Skip to main content

​
Implementing Global API Deployments with Tyk MDCB

Tyk MDCB enables you to deploy your API management platform globally while maintaining centralized control. This guide shows you how to implement worldwide API availability with optimized performance for users in different geographic regions.

​
Global Deployment Fundamentals

Global API deployments bring your APIs closer to consumers by:
  • Reducing latency through geographic proximity
  • Improving reliability with regional redundancy
  • Meeting regional compliance requirements
  • Optimizing user experience across the world
Tyk’s MDCB facilitates this by separating the control plane (where you manage APIs) from data planes (where API traffic is processed).

​
Global Deployment Patterns

​
Hub and Spoke Pattern

[Diagram of hub and spoke deployment pattern] A central control plane with multiple regional data planes:
  • Control Plane: Central management hub with Dashboard, MDCB, and control Gateway
  • Data Planes: Regional processing centers with worker Gateways and local Redis
  • Management Flow: All administration through the central Dashboard
  • Synchronization: MDCB pushes configurations to all data planes
Best for: Organizations with centralized IT governance and distributed operations

​
Regional Clusters Pattern

[Diagram of regional clusters pattern] Multiple semi-autonomous regional deployments:
  • Regional Control Planes: Each region has its own control plane
  • Regional Data Planes: Data planes connect to their regional control plane
  • Limited Synchronization: Optional synchronization between control planes
  • Regional Administration: Each region can have its own administrators
Best for: Organizations with strong regional autonomy or complex compliance requirements

​
Edge Deployment Pattern

[Diagram of edge deployment pattern] Lightweight edge Gateways with centralized management:
  • Central Control: Centralized control plane for management
  • Edge Gateways: Smaller, focused Gateway deployments at the edge
  • CDN Integration: Often combined with Content Delivery Networks
  • Caching Focus: Heavy emphasis on caching at the edge
Best for: Content-heavy APIs or applications requiring ultra-low latency

​
Regional Implementation

​
Region Selection Strategy

Select regions based on:
  • User Distribution: Deploy close to your API consumers
  • Backend Proximity: Consider proximity to backend services
  • Latency Requirements: More regions for latency-sensitive applications
  • Cost Considerations: Balance performance needs with infrastructure costs
  • Compliance Requirements: Address data residency regulations
Recommended starting regions for global coverage:
  • North America (US East/West)
  • Europe (Central/West)
  • Asia Pacific (Singapore/Tokyo)
  • Optional: South America, Australia, Middle East

​
Cross-Region Communication

Optimize communication between regions: [Diagram showing cross-region communication patterns]
  • Control Plane to Data Plane: Secure, efficient RPC communication
  • Data Plane to Backend: Direct connection or regional proxying
  • Cross-Region Requests: Avoid when possible for performance
  • Analytics Aggregation: Efficient collection of operational data

​
Load Balancing Approach

Implement global load balancing:
  • DNS-Based Routing: Geographic DNS routing to nearest data plane
  • Anycast IP: Network-level routing to closest instance
  • Application-Level Routing: Custom routing based on user attributes
  • Health Checking: Regular verification of regional availability

​
Performance Optimization

​
Latency Reduction Techniques

Minimize latency for global users: 
{
"analytics_config": {
"enable_detailed_recording": false,
"enable_geo_ip": true
},
"cache": {
"global_cache": {
"redis": {
"host": "redis-local",
"port": 6379
}
}
}
}
Key techniques:
  • Disable detailed recording for performance-critical APIs
  • Enable geographic data for analytics
  • Configure local caching for frequently accessed data
  • Optimize TLS session resumption

​
Caching Strategies

Implement effective caching: 
{
"cache_options": {
"cache_timeout": 60,
"enable_cache": true,
"cache_all_safe_requests": true,
"cache_response_codes": [200, 301, 302],
"enable_upstream_cache_control": true
}
}
Key strategies:
  • Cache common responses at the Gateway level
  • Respect upstream cache control headers
  • Implement different cache timeouts by region
  • Consider edge caching for static content

​
Traffic Routing Optimization

Optimize how traffic flows:
  • API Segmentation: Deploy different APIs to different regions
  • Backend Routing: Connect to closest backend services
  • Response Transformation: Optimize response size for different regions
  • Protocol Selection: Use HTTP/2 or HTTP/3 where supported

​
Monitoring Global Deployments

Implement comprehensive monitoring:
  • Regional Health: Monitor each region independently
  • Global Dashboard: Unified view of all regions
  • Latency Tracking: Monitor latency from different regions
  • Traffic Distribution: Track request distribution across regions
  • Alert Correlation: Correlate alerts across regions

​
Implementation Example: E-commerce Global API Platform

This example demonstrates how a global e-commerce company implemented worldwide API availability for their platform. [Detailed diagram of global e-commerce API deployment] The company implemented:
  1. Control Plane: Deployed in US East with disaster recovery in US West
  2. Data Planes: Six regional deployments:
    • North America (US East, US West)
    • Europe (Ireland, Frankfurt)
    • Asia (Singapore, Tokyo)
  3. Configuration Highlights: 
{
"gateway": {
"optimizations_use_async_session_write": true,
"enable_custom_domains": true,
"enforce_org_quotas": false
},
"slave_options": {
"group_id": "region-eu-west",
"use_ssl": true,
"ssl_insecure_skip_verify": false
}
}
  1. Caching Strategy:
  • 30-second cache for product listings
  • 5-minute cache for product details
  • No cache for cart and checkout APIs
  • Edge caching for images and static content
  1. Load Balancing:
  • DNS-based geographic routing
  • Health checks every 30 seconds
  • Automatic failover to next closest region
Results:
  • 65% reduction in API latency
  • 99.99% global availability
  • Seamless handling of regional traffic spikes
  • Compliance with regional data regulations

​
Next Steps