Deploying RHEL with Ansible Automation Platform | RHEL Presents Ep. 83

Episode 83 was a full house. My cohost Brian Smith and I brought on three guests, Martin Jackson from the validated patterns team, Paul Armstrong from RHEL Canada, and chief architect Mike Savage, to take all the concepts we talk about in pieces, automation, identical systems, Ansible, and package them into one end-to-end picture. The goal was to actually deploy and test a RHEL workload on Azure live on the show, and this also happened to be my last episode as host as I move over to Red Hat’s cloud partner team.

A few things worth carrying away from the conversation:

• GitOps needs Ansible Automation Platform, not just Ansible on the command line. Martin laid out the four GitOps principles and made the point that plain CLI Ansible can be declarative and versioned but can’t be pulled automatically or continuously reconciled, which is exactly what the platform adds.
• One reusable workflow deploys and tests across any cloud by changing a variable. Paul’s workflow published Satellite content views, spun up ephemeral hosts on Azure, deployed apps, ran tests, and promoted to QA, and Mike stressed the same pipeline runs identically on EC2, VMware, KVM, or bare metal.
• They deliberately broke a build to show why you want to catch failures early. Paul contrived a failed JBoss deployment to demonstrate “celebrating failure” and shifting problems left into dev and QA, tied to the Azure landing zone work Mike described with Microsoft.

If you’ve wanted to see automation, identity, and content management all working together to stand up RHEL in the cloud, this episode ties the whole thing together.