Guía Completa de Argo CD: Entrega Continua GitOps para Kubernetes

What Is Argo CD and GitOps?

Argo CD is a declarative, GitOps-based continuous delivery tool for Kubernetes, maintained as a CNCF graduated project. It continuously monitors Git repositories containing Kubernetes manifests and automatically synchronizes the desired application state to target clusters. When someone pushes a change to Git, Argo CD detects the difference between the live state and the desired state and reconciles them — either automatically or with manual approval.

GitOps is an operational framework that takes DevOps best practices — version control, collaboration, compliance, CI/CD — and applies them to infrastructure automation. The core principle is simple: the entire system state is described declaratively in Git, and an automated agent (Argo CD) ensures the live system matches that description at all times. Every change is a Git commit, every rollback is a Git revert, and every audit trail is a Git log.

Traditional CI/CD pipelines push changes to clusters using kubectl apply or helm upgrade from a CI runner. This push model has problems: CI credentials with cluster admin access are a security risk, pipeline failures can leave the cluster in a partially-applied state, and there is no continuous reconciliation — if someone manually changes a resource, nobody knows. Argo CD solves all of these by running inside the cluster, pulling desired state from Git, and continuously reconciling.

Installation: kubectl and Helm

Install with kubectl (Quick Start)

# Create namespace and install Argo CD
kubectl create namespace argocd
kubectl apply -n argocd -f \
  https://raw.githubusercontent.com/argoproj/argo-cd/stable/manifests/install.yaml

# Wait for all pods to be ready
kubectl wait --for=condition=Ready pod --all -n argocd --timeout=300s

# Get the initial admin password
argocd admin initial-password -n argocd

# Port-forward the API server
kubectl port-forward svc/argocd-server -n argocd 8080:443

# Login via CLI
argocd login localhost:8080 --username admin --password <initial-password>

# Change the admin password immediately
argocd account update-password

Install with Helm (Production)

# Add the Argo Helm repository
helm repo add argo https://argoproj.github.io/argo-helm
helm repo update

# Install with custom values
helm install argocd argo/argo-cd \
  --namespace argocd \
  --create-namespace \
  --set server.service.type=LoadBalancer \
  --set server.extraArgs[0]=--insecure \
  --set configs.params."server\.insecure"=true \
  --set ha.enabled=true \
  --set controller.replicas=2 \
  --set redis-ha.enabled=true \
  -f values-production.yaml

# values-production.yaml example:
# server:
#   replicas: 2
#   ingress:
#     enabled: true
#     hosts:
#       - argocd.example.com
#     tls:
#       - secretName: argocd-tls
#         hosts:
#           - argocd.example.com

Architecture: Core Components

Argo CD consists of three main components that work together to implement the GitOps reconciliation loop. Understanding this architecture is essential for troubleshooting, scaling, and securing your Argo CD installation.

Argo CD Architecture

  Git Repository          Argo CD (in-cluster)            Target Cluster(s)
  ┌──────────┐     ┌─────────────────────────────┐     ┌──────────────────┐
  │ manifests│     │  ┌────────────────────────┐  │     │                  │
  │ helm/    │◄────┤  │    Repo Server         │  │     │  Namespaces      │
  │ kustomize│     │  │  (clone, render, cache) │  │     │  ┌────────────┐ │
  │ plain/   │     │  └───────────┬────────────┘  │     │  │ Deployments│ │
  └──────────┘     │              │                │     │  │ Services   │ │
                   │  ┌───────────▼────────────┐  │     │  │ ConfigMaps │ │
                   │  │  Application Controller │  │────►│  │ Secrets    │ │
                   │  │  (diff, sync, health)   │  │     │  │ Ingress    │ │
                   │  └───────────┬────────────┘  │     │  └────────────┘ │
                   │              │                │     └──────────────────┘
  Web UI / CLI     │  ┌───────────▼────────────┐  │
  ┌──────────┐     │  │      API Server        │  │
  │ Browser  │◄────┤  │  (gRPC/REST, auth,     │  │
  │ argocd   │     │  │   RBAC, web UI)        │  │
  │  CLI     │     │  └────────────────────────┘  │
  └──────────┘     └─────────────────────────────┘

API Server

The API Server is the central component that exposes the gRPC and REST API consumed by the web UI, CLI, and CI/CD systems. It handles authentication (local accounts, SSO via OIDC/SAML/LDAP), authorization (RBAC policies), and serves the built-in web dashboard. It manages Application and AppProject resources and triggers sync operations.

