SI2CI stands for Spring Initializr to Concourse CI.

This repo is home to Demo grade code that glues Spring Initializr with Concourse. The idea was to showcase a complete onboarding experience for an Enterprise Developer by bridging Spring Initializr portal with workflow that still needs to happen. This workflow is driven by Concourse to take actions such as; Create a new git Repostiory, Check in Mock code and Enterprise templates, Build a CI/CD Delivery pipeline ...

Below is the folder structure of the Repository and its contents

• infra : Terraform specs to setup concourse on GCP. Currently only setups a cloud machine on GCP
• pipeline : Sample Concourse pipeline that can get executed via Spring Initialzr
• templates : Template folder gets coppied into GitHub Repo that's created
• pkg/concourse : Concourse wrapper in use by the project
• vendor : Go dependencies

Spring Initialzr screen has a form that will perform a GET action against its /starter.zip endpoint based on the user input.

Since Pivotal Cloud Foundry user owns the control over the routing stack a second application can "hijack" the ./starter.zip endpoint to execute a custom workflow. SI2CI is intended to sit behind this endpoint and capture the parameteres that would have otherwise generate a zip file.

SI2CI is developed to catch Spring Initializr form passed parameters:

• SI2CI_NAME
• SI2CI_LANGUAGE
• SI2CI_BOOTVERSION
• SI2CI_BASEDIR
• SI2CI_GROUPID
• SI2CI_ARTIFACTID
• SI2CI_DESCRIPTION
• SI2CI_PACKAGENAME
• SI2CI_PACKAGING
• SI2CI_JAVAVERSION

and append them to Concourse Pipeline variables/credentials file, so that user input from Spring Initialzr can be utilized during the pipeline execution. In sample workflow provided in the repo SI2CI_NAME field which is generated from Spring Project name is utilized as; Appname in Pivotal Application Services, Repo name on Github etc.

At a high level there are few steps involved to get the environment up:

1. Bring up Spring Initializr
2. Create a github account for provisioning
3. Deploy SI2CI
4. Modify Pipeline and Credentials
5. Provide SI2CI runtime parameters

The official Spring Initializr source code is location in; https://github.com/spring-io/initializr . README documentation in the repository provides details for deploying Spring Initializr in your own infrastructure. At a high level;

Clone the repository and navigate into the folder

git clone https://github.com/spring-io/initializr.git
cd initializr

Build the Spring Initialzr project

./mvnw clean install

Create a new manifest.yml file to push Spring Initialzr to Pivotal Cloud Foundry

---
applications:
- name: initializr
  memory: 2048M
  instances: 1
  path: initializr-service/target/initializr-service.jar

Push the application to Pivotal Cloud Foundry

cf push

Validate the Spring Initialzr is working as expected.

While this step is not strictly required for SI2CI to function, the default Pipeline located in ./pipeline folder of this repository assumes the first step of the pipeline is to create a GitHub Repository. In order to achive this goal, a small utility is used to programatically create a github.com repo given that a GITHUB_TOKEN environmental variable is provided.

Recommendation is to create a new github user such as "BotOfAleks" and follow the instructions to generate a token for GitHub API access. Make sure that the token has access to repo scope.

SI2CI is a golang application with a sample manifest.yml located at the root of this repository. Target the same Org and Space utilized by Initialzr, then cf push the application. Pivotal Cloud Foundry will build the application.

NOTE THAT application manifest.yml has the no-route parameter set to true. As a result no route is assigned to the application.

You should end up with a similar setup to screenshot below

Map the starter.zip endpoint of the Spring Initialzr app onto si2ci app

cf map-route si2ci cfapps.io --hostname demoinitializr --path starter.zip

Validate the setup

cf apps
Getting apps in org fakeOrg / space initalizr as fakeUser@pivotal.io...
OK

name             requested state   instances   memory   disk   urls
demoinitializr   started           1/1         2G       1G     demoinitializr.cfapps.io
si2ci            started           1/1         512M     1G     demoinitializr.cfapps.io/starter.zip

Create a credentials file out of sample file.

mv ./pipeline/credentials.yml.sample ./pipeline/credentials.yml

Note that credentials.yml is already part of gitignore file.

Modify the ./pipeline/pipeline.yml and ./pipeline/credentials.yml as you see fit. These files will be used by SI2CI to create the initial pipeline

Runtime Environmental Variables:

• CONCOURSE_USERNAME: Concourse Username of the user that pushes the pipelines via this code
• CONCOURSE_PASSWORD: Password for the Concourse User
• CONCOURSE_TARGET: Target conconcourse environment URL, ie: https://concourse.company.com
• PIPELINE_YML: Contents of pipeline.yml . This can as an be generated by; export PIPELINE_YML=$(cat ./pipeline/pipeline.yml)
• CREDENTIALS_YML: Contents of pipeline.yml . This can as an be generated by; export CREDENTIALS_YML=$(cat ./pipeline/creds.yml)

Modify the application runtime environmental variables

export APPNAME=si2ci
export CONCOURSE_USERNAME="concourseadmin"
export CONCOURSE_PASSWORD="Password1!"
export CONCOURSE_TARGET="https://concourse.company.com"
export PIPELINE_YML=$(cat ./pipeline/pipeline.yml)
export CREDENTIALS_YML=$(cat ./pipeline/creds.yml)
cf set-env $APPNAME CONCOURSE_USERNAME "$CONCOURSE_USERNAME"
cf set-env $APPNAME CONCOURSE_PASSWORD "$CONCOURSE_PASSWORD"
cf set-env $APPNAME CONCOURSE_TARGET "$CONCOURSE_TARGET"
cf set-env $APPNAME PIPELINE_YML "$PIPELINE_YML"
cf set-env $APPNAME CREDENTIALS_YML "$CREDENTIALS_YML"

Restage the application

cf restage $APPNAME