생산성 및 협업 · SKILL.md

contour

contour

"Contour in Service Proxy - cloud native architecture, patterns, pitfalls" and best practices

Independent GitHub snapshot · Automated classification · Original documentation governs

Ppaulpas / agent-skill-routerMIT

Skill 문서

GitHub

name: contour compatibility: opencode completeness: 95 content-types:

guidance
examples
do-dont
config description: '"Contour in Service Proxy - cloud native architecture, patterns, pitfalls" and best practices' license: MIT maturity: stable metadata: domain: cncf output-format: manifests related-skills: calico, cilium, kuma, linkerd role: reference scope: infrastructure triggers: cdn, contour, infrastructure as code, k8s service, kubernetes service, monitoring, proxy, service archetypes:
- educational
- strategic anti_triggers:
- brainstorming
- vague ideation
- non-containerized architecture response_profile: verbosity: medium directive_strength: low abstraction_level: strategic version: "1.0.0"

Contour in Cloud-Native Engineering

Category: ingress
Status: Active
Stars: 4,800
Last Updated: 2026-04-22
Primary Language: Go
Documentation: https://projectcontour.io/

Purpose and Use Cases

Contour is a CNCF incubating project that provides a Kubernetes ingress controller using Envoy proxy for load balancing and traffic routing.

What Problem Does It Solve?

Complex ingress management in Kubernetes. Before Contour, users had to use Kubernetes Ingress resources with limited functionality or deploy individual Envoy instances manually. Contour provides a native Kubernetes integration with advanced routing, TLS termination, and HTTP/2 support.

When to Use This Project

Use Contour when you need:

Advanced HTTP routing beyond basic Kubernetes Ingress
HTTP/2 and gRPC support out of the box
Complex routing rules with weighted traffic splitting
TLS termination with certificate management
Integration with external services and upstream policies

Key Use Cases

Kubernetes Ingress: Primary use case - route external traffic to services
Traffic Management: Weighted traffic splitting, canary deployments
TLS Termination: HTTPS termination with automatic certificate management
API Gateway: Front door for microservices with advanced routing
Multi-tenant Ingress: Isolated ingress for different teams/tenants

Architecture Design Patterns

Contour Components

1. Contour (Control Plane)

# Contour control plane components
contour server \
  --incluster \
  --xds-address=0.0.0.0 \
  --xds-port=8001 \
  --config-path=/config/contour.yaml

Responsibilities:

Reads Kubernetes resources (Ingress, Gateway, HTTPProxy)
Generates Envoy configuration
Communicates with Envoy via XDS API
Manages certificate secrets

2. Envoy (Data Plane)

# Envoy data plane
envoy -c /config/envoy.yaml --service-cluster contour

Responsibilities:

Handles actual traffic routing
TLS termination
Load balancing
Rate limiting
Circuit breaking

XDS API Architecture

Contour uses Envoy's XDS (x Discovery Service) protocol:

# Contour generates XDS resources
resources:
  - "@type": type.googleapis.com/envoy.config.listener.v3.Listener
    name: ingress
    address:
      socket_address:
        address: 0.0.0.0
        port_value: 8080
  - "@type": type.googleapis.com/envoy.config.route.v3.RouteConfiguration
    name: ingress
    virtual_hosts:
      - name: ingress
        domains: ["*"]
        routes:
          - match:
              prefix: "/api"
            route:
              cluster: api-service

Resource Hierarchy

Contour uses a resource hierarchy:

# HTTPProxy - Contour's custom resource
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: example
  namespace: default
spec:
  virtualhost:
    fqdn: example.com
    tls:
      secretName: example-tls
  routes:
    - conditions:
        - prefix: "/"
      services:
        - name: frontend
          port: 80
    - conditions:
        - prefix: "/api"
      services:
        - name: api
          port: 8080

Certificate Management

1. Standard Kubernetes Secrets

# TLS certificate from Kubernetes Secret
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: secure
spec:
  virtualhost:
    fqdn: secure.example.com
    tls:
      secretName: secure-example-com-tls  # Kubernetes Secret

2. ACME Certificate Management

# ACME certificate via cert-manager
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: acme
spec:
  virtualhost:
    fqdn: acme.example.com
    tls:
      secretName: acme-example-com-tls
      minimumProtocolVersion: "1.2"
      passthrough: false

Gateway API Integration

Contour supports the Gateway API for modern ingress management:

# Gateway API resource
apiVersion: gateway.networking.k8s.io/v1beta1
kind: GatewayClass
metadata:
  name: contour
spec:
  controllerName: projectcontour.io/contour-gateway-controller
  parametersRef:
    group: projectcontour.io
    kind: GatewayParameters
    name: contour-params
---
apiVersion: gateway.networking.k8s.io/v1beta1
kind: Gateway
metadata:
  name: example
spec:
  gatewayClassName: contour
  listeners:
    - name: http
      port: 80
      protocol: HTTP
    - name: https
      port: 443
      protocol: HTTPS
      tls:
        mode: Terminate
        certificateRefs:
          - name: example-tls