Repo Server

The Repo Server clones Git repositories, caches them locally, and renders Kubernetes manifests. It supports plain YAML, Helm charts (helm template), Kustomize overlays (kustomize build), Jsonnet, and custom config management plugins. The Repo Server is stateless and horizontally scalable — for large installations with hundreds of applications, run multiple replicas.

Application Controller

The Application Controller is the brain of Argo CD. It continuously monitors running applications, compares the live state against the desired state from the Repo Server, detects drift (OutOfSync), and optionally auto-syncs. It also evaluates resource health using built-in and custom health checks. The controller runs the reconciliation loop on a configurable interval (default: 3 minutes for full reconciliation, with webhook-triggered instant checks).

The Application CRD

The Application custom resource is the fundamental building block in Argo CD. It defines what to deploy (source), where to deploy (destination), and how to sync (sync policy). Applications can be created via the UI, CLI, or declaratively as YAML — the latter is the GitOps-recommended approach, often called the "app of apps" pattern.

# application.yaml - Argo CD Application CRD
apiVersion: argoproj.io/v1alpha1
kind: Application
metadata:
  name: my-web-app
  namespace: argocd
  finalizers:
    - resources-finalizer.argocd.argoproj.io
spec:
  project: default
  source:
    repoURL: https://github.com/org/k8s-manifests.git
    targetRevision: main
    path: apps/web-app/overlays/production
  destination:
    server: https://kubernetes.default.svc
    namespace: production
  syncPolicy:
    automated:
      prune: true
      selfHeal: true
      allowEmpty: false
    syncOptions:
      - CreateNamespace=true
      - PrunePropagationPolicy=foreground
      - PruneLast=true
    retry:
      limit: 5
      backoff:
        duration: 5s
        factor: 2
        maxDuration: 3m

AppProjects: Multi-Tenancy and Access Control

AppProjects provide logical grouping and access control for Applications. They restrict which Git repositories an Application can use as source, which clusters and namespaces it can deploy to, and which Kubernetes resource kinds are allowed or denied. Projects are essential for multi-tenancy — different teams get different projects with scoped permissions.

apiVersion: argoproj.io/v1alpha1
kind: AppProject
metadata:
  name: team-backend
  namespace: argocd
spec:
  description: Backend team project
  sourceRepos:
    - https://github.com/org/backend-manifests.git
    - https://github.com/org/shared-charts.git
  destinations:
    - server: https://kubernetes.default.svc
      namespace: backend-*
    - server: https://staging-cluster.example.com
      namespace: backend-*
  clusterResourceWhitelist:
    - group: ""
      kind: Namespace
  namespaceResourceBlacklist:
    - group: ""
      kind: ResourceQuota
    - group: ""
      kind: LimitRange
  roles:
    - name: backend-dev
      description: Backend developer role
      policies:
        - p, proj:team-backend:backend-dev, applications, get, team-backend/*, allow
        - p, proj:team-backend:backend-dev, applications, sync, team-backend/*, allow
      groups:
        - backend-developers

Sync Policies: Automated vs Manual

Sync policies determine how Argo CD reconciles the live state with the desired state. Manual sync requires explicit user action (via UI or CLI) to apply changes. Automated sync triggers synchronization whenever the desired state in Git diverges from the live state. Each mode has distinct use cases depending on your deployment risk tolerance.

# Manual sync via CLI
argocd app sync my-web-app --prune --force

# Automated sync in Application spec
# syncPolicy:
#   automated:
#     prune: true        # Delete resources removed from Git
#     selfHeal: true     # Revert manual cluster changes
#     allowEmpty: false  # Prevent sync if Git returns empty manifests

Sync Waves and Hooks

Sync waves control the order in which Argo CD applies resources during a sync. Resources are assigned wave numbers via annotations: lower waves are applied first, and Argo CD waits for all resources in a wave to be healthy before proceeding to the next. Hooks are special resources (usually Jobs) that run at specific lifecycle phases.

