Skip to content

Certified Argo Project Associate (CAPA) Fact Sheet

Exam Overview

Exam Code: CAPA Exam Name: Certified Argo Project Associate Level: Associate Duration: 90 minutes Format: Multiple choice and multiple select, online proctored Questions: 60 Passing Score: 75% Cost: USD 250 (includes one free retake) Valid For: 2 years Delivery: Online proctored through PSI Prerequisites: None; Kubernetes fundamentals assumed

Verify before booking. Confirm current details on the official pages below.

📖 CAPA certification page - registration and curriculum 📖 Linux Foundation CAPA page - logistics 📖 Argo Project documentation - the four projects 📖 CNCF curriculum repository - published exam domains

Four projects, one exam

The Argo Project is four separate tools that share a name and a community. CAPA covers all four, and the weighting surprises people who assume it is an Argo CD exam:

Project Weight What it does
Argo Workflows 36% Container-native workflow engine: DAGs and step sequences run as pods
Argo CD 34% GitOps continuous delivery: reconciles cluster state against a repository
Argo Rollouts 18% Progressive delivery: canary and blue-green with automated analysis
Argo Events 12% Event-driven automation: event sources, sensors, and triggers

Workflows is the largest domain, at slightly more than Argo CD. Candidates who study only Argo CD are studying a third of the exam.

Target Audience

  • Platform engineers running any part of the Argo stack
  • SREs and DevOps engineers implementing GitOps or progressive delivery
  • Data and ML engineers using Argo Workflows for pipelines
  • Anyone holding CKA or KCNA extending into delivery tooling

Assumed background: Kubernetes objects, custom resources, and controllers. CAPA does not teach Kubernetes.

Exam Domains

Domain 1: Argo Workflows (36%)

Key Concepts: - The Workflow custom resource and the workflow controller - Templates: container, script, resource, suspend, and the DAG and steps orchestration templates - Template invocation, templateRef, and WorkflowTemplate versus ClusterWorkflowTemplate - Parameters and artifacts: inputs, outputs, and passing values between steps - Artifact repositories (S3, GCS, Azure Blob, MinIO) and artifact garbage collection - Volumes, volumeClaimTemplates, and sharing data between steps - Conditionals (when), loops (withItems, withParam, withSequence), and recursion - Retry strategy, timeouts, and activeDeadlineSeconds - Exit handlers and lifecycle hooks - CronWorkflow for scheduled execution - Workflow archive, pod garbage collection, and TTL strategy - Synchronization: mutexes and semaphores for concurrency limits - Security: service accounts, podSpecPatch, and the workflow executor - The Argo Workflows UI, CLI, and Events integration

📖 Argo Workflows documentation - templates, artifacts, and orchestration

Domain 2: Argo CD (34%)

Key Concepts: - Architecture: API server, repository server, application controller, Redis, and Dex or an external OIDC provider - The Application custom resource: source, destination, project, and sync policy - ApplicationSet and its generators: list, cluster, Git, matrix, merge, pull request, SCM provider - AppProject for multi-tenancy: allowed sources, destinations, and resource kinds - Sync policies: manual versus automated, prune, self-heal - Sync options, waves, hooks (PreSync, Sync, PostSync, SyncFail), and resource ordering - Health assessment and custom health checks - Diffing, ignoreDifferences, and managing fields owned by other controllers - Tools support: Kustomize, Helm, jsonnet, and config management plugins - Multi-cluster management and cluster registration - RBAC, projects, and SSO integration - Notifications and the Argo CD Image Updater - Declarative setup: managing Argo CD itself with Argo CD (app of apps)

📖 Argo CD documentation - applications, projects, sync

Domain 3: Argo Rollouts (18%)

Key Concepts: - The Rollout custom resource as a Deployment replacement - Canary strategy: steps, setWeight, pause, and traffic routing - Blue-green strategy: active and preview services, autoPromotionEnabled, scale-down delay - Traffic management integrations: Istio, NGINX, ALB, SMI, Gateway API - AnalysisTemplate, ClusterAnalysisTemplate, and AnalysisRun - Metric providers: Prometheus, Datadog, New Relic, CloudWatch, Wavefront, Job, Web - Automated promotion and automatic rollback on failed analysis - Experiments for comparing versions side by side - The Rollouts dashboard and kubectl argo rollouts plugin

📖 Argo Rollouts documentation - strategies and analysis

Domain 4: Argo Events (12%)

Key Concepts: - Architecture: EventSource, Sensor, EventBus - Event sources: webhook, S3, calendar, Kafka, SQS, GitHub, GitLab, Redis, resource - Sensors: dependencies, filters, and triggers - Triggers: Argo Workflow, Kubernetes object, HTTP, AWS Lambda, Kafka, Slack - Trigger parameterization from event payloads - Event filtering by data, context, time, and expression - The EventBus implementations (NATS, Jetstream, Kafka)

📖 Argo Events documentation - event sources, sensors, triggers

Which tool for which job

Requirement Tool
Run a multi-step batch or ML pipeline in containers Argo Workflows
Keep clusters matching a Git repository Argo CD
Release a new version to 10% of traffic and roll back on errors Argo Rollouts
Trigger something when a file lands in S3 or a webhook fires Argo Events
Run a workflow on a schedule CronWorkflow (Argo Workflows)
Generate one Application per cluster from a template ApplicationSet (Argo CD)