Canary deployment
A canary deployment is a release strategy that exposes a change to a limited portion of users, traffic, or infrastructure before expanding it more broadly.
Platform Engineering on Kubernetes
A canary release is a deployment technique that gradually routes live traffic to a new software version while retaining traffic on a stable version for early validation and rollback.
- How it is used
- Reduces release risk by validating a new service version against a controlled share of live traffic before full promotion.
- How deep it goes
- Works through replica-based splitting, Knative revision rules, and Argo Rollouts steps; defers detailed Istio configuration.
- Examples
- Uses a notifications service, 50/50 Knative revision routing, curl responses from two versions, and a staged Argo Rollout.
- What you can do after
- Apply it via Knative Service traffic rules or Argo Rollouts steps, then validate, promote, or redirect traffic to the stable version.
Read how Platform Engineering on Kubernetes explains canary deployment →
Kubernetes in Action, Second Edition
A release technique that exposes a new application version to a limited subset of instances or traffic before expanding it broadly.
- How it is used
- Limits exposure to a new application version while its behavior is checked before a wider rollout.
- How deep it goes
- Explains two-Deployment traffic distribution and manual rollout controls, but gives no automated approval procedure.
- Examples
- Uses stable and canary pods, rollout pause and resume commands, and a three-replica StatefulSet partitioned to update one pod.
- What you can do after
- Practice/perform a limited rollout using separate Deployments, a paused Deployment update, or a StatefulSet partition.
Read how Kubernetes in Action, Second Edition explains canary deployment →
Infrastructure as Code, Patterns and Practices: With examples in Python and Terraform
A canary deployment is a release pattern that incrementally exposes a change to a limited portion of users, traffic, or infrastructure before expanding it broadly.
- How it is used
- Deploys significant infrastructure changes gradually to detect failures before they affect all requests.
- How deep it goes
- Explains traffic weighting, rollback, monitoring, regression windows, and stateful cutover, but leaves increment selection context-dependent.
- Examples
- Uses blue-green networks, weighted load-balancer services, application instances, and staged database writes and reads.
- What you can do after
- Operate/configure incremental traffic shifts, monitor regressions, and revert to the previous environment when failures emerge.
Machine Learning System Design
Canary deployment is a release strategy that sends a limited portion of users or traffic to a new version before fully switching over.
- How it is used
- Controls production risk by testing a release with a small user portion before completing the switch.
- How deep it goes
- Defines concurrent versions, granular routing, rollback, and A/B-test distinction without detailing rollout controls or evaluation metrics.
- Examples
- Uses a search system that releases ranking models separately from its index, filter, and API layers.
- What you can do after
- Operate/configure staged production releases with fallback versions and independently deployed ML components.
Read how Machine Learning System Design explains canary deployment →
Google Anthos in Action
A canary deployment is a release strategy that exposes a new version to a small portion of traffic before expanding its availability.
- How it is used
- Supports safer microservice releases by testing new versions with limited production traffic before broad rollout.
- How deep it goes
- Explains rollout stages, concurrent versions, percentage and user-agent routing, but does not detail deployment configuration.
- Examples
- Uses a 5% traffic split and user-agent targeting for Service A through Istio.
- What you can do after
- Operate progressive releases by routing selected traffic to a new version and expanding exposure after testing.
Read how Google Anthos in Action explains canary deployment →
Think Like a Software Engineering Manager
A canary deployment releases a software change to a limited subset of users or infrastructure before expanding it to the full population.
- How it is used
- Improves production safety by releasing changes gradually to a subset of users before full deployment.
- How deep it goes
- Defines gradual release and feedback-based expansion, but does not detail rollout mechanics or measurement.
- Examples
- Uses subset releases alongside alpha, beta, pre-production, and production deployment stages.
- What you can do after
- Apply it via gradual production rollout, using feedback and stability to guide expansion.
Read how Think Like a Software Engineering Manager explains canary deployment →
Kubernetes for Developers
A rollout is the controlled propagation of an application or platform update across its deployed instances or locations.
- How it is used
- Uses limited traffic exposure to validate an application version before completing a production cutover.
- How deep it goes
- Describes staged users, traffic percentages, validation, and decisions, but stops short of defining detailed analysis criteria.
- Examples
- Uses internal users followed by a percentage of external production traffic before full cutover.
- What you can do after
- Apply it via Istio and Spinnaker to split traffic and automate deployment decisions.
Read how Kubernetes for Developers explains canary deployment →
Related concepts
- blue-green deploymentalternative to · Blue-green deployment is a release pattern that maintains separate active and replacement environments and switches traffic between them after the replacement is validated.
- deploymentkind of · Deployment is the process of making software, models, or other computational systems available to operate in a target environment.
- testing in productionused by · Testing in production is the practice of evaluating software or infrastructure behavior in a live production environment.
- evaluation suiteuses · An evaluation suite is an organized collection of tests, cases, or tasks used to assess a system against specified behaviors or criteria.
- experimentationcontrasts with · Experimentation is the deliberate testing of an idea, intervention, or alternative by observing outcomes to gather evidence or compare effects.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.