# Wave -1: Namespace and RBAC (created first)
apiVersion: v1
kind: Namespace
metadata:
  name: my-app
  annotations:
    argocd.argoproj.io/sync-wave: "-1"
---
# Wave 0: ConfigMaps and Secrets
apiVersion: v1
kind: ConfigMap
metadata:
  name: app-config
  annotations:
    argocd.argoproj.io/sync-wave: "0"
data:
  DATABASE_HOST: postgres.db.svc
---
# Wave 1: Database migration (PreSync hook)
apiVersion: batch/v1
kind: Job
metadata:
  name: db-migrate
  annotations:
    argocd.argoproj.io/hook: PreSync
    argocd.argoproj.io/hook-delete-policy: HookSucceeded
spec:
  template:
    spec:
      containers:
        - name: migrate
          image: my-app:latest
          command: ["./migrate", "--up"]
      restartPolicy: Never
---
# Wave 2: Application Deployment
apiVersion: apps/v1
kind: Deployment
metadata:
  name: my-app
  annotations:
    argocd.argoproj.io/sync-wave: "2"
---
# PostSync hook: Smoke test
apiVersion: batch/v1
kind: Job
metadata:
  name: smoke-test
  annotations:
    argocd.argoproj.io/hook: PostSync
    argocd.argoproj.io/hook-delete-policy: HookSucceeded
spec:
  template:
    spec:
      containers:
        - name: test
          image: curlimages/curl:latest
          command: ["curl", "-f", "http://my-app.my-app.svc/healthz"]
      restartPolicy: Never

Hook Types Reference

Health Checks

Argo CD evaluates the health of every resource it manages. Built-in health checks cover Deployments (are all replicas available?), StatefulSets, DaemonSets, Ingress, PersistentVolumeClaims, and more. For custom resources (CRDs), you can define custom health check scripts in Lua that Argo CD evaluates to determine Healthy, Progressing, Degraded, Suspended, or Missing status.

# Custom health check for a CronJob (in argocd-cm ConfigMap)
# data:
#   resource.customizations.health.batch_CronJob: |
#     hs = {}
#     if obj.status ~= nil then
#       if obj.status.lastScheduleTime ~= nil then
#         hs.status = "Healthy"
#         hs.message = "CronJob is scheduled"
#       else
#         hs.status = "Progressing"
#         hs.message = "Waiting for first schedule"
#       end
#     else
#       hs.status = "Progressing"
#       hs.message = "No status yet"
#     end
#     return hs

Multi-Cluster Management

Argo CD can manage applications across multiple Kubernetes clusters from a single control plane. This is critical for organizations with separate clusters for development, staging, production, or regional deployments. External clusters are registered and their credentials are stored securely as Kubernetes Secrets in the Argo CD namespace.

# Register an external cluster
argocd cluster add staging-context --name staging-cluster
argocd cluster add prod-us-east --name prod-us-east
argocd cluster add prod-eu-west --name prod-eu-west

# List registered clusters
argocd cluster list

# Deploy to a specific cluster
apiVersion: argoproj.io/v1alpha1
kind: Application
metadata:
  name: web-app-prod-us
  namespace: argocd
spec:
  project: production
  source:
    repoURL: https://github.com/org/manifests.git
    path: overlays/prod-us-east
    targetRevision: main
  destination:
    server: https://prod-us-east.example.com
    namespace: web-app

SSO and RBAC

Argo CD supports Single Sign-On via OIDC (Okta, Auth0, Keycloak, Google), SAML 2.0, LDAP, and GitHub/GitLab OAuth. RBAC policies use a Casbin model to control who can view, sync, delete, or override applications at the project or global level. This enables fine-grained access control for multi-team environments.

