GitOps is not enough: taming environment promotion in Kubernetes with Kargo

You adopted GitOps.
Your deployments are declarative, auditable and reproducible. And yet promoting a release from dev to staging to production is still a mess of shell scripts, manual PRs and Slack messages.

Sound familiar?

GitOps solves the "how do I deploy" problem brilliantly. What it does not solve is the "how do I safely move an artefact across environments" problem and that gap is where most teams quietly give up and reintroduce the manual steps they were trying to eliminate.

In this talk we will look at why environment promotion is genuinely hard in a GitOps world, and then explore the tools that are finally starting to solve it properly.

We will start with the patterns: what does a healthy promotion pipeline actually look like? Single Git repo vs multi-repo, image tags vs digests, who owns the promotion decision and how do you enforce gates (tests, approvals, metrics thresholds)?

Then we will introduce Kargo (by Akuity, the team behind Argo CD), the first tool designed specifically for Kubernetes-native environment promotion.

Live demo: a full promotion pipeline, image built by CI, picked up by Kargo, promoted through three environments with automated verification and a manual approval gate before production.

We will also briefly compare the Kargo approach against Argo CD ApplicationSets and Flux Promoter, so you can make an informed choice for your own setup.