Gloo Edge 1.0: Production-Ready API Gateway

Introduction

On November 14, 2019, Solo.io announced Gloo Edge 1.0, marking the first production-ready release of the Envoy-based API gateway. Originally open-sourced in March 2018, Gloo Edge 1.0 delivers enterprise-grade features for routing, security, and observability.

If your cluster has outgrown “Ingress rules + annotations”, gateways like Gloo typically show up when you need consistent routing policy, strong auth, and real troubleshooting signals across dozens (or hundreds) of services.

Production-Ready Features

• Enterprise stability: Predictable behavior under load and clearer day-2 workflows (upgrades, rollback, drift management).
• Advanced routing: Header/path/query matching for canaries, tenant routing, and API versioning.
• Function discovery: Route to serverless functions, microservices, and legacy applications without hand-wiring every upstream.
• Multi-protocol support: HTTP, gRPC, and WebSocket traffic in one gateway model.

Envoy Integration

1. Envoy Proxy foundation leverages Envoy’s high-performance data plane for request handling.
2. xDS API integration provides dynamic configuration updates without restarts.
3. Filter chain support enables custom request/response processing with Envoy filters.
4. Performance optimizations deliver low latency and high throughput for API traffic.

Kubernetes Integration

• Native Kubernetes support provides seamless integration with Kubernetes services and ingresses.
• CRD-based configuration enables managing routes and virtual services as Kubernetes resources.
• Service discovery automatically discovers Kubernetes services and endpoints.
• RBAC integration provides fine-grained access control for gateway configuration.

Practical tip: treat gateway config like application code — code review, staged rollout, and a fast rollback path. “Small YAML tweaks” at the edge can become production outages.

Security Features

• TLS termination provides secure HTTPS connections with automatic certificate management.
• Authentication support includes OAuth2, JWT, and API key authentication.
• Authorization policies enable fine-grained access control based on user identity and roles.
• Rate limiting protects backends from traffic spikes and abuse.

Two common edge cases to plan for:

• Retry + rate limit interactions: retries can amplify load and unexpectedly trip limits unless budgets are coordinated.
• Multi-tenant clusters: you’ll want clear ownership boundaries for routes so one team can’t accidentally break another’s APIs.

Observability & Monitoring

• Metrics integration exposes detailed gateway metrics for Prometheus and Grafana.
• Distributed tracing support enables correlation of requests across services with OpenTracing.
• Access logs provide detailed request/response logging for debugging and audit.
• Health checks monitor backend service health and route traffic accordingly.

Getting Started

helm repo add gloo https://storage.googleapis.com/solo-public-helm
helm install gloo gloo/gloo --namespace gloo-system --create-namespace

Create a VirtualService:

apiVersion: gateway.solo.io/v1
kind: VirtualService
metadata:
  name: my-app
  namespace: gloo-system
spec:
  virtualHost:
    domains:
    - my-app.example.com
    routes:
    - matchers:
      - prefix: /
      routeAction:
        single:
          upstream:
            name: my-service
            namespace: gloo-system

Summary

Gloo Edge 1.0 is the kind of release that matters once your “edge” becomes a shared platform surface: it moves routing, security, and visibility into a more explicit gateway model that scales better than per-service Ingress tweaks.