Rate Limits Overview

QNSP enforces rate limits at the edge gateway to protect platform stability.

Default Rate Limit

From apps/edge-gateway/src/config/env.ts:

Setting	Environment Variable	Default
Global rate limit	EDGE_RATE_LIMIT_RPS	5,000 RPS

How Rate Limiting Works

From apps/edge-gateway/src/routes/proxy.ts:

• Limits are per-tenant, per-route
• Algorithm: Token bucket
• Key format: tenant:<tenantId>:route:<routeId>

Response Headers

When rate limiting is applied, responses include rate limit headers:

X-RateLimit-Limit: 50
X-RateLimit-Remaining: 45
X-RateLimit-Reset: 1703424900

Header	Description
X-RateLimit-Limit	Configured RPS for this route
X-RateLimit-Remaining	Tokens remaining in bucket
X-RateLimit-Reset	Unix timestamp when bucket refills

Rate Limited Response

HTTP 429 Too Many Requests:

{
  "statusCode": 429,
  "error": "TOO_MANY_REQUESTS",
  "message": "Rate limit exceeded",
  "details": {
    "routeId": "<route-id>",
    "limit": 50,
    "retryAfter": 2
  }
}

The Retry-After header indicates seconds to wait before retrying.

Metrics

Rate limit events are tracked via telemetry:

// From edge-gateway/src/routes/proxy.ts
telemetry.metrics.blockedRequestsTotal.add(1, {
  tenant_id: effectiveTenantId,
  reason: "rate_limit_exceeded",
  route_id: matchingRoute.id,
});

Increasing Limits

Contact support for limit increases. See Limit Upgrades.