Q: You implement a full GitOps pipeline where ArgoCD manages production. A developer accidentally pushes a Kubernetes deployment with `replicas: 0` to the GitOps repository. ArgoCD syncs it, silently taking down the entire service. How do you add a policy gate that prevents zero-replica deployments from ever reaching the GitOps repository?
Two layers of defence — one at the Git level, one at the cluster level:
Want to master this scenario in a live sandbox? KodeKloud's Enterprise GitOps with ArgoCD & Kubernetes Rollouts covers this exact problem with hands-on terminal drills.
🛠️ Production Runbook & Step-by-Step Resolution
Direct Answer: "In our delivery pipeline supporting multiple engineering squads, pipeline reliability was paramount. The interviewer is testing: Pre-merge policy gates, Conftest in CI, OPA/Kyverno at admission.. I structure my answer around systematic triage first, root cause analysis second, and permanent remediation third."
Production Solution & Architecture
Two layers of defence — one at the Git level, one at the cluster level: Layer 1 — CI policy gate on PR: In the CI pipeline that validates the GitOps repo, add a conftest check: Run conftest test on every PR diff. The PR is blocked from merging until this passes. Layer 2 — Admission controller in the cluster: Install a Kyverno policy as a final defence: Even if the Git gate is bypassed, the Kubernetes API server rejects the manifest. ArgoCD sync fails with a clear error message, and the service remains unaffected.
deny[msg] {
input.kind == "Deployment"
input.spec.replicas == 0
msg := sprintf("Deployment %v has replicas=0 which would cause an outage", [input.metadata.name])
}
- Immediate Triage: Two layers of defence — one at the Git level, one at the cluster level:
- Run targeted verification commands before modifying configuration.
- Automate permanent guardrails (CI check, alerts, IaC policy) to prevent recurrence.
Senior hiring managers and bar raisers evaluate how you prioritize during outages. Avoid these common anti-patterns:
- Promoting application releases to production without an automated, canary-validated rollback mechanism when configuring implement full GitOps Pipelines that lack automated rollback mechanisms force engineering teams to scramble manual hotfixes during severe production regressions.
- Running concurrent deployment pipeline jobs without stage mutex locking on shared target environments when configuring implement full GitOps Concurrent pipeline runs on the same environment race to apply state changes, causing out-of-order commits and unpredictable deployment state.
- Omitting dependency lockfiles (package-lock.json, go.sum, Cargo.lock) from continuous integration builds when configuring implement full GitOps Omitting lockfiles allows upstream package updates to break builds unpredictably without any changes to application codebase.
- Failing to define dynamic cache invalidation keys for CI package managers and layer caches when configuring implement full GitOps Stale build caches mask underlying dependency incompatibilities or pull outdated libraries, while un-cached runs slow feedback loops.