Traffic Routing
Implements round robin, least connections, weighted routing, and consistent hashing with runtime strategy switching.
Why This Project Signals Engineering Depth
Reliability
Active health checks remove failing nodes from rotation and restore them after recovery.
Operational Clarity
Admin APIs expose backend health and connection load for observability and decision-making.
Production Controls
Supports JWT/Cognito auth, rate limiting, retries/backoff, trace IDs, panic recovery, and cost awareness.
Live Control Plane
Reads from /admin/backends and /admin/strategy.
Current Strategy:
Loading...
Proxy Endpoint:
/proxy/
Estimated Cost:
loading...
How Request Routing Works
- Client request reaches
/proxy/*. - Balancer chooses backend using active strategy.
- Health checks keep unhealthy backends out of rotation.
- Proxy forwards request and updates connection counters.
- Failures trigger automatic rerouting to healthy nodes.
Routing Trade-offs
Round Robin
Simple and fair when all nodes are similar in capacity and latency profile.
Least Connections
Adapts to uneven request duration by sending new traffic to less-busy nodes.
Weighted Routing
Distributes traffic by node capacity so stronger backends handle a larger share.
Consistent Hashing
Best for affinity and cache locality because key-to-node mapping stays stable.
Health checks and failover apply across all strategies to keep unhealthy nodes out of rotation.
AI Copilot
Ask architecture, reliability, or optimization questions based on live cluster state.
AI Provider:
loading...
AI response will appear here.