Upstream Policy

# Upstream policy for backend services
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: policy-example
spec:
  routes:
    - services:
        - name: backend
          port: 80
      policy:
        loadBalancer:
          policy: RingHash
          requestHashPolicies:
            - header:
                headerName: "x-request-id"
        retryPolicy:
          count: 3
          perTryTimeout: 2s
        connectionPool:
          tcp:
            maxConnections: 100
          http:
            h2ProtocolSettings:
              maxConcurrentStreams: 100

Integration Approaches

Kubernetes Native Integration

1. Ingress Resource

# Standard Kubernetes Ingress
apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  name: example
  annotations:
    kubernetes.io/ingress.class: contour
spec:
  rules:
    - host: example.com
      http:
        paths:
          - path: /
            pathType: Prefix
            backend:
              service:
                name: frontend
                port:
                  number: 80
          - path: /api
            pathType: Prefix
            backend:
              service:
                name: api
                port:
                  number: 8080

2. HTTPProxy Resource

# Contour's HTTPProxy with advanced features
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: advanced
spec:
  virtualhost:
    fqdn: advanced.example.com
    tls:
      secretName: advanced-tls
      minimumProtocolVersion: "1.2"
  routes:
    - conditions:
        - prefix: "/v1"
      services:
        - name: api-v1
          port: 8080
      timeoutPolicy:
        response: 30s
        idle: 5m
      retryPolicy:
        count: 3
        perTryTimeout: 2s
    - conditions:
        - prefix: "/v2"
      services:
        - name: api-v2
          port: 8080
      rewritePolicy:
        prefix: "/api"

3. Gateway API Resource

# Gateway API with Contour
apiVersion: gateway.networking.k8s.io/v1beta1
kind: Gateway
metadata:
  name: main-gateway
spec:
  gatewayClassName: contour
  listeners:
    - name: http
      port: 80
      protocol: HTTP
    - name: https
      port: 443
      protocol: HTTPS
      tls:
        mode: Terminate
        certificateRefs:
          - name: main-cert
---
apiVersion: gateway.networking.k8s.io/v1beta1
kind: HTTPRoute
metadata:
  name: main-route
spec:
  parentRefs:
    - name: main-gateway
  hostnames:
    - example.com
  rules:
    - matches:
        - path:
            type: Prefix
            value: /
      backendRefs:
        - name: frontend
          port: 80
    - matches:
        - path:
            type: Prefix
            value: /api
      backendRefs:
        - name: api
          port: 8080

External Services Integration

1. External Service Reference

# Reference external service
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: external
spec:
  routes:
    - conditions:
        - prefix: "/external"
      services:
        - name: external-service
          port: 443
          protocol: https
          urlScheme: https
      # External service load balancing
      loadBalancer:
        policy: RoundRobin

2. Weighted Services

# Traffic splitting between services
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: canary
spec:
  routes:
    - conditions:
        - prefix: "/"
      services:
        - name: primary
          port: 80
          weight: 90
        - name: canary
          port: 80
          weight: 10
      # Percent-based traffic splitting
      policy:
        loadBalancer:
          policy: RoundRobin

Load Balancer Integration

1. NLB Integration (AWS)

# AWS NLB with Contour
apiVersion: v1
kind: Service
metadata:
  name: contour-envoy
  namespace: projectcontour
  annotations:
    service.beta.kubernetes.io/aws-load-balancer-type: "external"
    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: "instance"
    service.beta.kubernetes.io/aws-load-balancer-cross-zone-load-balancing-enabled: "true"
spec:
  type: LoadBalancer
  ports:
    - name: http
      port: 80
      targetPort: 8080
    - name: https
      port: 443
      targetPort: 8443
  selector:
    app: contour-envoy

2. Service Mesh Integration

# Istio integration with Contour
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: mesh-integration
spec:
  routes:
    - conditions:
        - prefix: "/api"
      services:
        - name: api
          port: 8080
      # Service mesh integration
      policy:
        loadBalancer:
          policy: Maglev

Monitoring Integration

1. Prometheus Metrics

# Enable Prometheus metrics
apiVersion: v1
kind: ConfigMap
metadata:
  name: contour
  namespace: projectcontour
data:
  contour.yaml: |
    xds-address: "0.0.0.0"
    xds-port: 8001
    debug: true
    # Contour metrics
    metrics:
      address: "0.0.0.0"
      port: 8002

2. Tracing Integration

# Jaeger/Zipkin tracing
apiVersion: v1
kind: ConfigMap
metadata:
  name: contour
  namespace: projectcontour
data:
  contour.yaml: |
    xds-address: "0.0.0.0"
    xds-port: 8001
    tracing:
      type: zipkin
      service-name: contour
      sampling-rate: 0.0001
      config:
        collector-host: zipkin.observability.svc.cluster.local
        collector-port: 9411

Common Pitfalls and How to Avoid Them

1. TLS Configuration

Pitfall: Missing TLS configuration for HTTPS traffic.

