should see the various stages go through the workflow from the CodePipeline console. For example:


You can also view additional details for the build process by navigating to the CodeBuild console where you will find messages for the various stages defined.


A complete log of the events is also detailed for you in this console.


Deploy Process

The final stage in our pipeline is to deploy the new docker image into our Fargate cluster. As part of the process, an imagedefinitions.json file is generated which contains the path to the newly created docker image(s) stored in ECR.

This file will then be used in Fargate’s task definition to pull the latest image containing your recent code changes.

So, now you should be able to confirm your containers are running by navigating to your Fargate console and see the following: