It has been replaced by https://github.com/redhat-cop/casl-ansible/tree/master/roles/openshift-applier and https://github.com/rht-labs/labs-ci-cd
This repository is home to the Automation API specification for Open Innovation Labs Push Button Infrastructure. Originally, it started as the home of all our Swagger / OpenAPI Specifications, System Designs and Documentation, but over time we've decided that we should have smaller, more discrete repos. The repository needs to be migrated to a new one to make the name more declarative.
See the contributor's guide. The following documentation is focused on end users.
Let's suppose for a moment that you're working on a team that is building a Java web application that you want to deploy to JBoss EAP on OpenShift. In your organization it's standard policy for teams to use a promotion process that deploys new releases of the application from a development environment (DEV), to a testing environment (UAT), and finally to production (PROD).
If you would like to read the schema for the engagement before following the example, now is the time to do it. You can use tools like swagger editor to visualize the schema as well.
The Automation API is currently composed of a single object hierarchy, the engagement. We're going to focus on the parts of the engagement that define OpenShift resources, like projects and applications inside an OpenShift cluster. Through out each section of the example, we'll show only the new JSON needed to add the functionality described. We'll put the whole document together at the end.
Most Automation API documents will start like this:
{
"openshift_clusters": [
{
"openshift_resources": {
"projects": []
}
}
]
}PBI leverages OpenShift S2I for all container builds. Thus, we need to tell our automation to set up an OpenShift project where these S2I builds can take place. Given that our organization has a policy for DEV => UAT => PROD promotion process, we'll use the DEV stage to build our images. Here is how to do that:
{
"projects": [
{
"name": "hello-world-dev",
"display_name": "Hello World - DEV",
"environment_type": "build",
"apps":[]
}
]
}PBI leverages Jenkins Pipeline to do application builds. While we can use OpenShift S2I to do application builds, Jenkins provides an ecosystem of 3rd party plugins for things like static analysis and artifact archiving that are very valuable. Therefore, we need to give Jenkins the information it needs to properly build our Java Application with maven. OpenShift will use the resulting WAR file from the Jenkins application build to create a container image with S2I binary source builds, so we need to tell OpenShift to use a JBoss EAP base container image.
{
"apps":[
{
"name": "hello-world",
"scm_url": "https://github.com/jboss-developer/jboss-eap-quickstarts.git",
"scm_ref": "master",
"build_tool": "mvn-3",
"build_application_commands": [
"mvn clean install"
],
"context_dir": "helloworld",
"base_image": "openshift/jboss-eap64-openshift"
}
]
}In the previous steps we've declared our DEV stage with the details of how to build our application and it's corresponding container image. We now need a way to promote this application to our UAT and PROD stages. PBI leverages Jenkins pipeline and OpenShift ImageStreams to provide this functionality. Therefore, we simply need to tell OpenShift which projects to create for our UAT and PROD stages, as well as the apps we expect in those project, and PBI will be smart enough to create pipelines in Jenkins to promote the application through those stages. Here is what that looks like:
{
"projects":[
{
"name": "hello-world-uat",
"display_name": "Hello World - UAT",
"environment_type": "promotion",
"apps": [
{
"name": "hello-world",
"base_image": "hello-world"
}
]
},
{
"name": "hello-world-prod",
"display_name": "Hello World - PROD",
"environment_type": "promotion",
"apps": [
{
"name": "hello-world",
"base_image": "hello-world"
}
]
}
]
}TODO
If we put it all together, here is the result.
{
"openshift_clusters": [
{
"openshift_resources": {
"projects": [
{
"name": "hello-world-dev",
"display_name": "Hello World - DEV",
"environment_type": "build",
"apps":[
{
"name": "hello-world",
"scm_url": "https://github.com/jboss-developer/jboss-eap-quickstarts.git",
"scm_ref": "master",
"build_tool": "mvn-3",
"build_application_commands": [
"mvn clean install"
],
"context_dir": "helloworld",
"base_image": "openshift/jboss-eap64-openshift"
}
]
},
{
"name": "hello-world-uat",
"display_name": "Hello World - UAT",
"environment_type": "promotion",
"apps": [
{
"name": "hello-world",
"base_image": "hello-world"
}
]
},
{
"name": "hello-world-prod",
"display_name": "Hello World - PROD",
"environment_type": "promotion",
"apps": [
{
"name": "hello-world",
"base_image": "hello-world"
}
]
}
]
}
}
]
}The examples directory contains some example files that can be used with the ansible-stacks implementation, and in particular the tests, found here.
The Automation API is designed to support multiple apps in a project. Remember that projects use kubernetes to control resource quotas, so you'll want to work with your operations team to figure out the right balance of projects and apps.
The Automation API supports a lot more than what was covered here. Review the schema to get an idea what those things are, and open an issue if you'd like more documentation.
See https://github.com/rht-labs/api-design/issues/53
TODO when console is ready