# ❌ Incorrect - no TLS configuration
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: insecure
spec:
  routes:
    - services:
        - name: backend
          port: 8080

# ✅ Correct - with TLS configuration
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: secure
spec:
  virtualhost:
    fqdn: secure.example.com
    tls:
      secretName: secure-tls
      minimumProtocolVersion: "1.2"
  routes:
    - services:
        - name: backend
          port: 8080

2. Route Conditions

Pitfall: Incorrect route condition matching.

# ❌ Incorrect - conflicting routes
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: conflicting
spec:
  routes:
    - conditions:
        - prefix: "/"
      services:
        - name: default
          port: 80
    - conditions:
        - prefix: "/"  # Conflicts with first!
      services:
        - name: other
          port: 80

# ✅ Correct - specific routes
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: correct
spec:
  routes:
    - conditions:
        - prefix: "/api"
      services:
        - name: api
          port: 80
    - conditions:
        - prefix: "/web"
      services:
        - name: web
          port: 80

3. Service Discovery

Pitfall: Backend service not in same namespace.

# ❌ Incorrect - cross-namespace service reference
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: cross-namespace
  namespace: ingress
spec:
  routes:
    - services:
        - name: backend
          namespace: production  # Must specify namespace
          port: 8080

# ✅ Correct - explicit namespace
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: same-namespace
  namespace: default
spec:
  routes:
    - services:
        - name: backend
          port: 8080  # Same namespace as HTTPProxy

4. HTTP/2 Configuration

Pitfall: HTTP/2 not properly configured.

# ❌ Incorrect - HTTP/2 misconfiguration
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: http2-missing
spec:
  routes:
    - services:
        - name: backend
          port: 8080
      # Missing HTTP/2 configuration

# ✅ Correct - HTTP/2 enabled
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: http2-enabled
spec:
  routes:
    - services:
        - name: backend
          port: 8080
      # HTTP/2 automatically enabled for gRPC
      policy:
        loadBalancer:
          policy: Maglev

5. Certificate Expiration

Pitfall: TLS certificates not rotated.

# ❌ Incorrect - manual certificate management
# Certificates expire and cause outages

# ✅ Correct - automatic rotation
# Use cert-manager with ACME
apiVersion: cert-manager.io/v1
kind: Certificate
metadata:
  name: example-com
spec:
  secretName: example-com-tls
  duration: 2160h  # 90 days
  renewBefore: 360h  # 15 days before expiration
  commonName: example.com
  dnsNames:
    - example.com
    - "*.example.com"
  issuerRef:
    name: letsencrypt
    kind: ClusterIssuer

6. Timeout Configuration

Pitfall: Missing timeout configuration.

# ❌ Incorrect - no timeouts
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: no-timeouts
spec:
  routes:
    - services:
        - name: slow-backend
          port: 8080

# ✅ Correct - with timeouts
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: with-timeouts
spec:
  routes:
    - services:
        - name: slow-backend
          port: 8080
      timeoutPolicy:
        response: 30s
        idle: 5m

Coding Practices

HTTPProxy Templates

// HTTPProxy template generation
package ingress

import (
    projectcontouriov1 "github.com/projectcontour/contour/api/projectcontour/v1"
    corev1 "k8s.io/api/core/v1"
    metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
)

// GenerateHTTPProxy creates a standard HTTPProxy
func GenerateHTTPProxy(name, namespace, fqdn string, routes []Route) *projectcontouriov1.HTTPProxy {
    proxy := &projectcontouriov1.HTTPProxy{
        TypeMeta: metav1.TypeMeta{
            APIVersion: "projectcontour.io/v1",
            Kind:       "HTTPProxy",
        },
        ObjectMeta: metav1.ObjectMeta{
            Name:      name,
            Namespace: namespace,
            Labels: map[string]string{
                "app": name,
            },
        },
        Spec: projectcontouriov1.HTTPProxySpec{
            VirtualHost: &projectcontouriov1.VirtualHost{
                Fqdn: fqdn,
                TLS: &projectcontouriov1.TLS{
                    SecretName: fqdnToSecretName(fqdn),
                },
            },
            Routes: convertRoutes(routes),
        },
    }
    
    return proxy
}

// Route represents an HTTP route
type Route struct {
    Path       string
    Service    string
    Port       int32
    Timeout    string
    Retries    int
    Weight     int32
}

func convertRoutes(routes []Route) []projectcontouriov1.Route {
    var result []projectcontouriov1.Route
    
    for _, r := range routes {
        route := projectcontouriov1.Route{
            Conditions: []projectcontouriov1.Condition{
                {
                    Prefix: r.Path,
                },
            },
            Services: []projectcontouriov1.Service{
                {
                    Name: r.Service,
                    Port: r.Port,
                },
            },
        }
        
