Why GitOps Exists (And Why CI/CD Alone Wasn't Enough)

Why It Matters

GitOps bridges the gap between declared intent and runtime reality, cutting outages and operational overhead for cloud‑native enterprises.

Key Takeaways

• Traditional CD deploys but can't guarantee cluster state matches Git.
• GitOps uses pull‑based agents like Argo CD to detect and correct drift.
• Drift visibility enables immediate alerts or automatic remediation of manual changes.
• Rollbacks become simple git revert operations without extra pipeline steps.
• Pull loop must own the cluster; mixing push pipelines reintroduces drift.

Summary

The video asks why GitOps emerged despite the existence of continuous deployment pipelines. Traditional CD tools such as Jenkins or GitHub Actions push manifests to Kubernetes, but they leave no mechanism to verify that the live cluster still mirrors the source‑of‑truth in Git, allowing undocumented, manual changes to creep in.

The presenter explains that GitOps flips the model: an in‑cluster agent (e.g., Argo CD or Flux) continuously pulls the desired state from Git and compares it to the actual state. When drift is detected, the agent either alerts operators or automatically reconciles the difference, turning drift into a visible, actionable event. This pull‑loop makes the repository the single source of truth and eliminates the need for separate rollback pipelines—simply reverting a commit restores the previous configuration.

Concrete examples illustrate the problem: a teammate SSHs into production at 2 a.m., edits a deployment, and the cluster diverges from Git. The agent would flag the change or revert it. Likewise, a buggy image tag merged into Git can be undone with a git revert, and the agent will redeploy the corrected version without additional tooling.

The implication is clear: organizations that let the pull‑based GitOps controller fully own the cluster gain continuous observability, faster remediation, and reduced operational risk. Mixing push‑based CD with GitOps defeats the purpose, as drift reappears, so teams must commit to the pull‑loop for true state fidelity.

Original Description

This is one of the best DevOps interview questions out there: "Why does GitOps even exist if we already had CD?"

Most people get it wrong because they think GitOps is just about using Git. It's not. It's about the pull loop — and letting the agent own the cluster.

In 60 seconds, we cover cluster drift, why Jenkins-based pipelines can't catch it, and how Argo CD and Flux solve it for real.

Perfect if you're studying for a DevOps or Kubernetes interview.

#GitOps #Kubernetes #DevOps #ArgoCD #CICD #CloudNative #DevOpsTips #K8s