# argocd-cm ConfigMap - OIDC SSO configuration
data:
  url: https://argocd.example.com
  oidc.config: |
    name: Okta
    issuer: https://org.okta.com/oauth2/default
    clientID: argocd-client-id
    clientSecret: $oidc.okta.clientSecret
    requestedScopes:
      - openid
      - profile
      - email
      - groups

# argocd-rbac-cm ConfigMap - RBAC policies
data:
  policy.default: role:readonly
  policy.csv: |
    # Admins can do everything
    p, role:admin, applications, *, */*, allow
    p, role:admin, clusters, *, *, allow
    p, role:admin, repositories, *, *, allow
    p, role:admin, projects, *, *, allow
    
    # Developers can view and sync their project apps
    p, role:developer, applications, get, team-backend/*, allow
    p, role:developer, applications, sync, team-backend/*, allow
    p, role:developer, applications, action/*, team-backend/*, allow
    
    # Map SSO groups to roles
    g, platform-team, role:admin
    g, backend-devs, role:developer

Secrets Management

Storing plain-text Kubernetes Secrets in Git violates security best practices. Argo CD integrates with multiple secret management solutions that keep encrypted or referenced secrets in Git while the actual values are resolved at sync time or by in-cluster controllers.

Sealed Secrets (Bitnami)

# Install Sealed Secrets controller
helm repo add sealed-secrets https://bitnami-labs.github.io/sealed-secrets
helm install sealed-secrets sealed-secrets/sealed-secrets -n kube-system

# Encrypt a secret (safe to commit to Git)
kubectl create secret generic db-creds \
  --from-literal=password=supersecret \
  --dry-run=client -o yaml | \
  kubeseal --controller-name=sealed-secrets \
           --controller-namespace=kube-system \
           --format yaml > sealed-db-creds.yaml

# The SealedSecret can be safely stored in Git
# Argo CD syncs it; the controller decrypts it in-cluster

External Secrets Operator

# ExternalSecret syncs from AWS Secrets Manager
apiVersion: external-secrets.io/v1beta1
kind: ExternalSecret
metadata:
  name: db-credentials
spec:
  refreshInterval: 1h
  secretStoreRef:
    name: aws-secrets-manager
    kind: ClusterSecretStore
  target:
    name: db-credentials
    creationPolicy: Owner
  data:
    - secretKey: username
      remoteRef:
        key: prod/db-credentials
        property: username
    - secretKey: password
      remoteRef:
        key: prod/db-credentials
        property: password

ApplicationSets: Managing Apps at Scale

The ApplicationSet controller generates Argo CD Application resources from templates using generators. Instead of manually creating hundreds of Application CRDs, you define a template and a generator that produces the parameter sets. This is essential for managing monorepos, multi-cluster deployments, and dynamic environments.

Cluster Generator (Multi-Cluster)

apiVersion: argoproj.io/v1alpha1
kind: ApplicationSet
metadata:
  name: web-app-all-clusters
  namespace: argocd
spec:
  generators:
    - clusters:
        selector:
          matchLabels:
            env: production
  template:
    metadata:
      name: web-app-{{name}}
    spec:
      project: production
      source:
        repoURL: https://github.com/org/manifests.git
        targetRevision: main
        path: overlays/{{metadata.labels.region}}
      destination:
        server: "{{server}}"
        namespace: web-app
      syncPolicy:
        automated:
          prune: true
          selfHeal: true

Git Directory Generator (Monorepo)

apiVersion: argoproj.io/v1alpha1
kind: ApplicationSet
metadata:
  name: monorepo-services
  namespace: argocd
spec:
  generators:
    - git:
        repoURL: https://github.com/org/monorepo.git
        revision: main
        directories:
          - path: services/*
  template:
    metadata:
      name: "{{path.basename}}"
    spec:
      project: default
      source:
        repoURL: https://github.com/org/monorepo.git
        targetRevision: main
        path: "{{path}}"
      destination:
        server: https://kubernetes.default.svc
        namespace: "{{path.basename}}"

