Argo Rollouts

Argo Rollouts
Name	Argo Rollouts
Developer	Intuit
Initial release	2019
Programming language	Go
Operating system	Linux, macOS, Windows
License	Apache-2.0

Argo Rollouts is a Kubernetes controller and set of Custom Resource Definitions for progressive delivery, providing advanced deployment strategies such as Canary and Blue-Green for cloud-native applications. It is developed by contributors at Intuit and an open-source community, designed to integrate with observability, CI/CD, and service mesh projects to automate release promotion and metrics-driven analysis. The project complements container orchestration projects and GitOps workflows in modern infrastructure stacks.

Overview

Argo Rollouts originated within the ecosystem driven by Intuit and contributors aligned with projects like Kubernetes, Cloud Native Computing Foundation, Linux Foundation, Docker, and CNCF Sandbox. It fills a niche alongside projects such as Argo CD, Flux, Jenkins X, Spinnaker, and Helm by providing fine-grained traffic management and automated promotion based on metrics from systems like Prometheus, Datadog, New Relic, and Grafana. The project’s development and governance intersect with organizations including Red Hat, Google, Amazon Web Services, Microsoft, and HashiCorp through integrations, contributions, and ecosystem support.

Features and Functionality

Argo Rollouts implements capabilities familiar to operators working with Istio, Linkerd, Envoy, NGINX, and Traefik. It exposes CRDs that allow declarative definitions comparable to resources in Kubernetes API and works with deployment strategies popularized by platforms such as Spinnaker and patterns from Martin Fowler. Key features include automated analysis leveraging metrics from Prometheus, Datadog, New Relic, and Stackdriver; traffic shifting integrations with Istio, Linkerd, Envoy, and NGINX Ingress Controller; and experiments and canary analysis inspired by practices used at Netflix, LinkedIn, Facebook, Google Cloud, and GitHub. Rollout lifecycle management aligns with observability practices used by Splunk, Elastic (company), Sentry, and Honeycomb.

Architecture and Components

The controller-oriented architecture is implemented in Go (programming language) and operates as a control plane component within Kubernetes clusters, interacting with objects similar to those managed by kubectl, kube-apiserver, kube-controller-manager, and kube-scheduler. Core components include the Rollout CRD, analysis templates, experiment controllers, and dashboard integrations comparable to interfaces offered by Grafana, Kiali, and Lens. Traffic routing relies on resources from Istio, Linkerd, Envoy, NGINX Ingress Controller, and Contour while metrics collection uses adapters to systems like Prometheus, Datadog, New Relic, and AWS CloudWatch. The project’s continuous integration and delivery touchpoints mirror patterns from Jenkins, GitLab CI/CD, GitHub Actions, and CircleCI.

Deployment Strategies

Supported strategies include Canary and Blue-Green releases, which have precedents in large-scale deployments at Netflix, Amazon, Google, Facebook, and Spotify. Canary workflows allow incremental traffic shifts with analysis windows, leveraging analysis templates that query Prometheus, Datadog, New Relic, and InfluxData backends. Blue-Green operations can be coordinated alongside service mesh approaches from Istio and Linkerd or ingress solutions such as NGINX and HAProxy. Advanced patterns include automated rollback on failing analysis, progressive traffic routing like the techniques used in Canary Deployment case studies at Microsoft Azure and Google Cloud Platform, and multi-environment promotion strategies akin to GitOps flows employed by Weaveworks and Flux.

Usage and Configuration

Users interact with the controller via annotations and Rollout CRs applied with tools like kubectl, helm, kustomize, Argo CD, Flux, and GitHub Actions. Configuration commonly references metric providers such as Prometheus, Datadog, New Relic, and AWS CloudWatch, and service mesh routing objects from Istio, Linkerd, Envoy, and NGINX Ingress Controller. CI/CD pipelines integrating Argo Rollouts often run in environments managed by Jenkins, GitLab, CircleCI, Travis CI, and GitHub Actions and coordinate with artifact registries like Docker Hub, Google Container Registry, Amazon ECR, and Harbor. Visualization and manual promotion are available through dashboards integrated with Grafana, Kiali, Lens, and the Kubernetes Dashboard.

Integrations and Ecosystem

Argo Rollouts is part of a broader ecosystem featuring tools and projects such as Argo CD, Flux, Helm, Kustomize, Prometheus, Grafana, Istio, Linkerd, Envoy, NGINX, Spinnaker, Jenkins, GitHub Actions, GitLab, CircleCI, HashiCorp Consul, Consul Connect, Fluentd, Loki (software), Elastic Stack, Sentry, and Honeycomb. Commercial cloud integrations include Amazon Web Services, Google Cloud Platform, Microsoft Azure, DigitalOcean, and IBM Cloud. The community contributions and adopters include enterprises like Intuit, Red Hat, VMware, Salesforce, and Pinterest.

Security and Governance

Security considerations parallel those in Kubernetes and service mesh projects such as Istio and Linkerd, including RBAC policies using Open Policy Agent, OPA Gatekeeper, and Kubernetes RBAC, TLS configurations aligned with SPIFFE and cert-manager, and supply chain protections via in-toto, Sigstore, and Notary. Governance of the project follows open-source contribution models seen in Linux Foundation projects and collaborative development practices used by CNCF-hosted communities. Operational security practices integrate with observability platforms like Prometheus, Grafana, Splunk, and Elastic (company) for alerting, auditing, and incident response workflows similar to processes at PagerDuty, VictorOps, and Sentry.

Category:Kubernetes