Argo CD Interview Preparation¶
Curated from multiple DevOps interview sources, deduplicated, and edited for REBASH Academy. Every question includes a model answer. Answer out loud first, then reveal it. Prefer judgement and verification over memorised lists.
How to practise
- Answer in two minutes without notes
- Name the first three commands or checks you would run
- Call out a failure mode and a rollback
- Tie the answer to least privilege and blast radius
Core concepts¶
1. What is Argo Rollouts, and how does it implement progressive delivery?
Reveal answer
In short: Argo Rollouts adds canary and blue-green progressive delivery on top of Kubernetes.
Key points
- Rollout CR replaces a plain Deployment for stepwise traffic shifts.
- Steps like
setWeight: 10, pause, and analysis gate promotion. - Works with Ingress, Service Mesh, or ALB for traffic splitting.
- Failed analysis triggers automated rollback to the stable ReplicaSet.
Trap
- Skipping analysis templates turns canary into a slow rolling update with no real safety.
2. What is the difference between GitOps and traditional CD pipelines?
Reveal answer
In short: Traditional CD pushes with pipeline credentials; GitOps pulls and reconciles from Git.
Key points
- Push CD — CI holds kubeconfig and runs
kubectl/helm upgrade. - GitOps — Git is the desired state; a controller syncs the cluster continuously.
- Drift shows as OutOfSync; self-heal can revert unauthorised live changes.
- Auditing improves because every change is a Git commit (or a logged sync).
Trap
- Emergency
kubectl applywithout committing creates silent drift GitOps will fight or hide.
3. What are the key components of ArgoCD?
Reveal answer
In short: Argo CD is a control plane of specialised controllers around Applications.
Key points
- argocd-server — UI, CLI, and API.
- repo-server — clones Git and renders Helm/Kustomize manifests.
- application-controller — compares live vs desired and syncs.
- Application / ApplicationSet — unit of delivery and multi-cluster fan-out.
- Redis and Dex (optional) support caching and SSO.
Trap
- Giving every Application
cluster-admincollapses blast-radius controls.
4. What is argocd and why we are using it?
Reveal answer
In short: Argo CD keeps Kubernetes matching Git so deploys are reviewable, repeatable, and self-healing.
Key points
- Point it at manifests, Helm charts, or Kustomize overlays in Git.
- Sync brings the cluster to the chosen revision; prune removes deleted resources.
- Multi-cluster: one control plane can manage many destinations.
- Teams review PRs instead of approving opaque pipeline kubectl steps.
Trap
- Tracking mutable tags (
latest) makes “Git is truth” meaningless.
5. What is ArgoCD and what problem does it solve?
Reveal answer
In short: Argo CD solves cluster drift and unsafe push-deploy credentials with continuous Git reconciliation.
Key points
- Desired state is Git; live state is the cluster API.
- Detects OutOfSync when someone hotfixes live or a sync fails.
- Reduces long-lived cluster-admin tokens in CI for every app.
- Enables progressive delivery when paired with Rollouts or sync waves.
Trap
- Without PR checks on the GitOps repo, GitOps just automates bad YAML faster.
Practice questions¶
6. How do you design GitOps for 1000+ clusters with environment drift detection, emergency hotfixes, and controlled manual overrides?
Reveal answer
In short: Scale with ApplicationSets, sharded Argo CD, labelled cluster rings, and strict break-glass.
Key points
- ApplicationSet generators onboard clusters by label (env, region, ring).
- App-of-apps or mono-repo overlays keep per-cluster config thin.
- Drift: auto-sync + self-heal for platform apps; manual sync for high-risk.
- Hotfix: short sync window disable under change control, then commit the fix.
- Shard controllers/repo-servers; never one mega-instance for everything.
Try this
argocd app list --output wideargocd app get <app> --refresh
Trap
- Manual overrides without a ticket leave 1000 clusters in unknown states.
7. How do you perform application synchronization in ArgoCD?
Reveal answer
In short: Sync applies the Git revision’s manifests so live state matches desired state.
Key points
- Trigger via UI,
argocd app sync, API, or automated sync policies. - Prune deletes resources removed from Git; use carefully in prod.
- ApplyOutOfSyncOnly and sync options reduce blast radius.
- Hooks and sync waves order CRDs, operators, then workloads.
Try this
argocd app sync my-app --pruneargocd app wait my-app --health
Trap
- Force sync with prune on the wrong app can delete production resources in seconds.
8. How does GitOps differ from traditional IaC?
Reveal answer
In short: IaC defines infrastructure; GitOps adds a continuous reconciler that keeps reality matching Git.
Key points
- IaC (Terraform etc.) often push-applies on demand with a state file.
- GitOps continuously compares live API objects to a Git revision.
- IaC state maps IDs; GitOps state is Git + cluster (no
tfstatefor manifests). - Many platforms use both: Terraform for cloud accounts, Argo CD for Kubernetes apps.
Trap
- Calling any YAML in Git “GitOps” without a reconciliation agent is just versioned push.
Related¶
- Course: Argo CD
- Hub: Interview Preparation