Pull Request Generator (Preview Environments)

apiVersion: argoproj.io/v1alpha1
kind: ApplicationSet
metadata:
  name: pr-previews
  namespace: argocd
spec:
  generators:
    - pullRequest:
        github:
          owner: org
          repo: web-app
          tokenRef:
            secretName: github-token
            key: token
        requeueAfterSeconds: 60
  template:
    metadata:
      name: preview-{{branch_slug}}
    spec:
      project: previews
      source:
        repoURL: https://github.com/org/web-app.git
        targetRevision: "{{head_sha}}"
        path: k8s/preview
      destination:
        server: https://kubernetes.default.svc
        namespace: preview-{{number}}

Argo CD Notifications

Argo CD Notifications (built-in since v2.6) sends alerts to Slack, Teams, email, webhooks, GitHub, and more when application events occur (sync succeeded, sync failed, health degraded). Notifications are configured via the argocd-notifications-cm ConfigMap with triggers and templates.

# argocd-notifications-cm ConfigMap
apiVersion: v1
kind: ConfigMap
metadata:
  name: argocd-notifications-cm
  namespace: argocd
data:
  service.slack: |
    token: $slack-token
  trigger.on-sync-succeeded: |
    - when: app.status.operationState.phase in ["Succeeded"]
      send: [app-sync-succeeded]
  trigger.on-sync-failed: |
    - when: app.status.operationState.phase in ["Error", "Failed"]
      send: [app-sync-failed]
  trigger.on-health-degraded: |
    - when: app.status.health.status == "Degraded"
      send: [app-health-degraded]
  template.app-sync-succeeded: |
    message: |
      Application {{.app.metadata.name}} sync succeeded.
      Revision: {{.app.status.sync.revision}}
  template.app-sync-failed: |
    message: |
      Application {{.app.metadata.name}} sync FAILED.
      Error: {{.app.status.operationState.message}}

# Subscribe an app to notifications (annotation)
# metadata:
#   annotations:
#     notifications.argoproj.io/subscribe.on-sync-succeeded.slack: deploys

Argo CD with Helm and Kustomize

Helm Integration

Argo CD renders Helm charts using helm template (not helm install), which means it treats Helm as a manifest generator rather than a release manager. You can specify Helm values inline, from files in the repo, or from external value files. This approach gives Argo CD full control over resource lifecycle while leveraging the Helm chart ecosystem.

apiVersion: argoproj.io/v1alpha1
kind: Application
metadata:
  name: nginx-ingress
  namespace: argocd
spec:
  project: infrastructure
  source:
    repoURL: https://kubernetes.github.io/ingress-nginx
    chart: ingress-nginx
    targetRevision: 4.9.1
    helm:
      releaseName: nginx-ingress
      valuesObject:
        controller:
          replicaCount: 3
          service:
            type: LoadBalancer
          metrics:
            enabled: true
  destination:
    server: https://kubernetes.default.svc
    namespace: ingress-nginx

Kustomize Integration

# Application using Kustomize overlays
apiVersion: argoproj.io/v1alpha1
kind: Application
metadata:
  name: api-prod
  namespace: argocd
spec:
  source:
    repoURL: https://github.com/org/manifests.git
    path: api/overlays/production
    targetRevision: main
    kustomize:
      namePrefix: prod-
      commonLabels:
        env: production
      images:
        - name=api-image
          newName=registry.example.com/api
          newTag=v2.5.0
  destination:
    server: https://kubernetes.default.svc
    namespace: api-production

# Corresponding kustomization.yaml in the repo
# apiVersion: kustomize.config.k8s.io/v1beta1
# kind: Kustomization
# resources:
#   - ../../base
# patches:
#   - path: replica-patch.yaml
#   - path: resource-limits-patch.yaml

CI/CD Pipeline Integration

In a GitOps workflow, CI and CD are cleanly separated. CI (GitHub Actions, GitLab CI, Jenkins) builds, tests, and pushes container images. CD is handled entirely by Argo CD watching the Git repo. The CI pipeline updates the image tag in the Git manifest repository, and Argo CD detects the change and syncs. This separation means CI never needs cluster credentials.