        if r.Timeout != "" {
            route.TimeoutPolicy = &projectcontouriov1.TimeoutPolicy{
                Response: r.Timeout,
            }
        }
        
        if r.Retries > 0 {
            route.RetryPolicy = &projectcontouriov1.RetryPolicy{
                Count: uint32(r.Retries),
            }
        }
        
        if r.Weight > 0 {
            route.Services[0].Weight = r.Weight
        }
        
        result = append(result, route)
    }
    
    return result
}

func fqdnToSecretName(fqdn string) string {
    // Convert FQDN to secret name
    // example.com -> example-com-tls
    return fqdn + "-tls"
}

Ingress Controller Integration

// Contour Ingress Controller
package controller

import (
    "context"
    "fmt"
    
    projectcontouriov1 "github.com/projectcontour/contour/api/projectcontour/v1"
    networkingv1 "k8s.io/api/networking/v1"
    apierrors "k8s.io/apimachinery/pkg/api/errors"
    "k8s.io/apimachinery/pkg/runtime"
    "k8s.io/apimachinery/pkg/watch"
    "k8s.io/client-go/tools/cache"
)

// IngressHandler handles Kubernetes Ingress resources
type IngressHandler struct {
    client       clientset.Interface
    proxyInformer cache.SharedIndexInformer
}

// HandleIngress converts Ingress to HTTPProxy
func (h *IngressHandler) HandleIngress(ing *networkingv1.Ingress) error {
    // Convert Ingress to HTTPProxy
    proxy := h.convertIngressToHTTPProxy(ing)
    
    // Create or update HTTPProxy
    _, err := h.client.ProjectcontourV1().HTTPProxies(ing.Namespace).Create(
        context.Background(),
        proxy,
        metav1.CreateOptions{},
    )
    
    if apierrors.IsAlreadyExists(err) {
        // Update existing HTTPProxy
        _, err = h.client.ProjectcontourV1().HTTPProxies(ing.Namespace).Update(
            context.Background(),
            proxy,
            metav1.UpdateOptions{},
        )
    }
    
    return err
}

func (h *IngressHandler) convertIngressToHTTPProxy(ing *networkingv1.Ingress) *projectcontouriov1.HTTPProxy {
    routes := make([]projectcontouriov1.Route, 0)
    
    for _, rule := range ing.Spec.Rules {
        if rule.IngressRuleValue.HTTP == nil {
            continue
        }
        
        for _, path := range rule.IngressRuleValue.HTTP.Paths {
            route := projectcontouriov1.Route{
                Conditions: []projectcontouriov1.Condition{
                    {
                        Prefix: path.Path,
                    },
                },
                Services: []projectcontouriov1.Service{
                    {
                        Name: path.Backend.Service.Name,
                        Port: path.Backend.Service.Port.Number,
                    },
                },
            }
            routes = append(routes, route)
        }
    }
    
    return &projectcontouriov1.HTTPProxy{
        TypeMeta: metav1.TypeMeta{
            APIVersion: "projectcontour.io/v1",
            Kind:       "HTTPProxy",
        },
        ObjectMeta: metav1.ObjectMeta{
            Name:      ing.Name + "-proxy",
            Namespace: ing.Namespace,
            Labels:    ing.Labels,
        },
        Spec: projectcontouriov1.HTTPProxySpec{
            Routes: routes,
        },
    }
}

Testing

// HTTPProxy testing
package ingress

import (
    "testing"
    
    projectcontouriov1 "github.com/projectcontour/contour/api/projectcontour/v1"
    corev1 "k8s.io/api/core/v1"
    metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
)

func TestGenerateHTTPProxy(t *testing.T) {
    routes := []Route{
        {
            Path:    "/api",
            Service: "api-service",
            Port:    8080,
            Timeout: "30s",
        },
        {
            Path:    "/",
            Service: "web-service",
            Port:    80,
        },
    }
    
    proxy := GenerateHTTPProxy("test", "default", "example.com", routes)
    
    // Validate HTTPProxy
    if proxy.Name != "test" {
        t.Errorf("expected name 'test', got '%s'", proxy.Name)
    }
    
    if proxy.Namespace != "default" {
        t.Errorf("expected namespace 'default', got '%s'", proxy.Namespace)
    }
    
    if proxy.Spec.VirtualHost.Fqdn != "example.com" {
        t.Errorf("expected FQDN 'example.com', got '%s'", proxy.Spec.VirtualHost.Fqdn)
    }
    
    if proxy.Spec.VirtualHost.TLS.SecretName != "example.com-tls" {
        t.Errorf("expected TLS secret 'example.com-tls', got '%s'", proxy.Spec.VirtualHost.TLS.SecretName)
    }
    
    if len(proxy.Spec.Routes) != 2 {
        t.Errorf("expected 2 routes, got %d", len(proxy.Spec.Routes))
    }
    
