Stability Fix

Author: alokemajumder

DEPLOYMENT_GUIDE.md

@@ -0,0 +1,217 @@
+# 🚀 FFprobe API Deployment Guide
+
+## 📋 Deployment Options Overview
+
+### 🟢 Simple Deployment (Recommended for Small/Test Organizations)
+**File**: `compose.simple.yml`  
+**Purpose**: Complete LLM-powered API setup without monitoring overhead
+
+**What's Included:**
+- ✅ FFprobe API service
+- ✅ PostgreSQL database  
+- ✅ Redis cache
+- ✅ **Ollama LLM (enabled by default)** - Essential for AI-powered analysis
+- ✅ **OpenRouter fallback** - Automatic fallback for enhanced reliability
+- ❌ No Prometheus/Grafana (enterprise-only monitoring)
+
+**Resource Usage:**
+- Memory: ~4-5GB total (includes LLM)
+- CPU: 2-4 cores recommended
+- Storage: ~8GB base + models + uploads
+
+**Command:**
+```bash
+docker compose -f compose.simple.yml up -d
+```
+
+**Perfect for:**
+- Small organizations
+- Test/staging environments  
+- Cost-conscious deployments
+- Quick demos with AI features
+
+---
+
+### 🟡 Development Deployment
+**File**: `compose.yml + compose.dev.yml`  
+**Purpose**: Local development with debugging tools and AI features
+
+**What's Included:**
+- ✅ All simple deployment features
+- ✅ Adminer (database GUI)
+- ✅ Redis Commander (Redis GUI)
+- ✅ Hot reload for development
+- ✅ Debug logging enabled
+- ✅ Full LLM capabilities
+
+**Command:**
+```bash
+docker compose -f compose.yml -f compose.dev.yml up -d
+```
+
+---
+
+### 🟠 Production Deployment  
+**File**: `compose.yml + compose.production.yml`
+**Purpose**: Medium-scale production with enhanced AI features
+
+**What's Included:**
+- ✅ All simple deployment features
+- ✅ **Enhanced Ollama setup** - Optimized for production workloads
+- ✅ **Multiple LLM models** - Better AI analysis variety
+- ✅ Production-optimized settings
+- ✅ Resource limits configured
+- ✅ **Intelligent LLM fallback** - Local-first, cloud backup
+- ❌ No monitoring stack (keeps it lightweight)
+
+**Resource Usage:**
+- Memory: ~6-8GB total  
+- CPU: 4-6 cores recommended
+- Storage: ~15GB base + models + uploads
+
+**Command:**
+```bash
+docker compose -f compose.yml -f compose.production.yml up -d
+```
+
+---
+
+### 🔴 Enterprise Deployment
+**File**: `compose.yml + compose.enterprise.yml`
+**Purpose**: Full-scale enterprise with monitoring and AI intelligence
+
+**What's Included:**
+- ✅ All production deployment features
+- ✅ **Prometheus monitoring**
+- ✅ **Grafana dashboards**
+- ✅ Load balancer (Nginx)
+- ✅ Message queue (RabbitMQ)
+- ✅ **Advanced LLM orchestration** - Multiple models with smart routing
+- ✅ Horizontal scaling support
+- ✅ Enhanced resource allocation
+
+**Resource Usage:**
+- Memory: ~12-16GB total
+- CPU: 8+ cores recommended
+- Storage: ~30GB base + monitoring data
+
+**Command:**
+```bash
+docker compose -f compose.yml -f compose.enterprise.yml up -d
+```
+
+---
+
+## 🤖 AI/LLM Features Across All Deployments
+
+All deployment options include **LLM-powered analysis** by default:
+
+### 🎯 **What's AI-Powered:**
+- ✅ **Video Analysis Reports** - Human-readable technical insights
+- ✅ **Quality Assessment** - Professional video quality evaluation  
+- ✅ **Comparison Analysis** - AI-driven before/after analysis
+- ✅ **Technical Recommendations** - FFmpeg optimization suggestions
+- ✅ **Format Suitability** - Delivery platform recommendations
+
+### 🔄 **Smart Fallback System:**
+1. **Local LLM First** - Uses Ollama for privacy and speed
+2. **OpenRouter Fallback** - Automatic cloud backup if local fails
+3. **Graceful Degradation** - API continues working without AI if both fail
+
+### ⚙️ **LLM Configuration:**
+```bash
+# Local LLM (default: enabled)
+ENABLE_LOCAL_LLM=true
+OLLAMA_URL=http://ollama:11434
+OLLAMA_MODEL=phi3:mini
+
+# OpenRouter fallback (optional)
+OPENROUTER_API_KEY=your-api-key-here
+```
+
+---
+
+## 🎯 Which Deployment Should You Choose?
+
+### Choose **Simple** if:
+- ✅ Small team (< 10 users)
+- ✅ Want AI features without complexity
+- ✅ Budget/resource constraints
+- ✅ Testing or staging environment
+- ✅ Don't need monitoring dashboards
+
+### Choose **Production** if:
+- ✅ Medium team (10-50 users)
+- ✅ Need enhanced AI performance
+- ✅ Production workload with AI requirements
+- ✅ Want optimized LLM processing
+
+### Choose **Enterprise** if:
+- ✅ Large team (50+ users)
+- ✅ Need comprehensive monitoring  
+- ✅ High availability requirements
+- ✅ Advanced AI orchestration needed
+- ✅ Compliance/audit requirements
+
+---
+
+## 🔧 Quick Setup Commands
+
+### Simple Deployment (LLM-Powered)
+```bash
+# 1. Clone repository
+git clone https://github.com/rendiffdev/ffprobe-api.git
+cd ffprobe-api
+
+# 2. Set environment variables
+cp .env.example .env
+# Edit .env with your values
+
+# 3. Deploy with AI features
+docker compose -f compose.simple.yml up -d
+
+# 4. Verify (should show LLM status)
+curl http://localhost:8080/health
+```
+
+### Test AI Features
+```bash
+# Upload a video and get AI analysis
+curl -X POST http://localhost:8080/api/v1/probe/file \
+  -H "X-API-Key: your-api-key" \
+  -F "[email protected]"
+
+# The response will include LLM-generated insights
+```
+
+---
+
+## 💡 Cost Optimization Tips
+
+1. **Start Simple**: Begin with `compose.simple.yml` - includes AI without monitoring overhead
+2. **Local LLM First**: Uses free Ollama models, only pays for OpenRouter fallback when needed
+3. **Smart Resource Limits**: Each deployment tier optimized for different scales
+4. **Optional Cloud LLM**: OpenRouter fallback is optional - works great with just local LLM
+
+---
+
+## 📊 Resource Requirements Summary
+
+| Deployment | Memory | CPU | Storage | AI Features | Monitoring |
+|------------|--------|-----|---------|-------------|------------|
+| **Simple** | 4-5GB | 2-4 cores | 8GB+ | ✅ Full LLM | Logs only |
+| **Production** | 6-8GB | 4-6 cores | 15GB+ | ✅ Enhanced LLM | Logs only |
+| **Enterprise** | 12-16GB | 8+ cores | 30GB+ | ✅ Advanced LLM | Full monitoring |
+
+---
+
+## 🔒 Security & Privacy
+
+- **Local LLM**: All AI processing can run locally for maximum privacy
+- **Encrypted Communication**: All external LLM calls use HTTPS
+- **API Key Security**: OpenRouter keys are optional and securely managed
+- **No Data Leakage**: Local-first approach means your videos stay on your infrastructure
+
+---
+
+*The FFprobe API is designed to be **AI-first** while maintaining complete flexibility in deployment scale and privacy requirements.*