# GitHub Actions CI pipeline that triggers Argo CD
name: Build and Update Manifests
on:
  push:
    branches: [main]

jobs:
  build:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4

      - name: Build and push image
        run: |
          docker build -t registry.example.com/app:$GITHUB_SHA .
          docker push registry.example.com/app:$GITHUB_SHA

      - name: Update manifest repo
        run: |
          git clone https://github.com/org/k8s-manifests.git
          cd k8s-manifests
          # Update image tag in kustomization.yaml
          cd apps/web-app/overlays/production
          kustomize edit set image \
            app=registry.example.com/app:$GITHUB_SHA
          git add .
          git commit -m "Update web-app to $GITHUB_SHA"
          git push
          # Argo CD detects change and syncs automatically

Image Updater (Automatic Image Tag Updates)

Argo CD Image Updater is a companion tool that watches container registries for new image tags and automatically updates the Application source. This eliminates the need for the CI pipeline to commit image tag changes to Git — the Image Updater handles it.

# Application annotation for Image Updater
metadata:
  annotations:
    argocd-image-updater.argoproj.io/image-list: |
      app=registry.example.com/app
    argocd-image-updater.argoproj.io/app.update-strategy: semver
    argocd-image-updater.argoproj.io/app.allow-tags: regexp:^v[0-9]+\.[0-9]+\.[0-9]+$
    argocd-image-updater.argoproj.io/write-back-method: git

Disaster Recovery

Since Argo CD follows the GitOps principle, disaster recovery is straightforward: your Git repositories contain the desired state of all applications. If the Argo CD installation is lost, reinstall it, reconnect the repositories, and all applications will be re-synced automatically. However, you should also back up Argo CD-specific configuration.

# Export all Argo CD resources for backup
argocd admin export -n argocd > argocd-backup.yaml

# Back up specific resource types
kubectl get applications -n argocd -o yaml > apps-backup.yaml
kubectl get appprojects -n argocd -o yaml > projects-backup.yaml
kubectl get secrets -n argocd -l argocd.argoproj.io/secret-type=repository -o yaml > repos-backup.yaml
kubectl get secrets -n argocd -l argocd.argoproj.io/secret-type=cluster -o yaml > clusters-backup.yaml

# Restore from backup
argocd admin import -n argocd < argocd-backup.yaml

# Best practice: Store Argo CD config as code
# Keep Application, AppProject, and config YAMLs
# in a dedicated Git repo managed by a bootstrap app

Best Practices

• Separate app repo from config repo — Keep application source code and Kubernetes manifests in separate Git repositories. CI builds images from the app repo; CD reads manifests from the config repo. This decouples application releases from infrastructure changes.
• Use the App of Apps pattern — Define a root Application that points to a directory of Application manifests. When you add a new app, commit its Application YAML to this directory, and the root app auto-syncs it.
• Enable automated sync with selfHeal for dev/staging — In non-production environments, automated sync with self-heal ensures the cluster always matches Git, catching manual changes immediately.
• Use manual sync with approval gates for production — In production, require explicit sync approval. Combine with CI checks (passing tests, security scans) before the manifest repo is updated.
• Pin targetRevision to tags or SHA in production — Avoid tracking HEAD of main in production. Pin to a specific Git tag or commit SHA for predictable, auditable deployments.
• Leverage sync waves for dependency ordering — Ensure CRDs are created before custom resources, databases before apps, and config before deployments.
• Set resource limits on Argo CD components — The Application Controller and Repo Server can be memory-intensive with hundreds of apps. Set appropriate resource requests and limits, and scale replicas horizontally.
• Use Projects to enforce least-privilege — Restrict each teams project to specific repos, clusters, namespaces, and resource kinds. Deny cluster-scoped resources unless explicitly needed.
• Integrate with Prometheus and Grafana — Argo CD exports Prometheus metrics. Monitor sync duration, sync failures, app health status, and API server request latency.
• Adopt Progressive Delivery — Combine Argo CD with Argo Rollouts for canary deployments, blue-green releases, and traffic shifting experiments before full promotion.

Argo CD vs Flux: Choosing a GitOps Tool

Argo CD and Flux are both CNCF graduated GitOps projects for Kubernetes. They share the same core principle (Git as source of truth, pull-based reconciliation) but differ significantly in architecture, user experience, and ecosystem integration.

