Ambassador · Rate Limits
Ambassador Rate Limits
Ambassador Edge Stack and Emissary-Ingress are self-hosted Kubernetes ingress controllers, so request-rate behavior is determined by the operator's own rate-limit policies (RateLimitService, RateLimit CRDs) configured against the bundled Envoy data plane. Edge Stack ships with a built-in rate-limit service supporting per-route, per-IP, and per-header limits; Emissary integrates with external rate-limit services. There are no public per-tier API throttle numbers because the gateway runs in the customer's cluster.
2 Limits
Throttle: 429
Rate LimitingAPI GatewayKubernetesAmbassadorGravitee
Limits
Self-hosted Envoy rate limits customer-configured
configured by RateLimit CRDs in the customer cluster
Edge Stack and Emissary-Ingress provide rate-limit primitives via Envoy; throughput is bounded by the customer's data-plane resources (replicas, CPU, memory).
Edge Stack RateLimitService customer-configured
configured per RateLimit CRD
Per-route, per-source-IP, per-header limits; supports time windows of seconds, minutes, hours.
Policies
Self-hosted operator responsibility
Customer operates the gateway in their own cluster, so capacity and rate-limit values are controlled by their CRD configuration, not by the vendor.
429 + Retry-After
Edge Stack RateLimitService returns HTTP 429 when configured thresholds are exceeded; include Retry-After in the response header per RFC 6585.
Sidecar scaling
Throughput scales with the number of Ambassador/Envoy replicas; horizontally scale the gateway deployment to absorb additional load.
External RLS
Emissary-Ingress integrates with external Envoy-compatible RateLimitService implementations for advanced policy.