Argo Rollouts
Argo Rollouts is a Kubernetes controller that provides advanced deployment strategies like canary releases and blue-green deployments. It progressively shifts traffic to a new version, analyzes metrics, and automatically promotes or rolls back based on configurable success criteria.
itCloud native tools and technologies | OpenSkills.info
Intro
Argo Rollouts
Argo Rollouts is a Kubernetes controller for controlled application releases. It manages Rollout resources and their ReplicaSets through canary or blue-green strategies.
The central mental model is exposure controlled by evidence.
new pod template
|
v
new ReplicaSet -> limited exposure -> evidence -> promote or abort
| | |
+----------- Rollout controller --+
A Kubernetes Deployment can replace Pods gradually. Argo Rollouts adds explicit stages, traffic control, metric analysis, experiments, manual judgment, and automated failure handling.
Argo Rollouts does not build an image or choose a safe release policy for you. It executes the strategy and analysis rules that you declare.
The control loop
A Rollout resembles a Deployment. It has a selector, replica count, and Pod template. A change to spec.template creates a new ReplicaSet.
The Rollout controller then compares actual state with the selected strategy. It scales ReplicaSets, advances steps, updates Service selectors, and adjusts configured traffic-router resources.
The controller recognizes two important revisions:
- stable ReplicaSet — the last version that completed promotion;
- new ReplicaSet — the version currently moving through the strategy.
Continue the course
This section is part of the paid course.
See pricing to subscribe, or log in if you already have access.
Where this skill leads
Relevant careers
See how this topic contributes to broader role-level skill maps.
Sources
- https://argo-rollouts.readthedocs.io/en/stable/
Supports
- Argo Rollouts as a Kubernetes controller and custom resource set for blue-green, canary, analysis, experiments, and progressive delivery
- Rollout controller creation, scaling, and deletion of ReplicaSets from Pod-template changes
- Stable ReplicaSet assignment after successful progression
- Main ingress, service-mesh, and metric-provider integration categories
- https://argo-rollouts.readthedocs.io/en/stable/concepts/
Supports
- Rollout resource relationship to a Kubernetes Deployment
- Progressive delivery, canary, and blue-green definitions
- Strategy selection based on mixed-version compatibility and traffic-management capabilities
- Blue-green as the recommended simpler starting strategy
- https://argo-rollouts.readthedocs.io/en/stable/getting-started/
Supports
- Official basic canary workflow and command sequence
- Indefinite pause and manual promotion behavior
- Full promotion skipping remaining steps and analysis
- Abort returning serving behavior to stable while leaving the Rollout degraded
- Reapplying a previous Pod template to restore healthy desired state
- https://argo-rollouts.readthedocs.io/en/stable/features/canary/
Supports
- Canary step sequence and setWeight behavior
- Replica-based weight approximation without traffic routing
- setCanaryScale use only with a traffic router
- Dynamic stable scaling, abort scale-down delay, anti-affinity, max surge, and max unavailable controls
- Canary strategy without steps mimicking rolling update behavior
- https://argo-rollouts.readthedocs.io/en/stable/features/bluegreen/
Supports
- Active and preview Service behavior
- Service selector updates using ReplicaSet hashes
- Manual and automatic promotion controls
- Preview replica count and scale-down delay
- Pre-promotion and post-promotion analysis behavior
- https://argo-rollouts.readthedocs.io/en/stable/features/analysis/
Supports
- AnalysisTemplate, ClusterAnalysisTemplate, AnalysisRun, and Experiment roles
- Successful, failed, inconclusive, and error outcomes
- Background, inline, pre-promotion, and post-promotion analysis placement
- Metric provider, argument, condition, failure-limit, dry-run, retention, and secret-reference behavior
- Post-promotion analysis failure returning traffic to the previous stable ReplicaSet
- https://argo-rollouts.readthedocs.io/en/stable/features/experiment/
Supports
- Temporary ReplicaSets and optional AnalysisRuns created by an Experiment
- Baseline and canary comparison use case
- Experiment lifecycle, completion, failure, and scale-down behavior
- Blocking canary experiment steps and weighted experiments with supported traffic routers
- https://argo-rollouts.readthedocs.io/en/stable/features/traffic-management/
Supports
- Traffic management as routing control independent from Kubernetes Service endpoint balancing
- Stable and canary Service requirement for traffic-routed canaries
- Stable ReplicaSet capacity behavior during traffic-routed updates
- Weighted, header-based, and mirrored traffic concepts
- Managed route ordering and cleanup behavior
- https://argo-rollouts.readthedocs.io/en/stable/features/kubectl-plugin/
Supports
- Official plugin visualization of Rollouts, Experiments, and Analysis resources
- Routine promote and retry operations
- Kubernetes API and kubeconfig authentication model
- https://argo-rollouts.readthedocs.io/en/stable/generated/kubectl-argo-rollouts/kubectl-argo-rollouts/
Supports
- Current command families for get, lint, status, promote, abort, retry, undo, set image, pause, and restart
- Status command completion and timeout behavior
- https://argo-rollouts.readthedocs.io/en/stable/features/rollback/
Supports
- Default replay of strategy steps when an older manifest is reapplied
- Existing fast paths for active stable revisions
- Rollback window fast-tracking of configured recent revisions
- https://argo-rollouts.readthedocs.io/en/stable/best-practices/
Supports
- Need to verify application compatibility with simultaneous versions
- Warning for shared-resource, worker, and infrastructure application categories
- One-workload and one-cluster scope
- Recommendation against preview environments and long-running parallel release versions
- Need for fast, trustworthy metrics and dry-run evaluation
- https://argo-rollouts.readthedocs.io/en/stable/FAQ/
Supports
- Standalone operation without Argo CD, GitOps, or a service mesh
- Argo CD health and resource-action integration
- Controller availability and version-skew boundaries
- No Git read or write during abort and rollback
- Degraded desired state after returning traffic to stable
- Lack of dependency-aware coordination across applications
- https://argo-rollouts.readthedocs.io/en/stable/migrating/
Supports
- Conversion and workloadRef migration paths between Deployments and Rollouts
- Side-by-side workload warning for live migration
- Traffic-management considerations during migration
- https://argo-rollouts.readthedocs.io/en/stable/features/specification/
Supports
- Complete Rollout fields for canary and blue-green strategies
- Analysis, experiment, pause, traffic route, scale, rollback window, and progress deadline syntax
- https://github.com/argoproj/argo-rollouts/releases
Supports
- Argo Rollouts v1.9.0 as the latest stable release on 2026-07-14
- v1.10.0-rc1 as a prerelease rather than the stable source baseline
- Release notes, assets, checksums, and version history
- https://kubernetes.io/docs/concepts/workloads/controllers/deployment/
Supports
- Kubernetes Deployment creation and management of ReplicaSets
- RollingUpdate and rollback behavior used for comparison with Rollouts
- https://kubernetes.io/docs/concepts/services-networking/service/
Supports
- Kubernetes Service selection and routing to matching Pod endpoints
- Service role in stable, canary, active, and preview traffic paths