    // Validate routes
    if proxy.Spec.Routes[0].Conditions[0].Prefix != "/api" {
        t.Errorf("expected prefix '/api', got '%s'", proxy.Spec.Routes[0].Conditions[0].Prefix)
    }
    
    if proxy.Spec.Routes[0].Services[0].Name != "api-service" {
        t.Errorf("expected service 'api-service', got '%s'", proxy.Spec.Routes[0].Services[0].Name)
    }
}

Fundamentals

Contour Architecture

1. Control Plane

Contour: Reads Kubernetes resources and generates Envoy config
XDS API: Communicates with Envoy data plane
Certificate Manager: Manages TLS certificates

2. Data Plane

Envoy: Handles traffic routing and load balancing
XDS Client: Connects to Contour control plane
Listeners: HTTP/HTTPS listeners on configured ports

Contour Deployment

# Contour deployment
kubectl apply -f https://projectcontour.io/quickstart/contour.yaml

# Verify deployment
kubectl -n projectcontour get pods

# Check Envoy pods
kubectl -n projectcontour get pods -l app=contour-envoy

Resource Types

1. HTTPProxy

apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: example
spec:
  virtualhost:
    fqdn: example.com
  routes:
    - services:
        - name: service
          port: 80

2. TLSService

apiVersion: projectcontour.io/v1
kind: TLSService
metadata:
  name: tls-service
spec:
  secretName: tls-secret
  minProtocolVersion: "1.2"

3. GatewayClass

apiVersion: gateway.networking.k8s.io/v1beta1
kind: GatewayClass
metadata:
  name: contour
spec:
  controllerName: projectcontour.io/contour-gateway-controller

Configuration Options

1. Contour Config

# contour.yaml
xds-address: "0.0.0.0"
xds-port: 8001
debug: true
metrics:
  address: "0.0.0.0"
  port: 8002
tracing:
  type: zipkin
  service-name: contour
  sampling-rate: 0.0001

2. Envoy Config

# Envoy bootstrap config
static_resources:
  listeners:
    - name: ingress
      address:
        socket_address:
          address: 0.0.0.0
          port_value: 8080
  clusters:
    - name: xds_cluster
      connect_timeout: 30s
      type: STRICT_DNS
      lb_policy: ROUND_ROBIN
      load_assignment:
        cluster_name: xds_cluster
        endpoints:
          - lb_endpoints:
              - endpoint:
                  address:
                    socket_address:
                      address: contour
                      port_value: 8001

Scaling and Deployment Patterns

High Availability

# HA Contour deployment
# Multiple Contour instances with leader election
kubectl apply -f https://projectcontour.io/quickstart/contour-ha.yaml

# Scale Envoy pods
kubectl -n projectcontour scale deployment contour-envoy --replicas=3

Auto-scaling

# Horizontal Pod Autoscaler for Envoy
apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
metadata:
  name: contour-envoy
  namespace: projectcontour
spec:
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: contour-envoy
  minReplicas: 2
  maxReplicas: 10
  metrics:
    - type: Resource
      resource:
        name: cpu
        target:
          type: Utilization
          averageUtilization: 70

Load Balancing Strategies

# Load balancing in HTTPProxy
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: lb-example
spec:
  routes:
    - services:
        - name: backend
          port: 8080
      policy:
        loadBalancer:
          policy: RoundRobin  # Random, RingHash,Maglev, WeightedLeastRequest

Rate Limiting

# Rate limiting configuration
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: rate-limit
spec:
  routes:
    - services:
        - name: api
          port: 8080
      policy:
        ratelimit:
          local:
            requests: 100
            period: 1s

Circuit Breaking

# Circuit breaking configuration
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: circuit-breaker
spec:
  routes:
    - services:
        - name: backend
          port: 8080
      policy:
        connectionPool:
          tcp:
            maxConnections: 100
          http:
            h2ProtocolSettings:
              maxConcurrentStreams: 100
            requestTimeout: 30s
            idleTimeout: 5m
        outlierDetection:
          consecutive5xxErrors: 5
          interval: 30s
          baseEjectionTime: 30s
          maxEjectionPercent: 50

Additional Resources

Official Documentation

Project Contour - Project website
Contour Documentation - Complete documentation
Contour GitHub - Source code
Contour API Reference - API documentation

Implementations

Kubernetes Ingress Contour - Kubernetes integration
Contour Operator - Operator for Contour
Contour Helm Chart - Helm installation

Community Resources

CNCF Contour - CNCF project page
Contour Slack - Community discussion
Contour Mailing List - Announcements

Learning Resources

Contour Getting Started - Tutorial
Contour Examples - Example configurations
Contour Webinars - Video content

This SKILL.md file was verified and last updated on 2026-04-22. Content based on CNCF Contour project and production usage patterns.

