Showcase OpenShift GitOps

[lijcam]

Develop a track to showcase OpenShift GitOps (v. 4.10). The narrative should focus on a project/application teams, persona: Developers, Operations. Using Argo to deploy a small (Pacman), and moderately complex application (Coffee Shop).

Concerns, problems, usecases

• How do I reduce the risk "opportunistic administration" going unrecorded/undocumented?
• Infrastructure as code is cool but, how do I know what's in my cluster reflects truth?
• What is an effective merge/pr process for application?
• Should my Infrastructure as code, ie deployments, live in the same or different repo?

[mattparko]

Ideas to leave behind:

• What does Infrastructure as Code look like?
• Potential approach for moving through lifecycle environments (git branching, tagging - git / images / etc)

First session:

• Slides: Intro to infra as code, GitOps in general, ArgoCD
• Lab: Manually make a common change in a deployment. It disappears - what happened? Was it logged/reported? Explore an ArgoCD application. Make a change via git that is synced into dev.
• Group discussion (discuss logs / alerts / audit)
• Slides: What does good like like? (SDLC, ....)
• Lab: SDLC. Change has already been made and pushed in dev. Pull into prod in git (assuming a different persona - like release mgmt)
• Group discussion and wrap up

Second session:

• Use Coffee Shop and dive into creation of an ArgoCD application, via either the operator or the console
• Deeper dive into YAML and git pull/push