Open Source • Cloud Native

The Future ofAPI Gateways

A powerful, cloud-native API Gateway built on NGINX with comprehensive features for modern microservices architectures. Deploy anywhere. Scale infinitely. Secure everything.

Enterprise Security
High Performance
Multi-Cloud Ready
10K+
Requests/sec
<10ms
Latency
99.9%
Uptime
Deploy on:
EKS
Amazon EKS
GKE
Google GKE
AKS
Azure AKS
K8s
Kubernetes

Comprehensive Features

Everything you need for a modern API Gateway, built with enterprise-grade security and performance in mind

Core Capabilities

High-Performance Proxy

NGINX-based reverse proxy with Lua scripting for maximum performance

Advanced Routing

Path-based, header-based, and weighted routing with intelligent load balancing

Protocol Support

HTTP/HTTPS, WebSocket, gRPC, and HTTP/2 support out of the box

Multi-Tenant

Team-based configuration via Custom Resource Definitions

Security & Authentication

JWT Validation

Built-in JWT token validation and claims extraction

OAuth2/OIDC Integration

Integration with Keycloak and OAuth2-Proxy for enterprise auth

API Key Management

Header, query parameter, and custom API key validation

mTLS & WAF Protection

Mutual TLS and ModSecurity integration for comprehensive security

Traffic Management

Rate Limiting

Advanced rate limiting with Redis backend and burst handling

Load Balancing

Multiple algorithms: round-robin, least-connections, IP hash

Circuit Breaker

Failure detection and automatic failover mechanisms

Request Transformation

Header manipulation, body transformation, and response caching

Developer Experience

GitOps Ready

Full declarative configuration via Kubernetes CRDs

Plugin System

Lua, Python, and WASM plugin support for custom logic

API Documentation

Auto-generated API documentation and developer portal

Cloud Agnostic

Deploy on any cloud provider or on-premises infrastructure

Ready to Experience the Power?

Deploy Cumulus API Gateway in under 5 minutes with our automated scripts. Start with our local development setup or go straight to production.

Cloud-Native Architecture

Built for Kubernetes with microservices-first design. Scale horizontally, deploy anywhere, monitor everything.

External Traffic (HTTPS)
Kubernetes Cluster

Ingress Layer

Load Balancer

External LB (GCP/AWS/Azure)

Ingress Controller

NGINX/Traefik Ingress

Cumulus Gateway Namespace

NGINX Gateway
• Deployment: 3 replicas
• HPA: 3-10 pods
• Service: ClusterIP
OAuth2-Proxy
• Deployment: 2 replicas
• ConfigMap: Auth config
• Secret: OAuth credentials
Redis Cache
• StatefulSet: 3 replicas
• PVC: 10Gi storage
• Service: Headless
Open Policy Agent
• DaemonSet: All nodes
• ConfigMap: Policy rules
Keycloak (Optional)
• Deployment: 2 replicas
• PostgreSQL backend

Microservices Namespaces

Service A
• Namespace: service-a
• Deployment: 5 replicas
• Ports: 8080, 9090
Service B
• Namespace: service-b
• Deployment: 3 replicas
• Ports: 8080, 9090
Database
• StatefulSet: 3 replicas
• PVC: 100Gi each
• Service: Headless

Observability Stack

Prometheus
• Metrics collection
• ServiceMonitor CRDs
• Alerting rules
Jaeger
• Distributed tracing
• Span collection
• UI dashboard
Grafana
• Visualization
• Custom dashboards
• Alerting

Traffic Flow & Security

1. External traffic hits Load Balancer
2. Ingress Controller routes to Gateway
3. OAuth2-Proxy validates authentication
4. OPA enforces authorization policies
5. Redis provides caching & rate limiting
6. NGINX proxies to target services

Technology Stack

Proxy EngineNGINX + Lua

High-performance reverse proxy

AuthenticationKeycloak + OAuth2-Proxy

Identity and access management

AuthorizationOpen Policy Agent

Policy-based access control

Rate LimitingRedis + Lua

Distributed rate limiting

MetricsPrometheus

Metrics collection and alerting

TracingJaeger

Distributed request tracing

ConfigurationKubernetes CRDs

Declarative API management

Architecture Benefits

Cloud Agnostic

Deploy on any Kubernetes cluster - GCP, AWS, Azure, or on-premises

Horizontally Scalable

Auto-scaling with HPA based on CPU, memory, or custom metrics

High Availability

Multi-zone deployment with automatic failover and health checks

Observable

Full observability with metrics, logs, traces, and alerting

See It In Action

Explore our interactive architecture diagram and deployment examples

Deploy Anywhere

Multi-cloud deployment options with automated setup scripts. From local development to production-ready clusters.

Available

Google Cloud Platform

Deploy on GKE with our optimized Helm charts and Terraform modules

Auto-scaling with GKE Autopilot
Cloud Load Balancer integration
Cloud Armor for DDoS protection
Cloud Operations monitoring
~10 minutes
Deploy Time
$160/month
Est. Cost
Coming Soon

Amazon Web Services

EKS deployment with AWS Application Load Balancer and CloudWatch

EKS Fargate for serverless
ALB ingress controller
AWS WAF integration
CloudWatch metrics & logs
~
Deploy Time
$
Est. Cost
Coming Soon

Microsoft Azure

AKS deployment with Azure Application Gateway and Monitor

AKS virtual nodes
Application Gateway ingress
Azure Security Center
Azure Monitor integration
~
Deploy Time
$
Est. Cost
Available

Local Development

Complete local setup with k3d, Docker Desktop, or kind clusters

One-command deployment
Mock services included
Local testing tools
Hot reload development
~2 minutes
Deploy Time
Free
Est. Cost

Quick Start Guide

1

Clone Repository

git clone https://github.com/fxellence-dev/cumulus-api-gateway.git

Get the latest source code

2

Run Setup Script

./scripts/quick-start-local.sh

Automated local deployment

3

Test Endpoints

curl http://localhost:30080/health

Verify deployment success

Additional Resources

Installation Guide

Step-by-step deployment instructions

Configuration Reference

Complete API and configuration docs

Best Practices

Production deployment guidelines

Troubleshooting

Common issues and solutions

Need Help?

