Code Stream is now available with the 8 release. You have now the choice of on-premises or as a service in The Cloud. This is an excellent solution to deploy your business app with pipelines, simplifying deployment and avoid error-prone.

I share with you as a glimpse how I set a use case for one on my customer. The main purpose is to avoid the (very) opinionated app deployment in Openshift (3.11).  This is why I used Openshift as a standard kubernetes cluster but orchestrate container image build and deployment in kubernetes by Code Stream.

Endpoints creation.

After endpoints creation, I'm able to consume them. So, I have now kubernetes endpoint, bitbucket and gitlab endpoints for code repo and one agent acting as proxy for my cPod running in OVH Cloud. In this context, I'm using Code Stream from Cloud Services.

Code Stream offerts to automaticaly create webhook in code repo. A webhook offers the possobility to trigger pipeline when an event occurs. This is the famouse "git push experience". Shortly, when a dev executes "git push" againt Bitbucket the pipeline is automatically triggered (executed).

Creation of webhook for Bitbucket

Files of my app is pushed to Bitbucket, I always write bash scripts that executes deployment without any additionnal tools. A good best practice to avoid lock-in. If  I need to adapt to another CD solutions I have to just reuse these scripts.

Repo in Bitbucket

Last step is to define the pipeline with all conditions, roll-back and all good SRE practices. I like to variablize everything for modularity and reusability.

Pipeline Designer in Code Stream

I let you to discover how to map variables between Code Stream and yaml files.

$${namespace} in yaml will be updated by Code Stream during pipeline execution.
$${namespace} variable binding in Deployment task.
namespace as pipeline input parameter.

Pods in Kubernetes means container image build means enterprise registry! This why I also integrate this pipeline with Harbor running in PhtonOS 3.0. Both are secure and fast to deploy.

Harbor as Regisry for Openshift scanning container image after push.

Finally everything is ready to be fired. Code Stream has also a nice Dashboards feature.

Pipeline execution and status
Pipeline standard dashboard, custom dashboards are also available.

Final Thouthgs: Code Stream is very simple to use, powerful with lot of endpoints and secure with password encryption. It's worth it to test it. Without any doubt, you will do incredible pipelines in few minutes.