Tutorial

This tutorial covers installation, configuration, and basic usage of Contour for Kubernetes ingress management.

Prerequisites

Before starting with Contour, ensure you have:

A running Kubernetes cluster (1.23+)
kubectl configured with cluster admin access
Basic understanding of Kubernetes services and ingress concepts
Helm 3.x (if using Helm installation)

Installation

Method 1: Using kubectl apply (Quick Start)

# Download and apply the latest Contour manifest
curl -sL https://projectcontour.io/quickstart/contour.yaml > contour.yaml

# Apply the manifest
kubectl apply -f contour.yaml

# Verify the installation
kubectl -n projectcontour get pods

# Expected output:
# NAME                      READY   STATUS    RESTARTS   AGE
# contour-7d8f9c6b9-abc12   1/1     Running   0          2m
# contour-7d8f9c6b9-def34   1/1     Running   0          2m
# contour-envoy-xyz56       1/1     Running   0          2m

Method 2: Using Helm (Production Recommended)

# Add the Contour Helm repository
helm repo add projectcontour https://projectcontour.github.io/charts
helm repo update

# Install Contour
helm install contour projectcontour/contour \
  --create-namespace \
  --namespace projectcontour \
  --set contour.config.filePath=/config/contour.yaml

# Verify installation
helm -n projectcontour list

Method 3: Custom Configuration

# Create namespace
kubectl create namespace projectcontour

# Apply Contour with custom config
kubectl apply -f contour.yaml

# Create custom ConfigMap for Contour
kubectl create configmap contour-config \
  --from-file=contour.yaml=./my-contour-config.yaml \
  -n projectcontour

# Patch deployment to use custom config
kubectl patch deployment contour \
  --namespace projectcontour \
  --type='json' \
  -p='[{"op": "add", "path": "/spec/template/spec/volumes/-", "value": {"name": "config", "configMap": {"name": "contour-config"}}}]'

Basic Configuration

Contour is configured via a YAML file and Kubernetes ConfigMap.

Contour Config File

Create contour.yaml:

# contour.yaml - Contour main configuration
xds-address: "0.0.0.0"
xds-port: 8001
debug:
  address: "0.0.0.0"
  port: 8000
metrics:
  address: "0.0.0.0"
  port: 8002
tracing:
  type: zipkin
  service-name: contour
  sampling-rate: 0.0001
  config:
    collector-host: zipkin.observability.svc.cluster.local
    collector-port: 9411
resources:
  requests:
    memory: "256Mi"
    cpu: "100m"
  limits:
    memory: "512Mi"
    cpu: "500m"

Enable Gateway API Support

# contour.yaml - Enable Gateway API
gateway:
  apiVersion: gateway.networking.k8s.io/v1beta1
  enabled: true

Configure TLS Settings

# contour.yaml - TLS configuration
tls:
  minimumProtocolVersion: "1.2"
  cipherSuites:
    - TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256
    - TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384

Usage Examples

Creating Your First HTTPProxy

# hello-world.yaml
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: hello-world
  namespace: default
spec:
  virtualhost:
    fqdn: hello-world.example.com
    tls:
      secretName: hello-world-tls
      minimumProtocolVersion: "1.2"
  routes:
    - conditions:
        - prefix: "/"
      services:
        - name: hello
          port: 80

Apply the configuration:

# Create a simple hello service first
kubectl create deployment hello --image=nginx --port=80
kubectl expose deployment hello --port=80 --target-port=80 --type=ClusterIP

# Apply HTTPProxy
kubectl apply -f hello-world.yaml

# Check status
kubectl get httpproxy hello-world -o yaml

Configure Path-Based Routing

# multi-path.yaml
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: multi-path
  namespace: default
spec:
  virtualhost:
    fqdn: app.example.com
    tls:
      secretName: app-tls
  routes:
    - conditions:
        - prefix: "/api"
      services:
        - name: api-service
          port: 8080
      timeoutPolicy:
        response: 30s
        idle: 5m
    - conditions:
        - prefix: "/web"
      services:
        - name: web-service
          port: 80
    - conditions:
        - prefix: "/admin"
      services:
        - name: admin-service
          port: 8080

Enable gRPC Support

# grpc-service.yaml
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: grpc-service
  namespace: default
spec:
  virtualhost:
    fqdn: grpc.example.com
    tls:
      secretName: grpc-tls
  routes:
    - conditions:
        - prefix: "/"
      services:
        - name: grpc-backend
          port: 50051
        # gRPC is automatically enabled when using HTTP/2