Join our community for support, bug reports, and feature requests.

Why Choose Cumulus Over Cloud Providers?

Compare Cumulus API Gateway against major cloud providers in performance, cost, and flexibility

8ms avg

Superior Performance

8x faster response times with sub-10ms latency

Unlimited

Complete Control

Full customization with NGINX + Lua scripting

$135/mo

Most Cost-Effective

Lowest 3-year TCO with pre-built solution

Any cloud

Zero Vendor Lock-in

Deploy on any cloud or on-premises Kubernetes

Full stack

Local Development

Complete local testing with k3d/minikube

Enterprise+

Advanced Features

Circuit breakers, custom metrics, protocol translation

Cost Analysis & TCO

Feature
C
Cumulus
A
AWS
A
Azure
G
GCP
3-Year TCO (10M req/month)$4,860$24,160$35,680$22,160
Monthly Cost (Est.)$90-150$350+$280+$300+
Monthly Average (3-Year)$135$671$991$615
Setup & DevelopmentPre-built (no cost)$4,000$4,000$2,000
Vendor Lock-in RiskNoneHighHighHigh

Performance & Scalability

Feature
C
Cumulus
A
AWS
A
Azure
G
GCP
Average Latency8ms75ms95ms65ms
P99 Latency25ms200ms300ms180ms
Max Throughput (RPS)10,000+5,0002,0003,000
Auto-scalingManual/HPAAutomaticLimitedAutomatic

Features & Capabilities

Feature
C
Cumulus
A
AWS
A
Azure
G
GCP
Circuit Breaker
Custom Lua Scripting
Local Development
Multi-Cloud Support
Custom Metrics
Protocol Translation

Security & Compliance

Feature
C
Cumulus
A
AWS
A
Azure
G
GCP
DDoS Protection
WAF Integration
Custom Firewalls
Certificate Management
Private Endpoints

When to Choose Each Solution

Choose Cumulus When

Need sub-10ms latency and high throughput (10,000+ RPS)
Complex business logic requiring custom Lua scripting
Multi-cloud strategy to avoid vendor lock-in
High-volume APIs (>10M requests/month) for maximum cost savings
Complete control over authentication and authorization
Advanced features like circuit breakers and protocol translation
Pre-built solution requiring no development investment
Strong DevOps and Kubernetes expertise in team
Compliance requiring specific security implementations

Choose Cloud Providers When

AWS API Gateway
Already using AWS services extensively
Need immediate production deployment
Heavy serverless (Lambda) integration
Variable or unpredictable traffic patterns
Azure APIM
Need enterprise features and developer portal
Hybrid deployment (on-premises + cloud)
Heavy Microsoft technology usage
API monetization and productization
GCP API Gateway
Using GCP as primary cloud platform
Need simple, straightforward solution
Strong OpenAPI specification usage
Cost-effective cloud-native option

Experience the Most Cost-Effective API Gateway Solution

Join developers who have achieved superior performance at 80% lower cost than cloud alternatives. Pre-built solution ready for immediate deployment with no development overhead.

8ms
Average Latency
10,000+
Requests/Second
$4.9k
3-Year TCO

Performance & Security

Built for production workloads with enterprise-grade security, performance optimization, and monitoring capabilities

Performance Benchmarks

Response Time

Average latency for simple GET requests

8ms
Cumulus
75ms
AWS
95ms
Azure
65ms
GCP

Throughput

Requests per second with 100 concurrent users

8,500 RPS
Cumulus
5,000 RPS
AWS
2,000 RPS
Azure
3,000 RPS
GCP

P99 Latency

99th percentile response time under load

25ms
Cumulus
200ms
AWS
300ms
Azure
180ms
GCP

Memory Usage

Memory footprint per gateway instance

256MB
Cumulus
N/A
AWS
N/A
Azure
N/A
GCP

Latency Comparison

Cumulus
8ms
AWS
75ms
Azure
95ms
GCP
65ms

Enterprise Security

DDoS Protection

Cloud Load Balancer + Rate Limiting

WAF Integration

ModSecurity with OWASP Core Rule Set

Certificate Management

Automated TLS/SSL with cert-manager

mTLS Support

Mutual TLS certificate validation

JWT Validation

Built-in token validation and claims

IP Filtering

Whitelist/blacklist with network policies

Scaling & Availability

Horizontal Pod Autoscaling

3-100 pods

Scale from 3 to 100+ pods based on CPU/memory

Multi-Region Deployment

Multi-AZ

Deploy across multiple zones and regions

Circuit Breaker

99.9% uptime

Automatic failover and failure detection

Load Balancing

Multiple algorithms

Round-robin, least-connections, IP hash

Production-Ready Performance

Join companies achieving 10,000+ RPS with sub-10ms latency. Experience the performance difference with our benchmarked solution.