CycleNest API
CycleNest API is a high-performance, cloud-integrated RESTful backend engineered for bike-sharing and logistics tracking networks. Built using the JAX-RS (Jersey) framework and deployed inside high-concurrency Apache Tomcat servlet containers, the API orchestrates real-time asset discovery, location-based distance calculations, and transaction records across global cloud boundaries.
Core Features
JAX-RS REST API Architecture
High-performance endpoint structures mapped under servlet containers handling concurrent HTTP methods (GET, POST, DELETE).
Self-Healing Storage Fallback
Automated active failover pattern redirecting traffic to local thread-safe ConcurrentHashMap stores in the event of database timeouts, ensuring 100% API uptime.
Asynchronous OSRM Router
Non-blocking coordinate driving evaluation using modern Java HttpClient and CompletableFuture pipelines.
Azure Cosmos DB Storage
Globally-distributed database persistence utilizing Azure's Java SDK with Session consistency levels for strong read-your-own-writes guarantees.
Operational Health Probes
Custom diagnostic debug endpoints checking latency, ping states, and network connectivity parameters in real-time.
Technical Deep Dive
Asynchronous Spatial Routing
Engineered a highly optimised routing client using Java's modern HttpClient API and CompletableFutures. Upon query activation, the backend dispatches non-blocking asynchronous calls to the Open Source Routing Machine (OSRM) service, resolving real road distances and durations on the fly. Calculated metrics are parsed efficiently using Jackson ObjectMapper and returned without locking primary execution threads.
Active Data-Store Failover Wrapper
To prevent API cold-starts or external database down-times from breaking the application lifecycle, I architected a robust Repository facade with automatic self-healing properties. If the Azure Cosmos DB connection fails during container initialization, the factory gracefully traps the error and binds the controller routes to an in-memory data store, keeping services live.
Cosmos Client & TCP Tuning
Configured CosmosRepository initialization blocks with performance tuning overrides, forcing Gateway connectivity modes and tuning reactor-netty thread-worker pools. Added low-level system properties like IPv4 stack prioritization and HTTP client timeout settings to prevent thread leaks and ensure fast API roundtrips.
Performance Benchmark
"Architected a resilient JAX-RS backend with an automated Cosmos DB to in-memory self-healing failover mechanism, maintaining 100% application availability during database timeouts."