Design for GitOps: What have we learned from Everything-as-Code at scale?

Learn about Everything-as-Code approach from our experience to enable effective and successful collaboration between hundreds of engineers by GitOps approach. In this session, I on-board you to “as-code” and GitOps practices and I share how have we married them to build “infrastructure-“, “observability-” and “permissions-“ as-code services. I will give you tips on how the system should be designed to help you start with GitOps quickly.

What’s Everything-as-Code?

Everything as Code is the practice of treating all parts of the system as code. This means, storing properties of your system along with its source code in a repository such as git. When we additionally apply GitOps practices – we get a single-click automation backed by review, dry-run workflow and auditing for free!

How have we applied GitOps in our company?

In Dynatrace, more than 1000 devs across multiple locations and timezones collaborate every day. Automation is key to support fast-paced innovation. We’ve started from a hackathon to build first version of project Monaco (aka MONitoring As COde) – where we essentially automate observability as code.  Then, we realized we can apply same approach to any area – for instance by “permissions-as-code”, anyone and anytime can elevate their permission to access new services by submitting just a pull request. 

How can I start with GitOps?

The most important enabler for GitOps is proper RESTful API. We’ll go through most important guidelines on how API should be structured. Then, I will teach you how a basic GitOps workflow looks like. Finally, based on open-source project Monaco, we’ll learn that applying a state from the repository can be easily achievable. 

I’m sure that after this session many of you will be inspired that your team and maybe even users of your platform can unleash unprecedented power of automation and effectiveness.

Scheduled on Thursday from 13:30 to 14:20 in Room 3