Common Operations

Checking Contour Status

# Check Contour pod status
kubectl -n projectcontour get pods -l app=contour

# Check Envoy pod status
kubectl -n projectcontour get pods -l app=contour-envoy

# View Contour logs
kubectl -n projectcontour logs -l app=contour -f

# View Envoy logs
kubectl -n projectcontour logs -l app=contour-envoy -f

Debugging Configuration Issues

# Validate HTTPProxy syntax
kubectl apply -f my-proxy.yaml --dry-run=client

# Check Contour XDS status
kubectl -n projectcontour exec -l app=contour -- contour status

# Check Envoy configuration
kubectl -n projectcontour exec -l app=contour-envoy -- envoy config dump

Monitoring with Prometheus

# Enable Prometheus scraping
apiVersion: v1
kind: ConfigMap
metadata:
  name: contour
  namespace: projectcontour
data:
  contour.yaml: |
    xds-address: "0.0.0.0"
    xds-port: 8001
    debug: true
    metrics:
      address: "0.0.0.0"
      port: 8002
---
# Prometheus scrape configuration
apiVersion: monitoring.coreos.com/v1
kind: ServiceMonitor
metadata:
  name: contour
  namespace: projectcontour
  labels:
    release: prometheus
spec:
  selector:
    matchLabels:
      app: contour
  endpoints:
    - port: metrics
      path: /metrics
      interval: 30s

Viewing TLS Certificate Status

# List certificates in secret
kubectl get secret -n projectcontour contour-envoy -o jsonpath='{.data.tls\.crt}' | base64 -d | openssl x509 -text -noout

# Check certificate expiration
kubectl get secret -n projectcontour contour-envoy -o jsonpath='{.data.tls\.crt}' | base64 -d | openssl x509 -enddate -noout

# View Contour TLS configuration
kubectl -n projectcontour get secret contour-envoy -o yaml

Traffic Management Examples

# Canary deployment with weighted routing
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: canary
  namespace: default
spec:
  virtualhost:
    fqdn: myapp.example.com
    tls:
      secretName: myapp-tls
  routes:
    - conditions:
        - prefix: "/"
      services:
        - name: myapp-primary
          port: 80
          weight: 90
        - name: myapp-canary
          port: 80
          weight: 10

# Retry configuration
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: with-retries
  namespace: default
spec:
  routes:
    - conditions:
        - prefix: "/api"
      services:
        - name: api
          port: 8080
      retryPolicy:
        count: 3
        perTryTimeout: 2s
        retryOn:
          - 5xx
          - reset
          - connect-failure

Best Practices for Contour

1. Use Namespace Isolation

# HTTPProxy in different namespaces
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: frontend
  namespace: web
spec:
  virtualhost:
    fqdn: frontend.example.com
    tls:
      secretName: frontend-tls
  routes:
    - services:
        - name: frontend
          port: 80
---
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: backend
  namespace: api
spec:
  virtualhost:
    fqdn: api.example.com
    tls:
      secretName: api-tls
  routes:
    - services:
        - name: backend
          port: 8080

2. Implement Timeout and Retry Policies

# Production-ready timeout and retry configuration
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: production
  namespace: default
spec:
  routes:
    - conditions:
        - prefix: "/"
      services:
        - name: backend
          port: 8080
      timeoutPolicy:
        response: 30s           # Response timeout
        idle: 5m                # Idle timeout
        reconnect: 3            # Number of retries
        perTryTimeout: 5s       # Per-try timeout
      retryPolicy:
        count: 3                # Retry count
        perTryTimeout: 2s       # Per-try timeout
        retryOn:
          - 5xx                 # Retry on 5xx errors
          - reset               # Retry on connection reset
          - connect-failure     # Retry on connection failures

3. Use Resource Limits

# Contour deployment with resource limits
apiVersion: apps/v1
kind: Deployment
metadata:
  name: contour
  namespace: projectcontour
spec:
  template:
    spec:
      containers:
        - name: contour
          resources:
            requests:
              memory: "256Mi"
              cpu: "100m"
            limits:
              memory: "512Mi"
              cpu: "500m"

4. Enable Access Logging

# Enable Envoy access logging
apiVersion: v1
kind: ConfigMap
metadata:
  name: contour
  namespace: projectcontour
data:
  contour.yaml: |
    xds-address: "0.0.0.0"
    xds-port: 8001
    debug: true
    accessLog:
      - type: file
        path: /var/log/contour/envoy.log
        format: "%LOCAL_PORT% %PROTOCOL% %STATUS% %STATUS_CODE% %RESPONSE_FLAGS% %UPSTREAM_HOST% %UPSTREAM_CLUSTER% %UPSTREAM_SERVICE_TIME% %REQUEST_SIZE% %RESPONSE_SIZE% %DOWNSTREAM_REMOTE_ADDRESS% %DOWNSTREAM_LOCAL_ADDRESS% %REQUEST_METHOD% %REQUEST_PATH% %REQUEST_HEADERS% %RESPONSE_HEADERS% %UPSTREAM_STREAM_RESPONSE_FLAGS%"

