Overview

This pull request implements comprehensive system optimizations, security hardening, detailed health monitoring, and containerized deployment configuration to make the FinWise backend ready for production.

Key Changes

1. Architectural & Core Utilities

• Logger Service: Added a custom level-based LoggerService that outputs structured JSON in production (safely redacting sensitive parameters like password, token, apiKey) and pretty-printed format in development.
• Cache Service: Built a pluggable caching wrapper using ioredis with an automated memory-safe fallback to prevent memory leaks if Redis is offline.
• Lock Service: Built a distributed lock mechanism using Redis (SET NX PX) with a local fallback to prevent overlapping background task processing.

2. Throttling & Concurrency Guardrails

• Rate Limiting: Re-implemented the Express rateLimitMiddleware to support Redis-backed rate limiting with a self-pruning memory-safe Map fallback.
• Worker Locking: Locked the background tick in NotificationWorker to ensure only one instance executes reminder checks at a time in scale-out container environments.

3. Caching & Performance Optimization

• System Categories Caching: Cached read-only system categories for 1 hour, bypassing the database for common app-startup list/tree requests.
• Financial Reports Caching: Cached report aggregates (overview, cash flow, categories spending, budget performance) for 5 minutes per user query.
• Cache Invalidation Triggers: Configured instant cache invalidation upon any mutations to Transactions, Wallets, Budgets, and Saving Goals.

4. System Health Monitoring

• Detailed Health Endpoint: Replaced the basic /health check with an active diagnostics controller returning:
  • Server resource utilization (memory rss/heap, process uptime).
  • Database availability & round-trip query latency (ms).
  • Cache adapter state (Redis connection vs. Memory fallback).
• Swagger API Docs: Updated the /health API schemas to document the detailed JSON response structure.

5. Deployment, CI/CD & Code Quality

• Dockerfile: Defined a multi-stage production Docker build using a lightweight Node.js Alpine base, executing under a non-root node user.
• Docker Compose: Configured healthy checks dependency order so that the Node application waits for PostgreSQL and Redis to be fully responsive.
• CI pipeline: Created a GitHub Actions workflow .github/workflows/ci.yml verifying TypeScript compilation, Prisma schemas, and ESLint.
• ESLint Configuration: Set up a flat eslint.config.mjs resolving all source-code formatting and type rules.

Verification Results

• Compilation: pnpm run build compiles successfully.
• Linting: pnpm run lint executes cleanly with zero errors.
• Database validation: pnpm exec prisma validate completes successfully.
• Health Check Probe: Proved database latency and Redis connection status:

  {
    "success": true,
    "status": "ok",
    "timestamp": "2026-08-06T10:18:58.937Z",
    "uptime": 11.58,
    "memory": { "rss": "72.16 MB", "heapTotal": "23.69 MB", "heapUsed": "20.54 MB" },
    "database": { "status": "up", "latencyMs": 3 },
    "cache": { "status": "up", "type": "redis" }
  }