Frequently Asked Questions

What is Argo CD and how does it differ from traditional CI/CD?

Argo CD is a declarative GitOps continuous delivery tool for Kubernetes. Unlike traditional push-based CI/CD where a pipeline runs kubectl apply or helm upgrade, Argo CD uses a pull-based model. It runs inside the cluster, continuously watches Git repositories, compares the desired state in Git against the live cluster state, and reconciles any differences. This means Git becomes the single source of truth, all changes are auditable through Git history, drift is automatically detected, and CI pipelines never need direct cluster access.

How do I install Argo CD?

Install Argo CD by creating the argocd namespace and applying the official install manifest with kubectl. For production, use the Argo Helm chart with HA (high availability) configuration — multiple API Server replicas, Redis HA, and the Application Controller sharding feature for large-scale deployments. After installation, retrieve the initial admin password from the argocd-initial-admin-secret, port-forward or expose the server via Ingress, and log in.

What are sync waves and hooks?

Sync waves control resource application order during a sync. Annotate resources with argocd.argoproj.io/sync-wave and an integer value — lower values are applied first. Hooks (PreSync, Sync, PostSync, SyncFail) are Jobs that run at specific lifecycle phases. For example, use a PreSync hook for database migrations before the app deployment, and a PostSync hook for smoke tests after deployment completes.

How does Argo CD handle secrets?

Never commit plain-text secrets to Git. Use Sealed Secrets (encrypts secrets client-side; the controller decrypts in-cluster), External Secrets Operator (syncs secrets from AWS Secrets Manager, Vault, Azure Key Vault), or SOPS with age/GPG encryption. Argo CD syncs these encrypted or reference resources like any other manifest, and the respective controller resolves the actual secret values within the cluster.

What is the App of Apps pattern?

The App of Apps pattern is a bootstrapping approach where a single root Application points to a Git directory containing other Application manifests. When Argo CD syncs the root app, it creates all child applications, which in turn sync their own resources. This enables managing your entire platform — infrastructure components, shared services, application deployments — from a single entry point. Adding a new application is as simple as committing a new Application YAML file.

How does multi-cluster deployment work?

Register external clusters using argocd cluster add, which creates a ServiceAccount on the target cluster and stores credentials in the Argo CD namespace. Applications can then target any registered cluster by specifying its server URL in the destination field. Combined with ApplicationSets and the cluster generator, you can deploy the same application to all clusters matching a label selector from a single Argo CD instance.

When should I use ApplicationSets?

Use ApplicationSets when you need to manage applications at scale: deploying across multiple clusters (cluster generator), managing microservices in a monorepo (Git directory generator), creating preview environments from pull requests (PR generator), or generating applications from a dynamic list of environments (list or matrix generator). ApplicationSets eliminate the need to manually create and maintain hundreds of Application CRDs.

How does Argo CD compare to Flux?

Both are CNCF graduated GitOps tools. Argo CD offers a rich web UI, built-in SSO/RBAC, multi-cluster management, and ApplicationSets — ideal for teams needing visibility and multi-tenancy. Flux uses a microservices architecture with separate controllers for sources, Kustomize, and Helm, and has built-in image automation. Flux suits teams preferring a minimal, CLI-driven, Kubernetes-native approach. For most teams starting with GitOps, Argo CD provides a smoother onboarding experience due to its UI and comprehensive feature set.

Conclusion

Argo CD brings the principles of GitOps to Kubernetes continuous delivery, replacing fragile push-based scripts with a robust, auditable, self-healing reconciliation loop. By making Git the single source of truth, every deployment is versioned, every change is reviewable, and every rollback is a git revert. The combination of Application CRDs, sync policies, waves, hooks, ApplicationSets, and multi-cluster support makes Argo CD suitable for everything from a single-app startup to an enterprise platform managing thousands of microservices across global clusters. Start simple — one Application, one cluster, manual sync — and progressively adopt automation, ApplicationSets, and progressive delivery with Argo Rollouts as your needs grow.