What is GitOps?

Question

What is GitOps?

Accepted Answer

**GitOps** is an approach to continuous deployment (especially for Kubernetes) where **Git is the single source of truth** for the desired state of infrastructure and applications, and an automated agent continuously **reconciles** the actual state to match Git. Changes are made via Git, and the system pulls and applies them.

## The GitOps model

```text
Git holds the DESIRED STATE (declarative configs: Kubernetes manifests, Helm charts, etc.):
  → an AGENT in the cluster continuously COMPARES actual state vs Git (desired state)
  → it RECONCILES: automatically applies changes to make reality match Git
  → to deploy/change: commit to Git → the agent detects and applies it (PULL-based)
→ Git = single source of truth; the system continuously converges to it.
```

## Push vs pull deployment

```text
TRADITIONAL CI/CD (push) → the pipeline PUSHES changes TO the cluster (CI has cluster
  credentials, runs kubectl apply)
GITOPS (pull) → an agent INSIDE the cluster PULLS from Git and applies changes:
  ✓ cluster credentials stay IN the cluster (more secure — CI doesn't need them)
  ✓ continuous reconciliation (drift is auto-corrected)
→ Tools: ArgoCD, Flux (Kubernetes GitOps operators).
```

## Benefits

```text
✓ Git as SOURCE OF TRUTH — all changes via Git (versioned, reviewed, audited, history)
✓ DECLARATIVE — describe the desired state; the system makes it so (and KEEPS it so)
✓ AUTO-RECONCILIATION — drift (manual changes) is detected and corrected automatically
✓ Easy ROLLBACK — revert the Git commit → the system reverts to that state
✓ SECURITY — pull-based (no external system needs cluster credentials)
✓ Visibility — Git history = deployment history; clear audit trail
```

## Why it matters

Understanding GitOps is valuable senior-level knowledge because it's an important, increasingly-popular approach to continuous deployment (especially for Kubernetes), so it's useful for modern cloud-native delivery.

The core GitOps model — **Git as the single source of truth** for desired state, with an **automated agent continuously reconciling** actual state to match Git — represents a clean, powerful approach: you change systems by committing to Git, and the system pulls and applies the changes, continuously converging reality to the declared desired state.

Understanding the **push vs pull distinction** is key: traditional CI/CD **pushes** changes to the cluster (requiring CI to have cluster credentials), while GitOps uses a **pull** model where an agent inside the cluster pulls from Git — which is **more secure** (cluster credentials stay in the cluster; no external CI system needs them) and provides **continuous reconciliation** (automatically correcting drift).

Understanding the **benefits** — Git as the source of truth (all changes versioned, reviewed, audited via Git's history), the declarative model (describing desired state and having the system maintain it), **automatic reconciliation** (detecting and correcting drift from manual changes), easy **rollback** (reverting a Git commit reverts the system), enhanced **security** (pull-based), and **visibility** (Git history as deployment history) — explains why GitOps is compelling for Kubernetes deployment.

Knowing the tools (ArgoCD, Flux) provides practical context.

GitOps is increasingly adopted for cloud-native, Kubernetes-based delivery, representing a modern evolution of CD practices.

Since GitOps is an important, growing approach to continuous deployment (with real benefits in security, auditability, drift correction, and rollback) especially for Kubernetes, and since understanding its model, the pull-based approach, and its benefits is valuable for modern delivery, understanding GitOps is valuable senior-level knowledge for cloud-native continuous deployment — a modern, increasingly-relevant CD paradigm reflecting the evolution of deployment practices, useful for teams adopting Kubernetes and declarative, Git-driven operations.