404Dev: GraphQL API, Prometheus & New Threading
Join CommunitySummary
The 404Development | Software Hub community announced a major update featuring a new GraphQL API powered by Strawberry, comprehensive Prometheus metrics for observability across various system components, and enhanced health checks. This release also introduces thread-safe atomic operations and a refined threading model that separates the API onto its own thread, ensuring better performance and preventing interference with the main Discord bot operations. These additions significantly improve mo...
โจ NEW FEATURES
๐ฏ GraphQL API
- Full GraphQL support using Strawberry GraphQL
- Endpoint:
POST /api/v2/graphql - Queries:
userTime,clockStatus,leaderboard,insights,systemHealth - Mutations:
clockIn,clockOut - Interactive playground at
/api/v2/graphql
๐ Prometheus Metrics & Monitoring
- 50+ metrics for comprehensive observability
- Metrics endpoint:
GET /metrics(Prometheus text format) - Categories tracked:
- API requests (count, duration, size, in-flight)
- Authentication (attempts, rate limits, active keys)
- Database (Redis/PostgreSQL operations, connection pools)
- Cache (hit/miss rates, sizes, evictions per layer)
- Threading (workers, queue size, lock contention)
- Business (active sessions, clock ops, WebSocket connections)
- Batch processing (operations, sizes, durations)
๐ฅ Enhanced Health Checks
- Basic health:
GET /health- Simple up/down status - Detailed health:
GET /health/detailed- Full system diagnostics - Thread pool status (active workers, queue size)
- Redis connection status
- PostgreSQL pool usage and database size
- Active sessions count
- Cache sizes (L1, L2, L3)
- Circuit breaker metrics
๐ Thread-Safe Atomic Operations
- New module:
Utils/atomic_operations.py - Decorators:
@atomic_redis,@atomic_db,@atomic_cache - Batch processor:
AtomicBatchProcessorfor efficient bulk operations - Context managers:
AtomicRedisContext,AtomicDBContext - Metrics integration: Automatic lock wait time tracking
- Thread-safe singleton: Base class for shared resources
๐งต Threading Model
- API in separate thread with its own asyncio event loop
- Thread pool: 4 workers for batch processing
- No bot interference: API operations don't block Discord bot
- Better CPU utilization: Multi-core usage
- Daemon thread: Automatic cleanup on bot shutdown
๐จ Enhanced API Documentation
- Swagger UI improvements:
- Search/filter enabled
- Request duration display
- Deep linking (shareable URLs)
- Monokai dark theme
- Persistent authorization
- Auto-collapsed models
- GraphQL Playground: Interactive GraphQL IDE
- ReDoc: Alternative documentation view