5. Certificate Management with cert-manager

# Certificate from cert-manager
apiVersion: cert-manager.io/v1
kind: Certificate
metadata:
  name: example-com
  namespace: default
spec:
  secretName: example-com-tls
  duration: 2160h  # 90 days
  renewBefore: 360h  # 15 days before expiration
  commonName: example.com
  dnsNames:
    - example.com
    - "*.example.com"
  issuerRef:
    name: letsencrypt-prod
    kind: ClusterIssuer
---
# Reference in HTTPProxy
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: secure
  namespace: default
spec:
  virtualhost:
    fqdn: example.com
    tls:
      secretName: example-com-tls
      minimumProtocolVersion: "1.2"

6. Health Check Endpoints

# Contour health endpoint
curl http://localhost:8000/healthz

# Contour metrics endpoint
curl http://localhost:8002/metrics

# Check Contour status
kubectl -n projectcontour exec -l app=contour -- contour status

7. Circuit Breaking Configuration

# Circuit breaker for backend services
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: circuit-breaker
  namespace: default
spec:
  routes:
    - conditions:
        - prefix: "/api"
      services:
        - name: api
          port: 8080
      policy:
        loadBalancer:
          policy: RingHash
          requestHashPolicies:
            - header:
                headerName: "x-request-id"
        connectionPool:
          tcp:
            maxConnections: 100
          http:
            h2ProtocolSettings:
              maxConcurrentStreams: 100
            requestTimeout: 30s
            idleTimeout: 5m
        outlierDetection:
          consecutive5xxErrors: 5
          interval: 30s
          baseEjectionTime: 30s
          maxEjectionPercent: 50

This SKILL.md file was verified and last updated on 2026-04-22. Content based on CNCF Contour project and production usage patterns.

Examples

Basic HTTPProxy Configuration

# Basic HTTPProxy for a single service
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: simple
  namespace: default
spec:
  virtualhost:
    fqdn: simple.example.com
    tls:
      secretName: simple-tls
      minimumProtocolVersion: "1.2"
  routes:
    - conditions:
        - prefix: "/"
      services:
        - name: my-service
          port: 80

TLS Passthrough

# TLS passthrough configuration
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: tls-passthrough
  namespace: default
spec:
  virtualhost:
    fqdn: secure.example.com
    tls:
      passthrough: true
  routes:
    - conditions:
        - prefix: "/"
      services:
        - name: my-service
          port: 443

Weighted Traffic Splitting

# Traffic splitting between versions
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: canary
  namespace: default
spec:
  virtualhost:
    fqdn: app.example.com
    tls:
      secretName: app-tls
  routes:
    - conditions:
        - prefix: "/"
      services:
        - name: app-v1
          port: 80
          weight: 95
        - name: app-v2
          port: 80
          weight: 5

External Service

# Route to external service
apiVersion: projectcontour.io/v1
kind: HTTPProxy
metadata:
  name: external
  namespace: default
spec:
  routes:
    - conditions:
        - prefix: "/external"
      services:
        - name: external-service
          port: 443
          protocol: https
          urlScheme: https
      policy:
        loadBalancer:
          policy: RoundRobin

When to Use

Use this skill when:

Integrating a CNCF project into Kubernetes infrastructure — You need to configure, deploy, or troubleshoot a cloud-native tool within a cluster
Designing cloud-native architecture — You are selecting and integrating CNCF tools to solve specific infrastructure challenges
Resolving operational issues — A CNCF component is misbehaving, underperforming, or needs configuration changes

Core Workflow

Assess Requirements — Understand the use case, scale, integration needs, and existing infrastructure. Checkpoint: Document requirements, constraints, and success criteria.
Design Architecture — Plan component interactions, data flow, and deployment strategy using cloud-native best practices. Checkpoint: Verify the architecture addresses all requirements and follows CNCF conventions.
Implement & Configure — Create manifests, configurations, and deployment scripts. Include resource limits, health checks, and observability hooks. Checkpoint: Validate all YAML against schema and test in a staging environment.
Deploy & Monitor — Apply manifests to the cluster, verify component health, and confirm observability is working. Checkpoint: Confirm all pods/services are running, probes passing, and metrics/alerts configured.

Constraints

MUST DO

Include at least one complete working YAML manifest example
Note when content is auto-generated vs. manually verified
Reference relevant CNCF project documentation

MUST NOT DO

Deploy manifests without testing in a staging environment first
Use deprecated API versions (e.g., apps/v1beta1)
Omit resource limits and requests in Kubernetes manifests

MAKE IT PART OF YOUR WORKFLOW

Skill 설치

TERMINAL

npx skills add paulpas/agent-skill-router --skill contour

터미널에서 명령을 실행하세요. 설치 전에 코드와 라이선스를 검토하세요.

ZIP · Complete skill directory with resources and licenses

SKILL.md · Instructions file only

리포지토리 스타: Not synced
라이선스: MIT
버전 / 브랜치: HEAD · d2e2ac6f
리포지토리 업데이트: Not synced

설치하기 전에 스크립트와 권한을 검토하세요.

소스 코드