📊 Glean Observability Dashboard

Production-ready observability platform for Glean deployments with USE methodology monitoring and comprehensive cost analytics.

---

🎯 Quick Links

• 🚀 Quick Deployment Guide - Deploy in under 10 minutes
• 📖 Complete Setup Guide - Detailed setup for local and production
• 🧪 Testing Guide - Local testing instructions
• 🔧 Migration Status - Track feature completeness

---

🚀 Tech Stack

• Frontend: Next.js 15 (React) + TypeScript + Tailwind CSS + Recharts
• Backend: FastAPI (Python 3.13) + Google Cloud APIs
• Monitoring: Prometheus (GKE Managed) + Cloud Monitoring API
• Deployment: Vercel (Frontend) + Cloud Run (Backend)
• Cost: ~$5-20/month for typical usage

✨ Features

USE Methodology Monitoring

• Utilization: CPU %, Memory %, Pod Availability
• Saturation: CPU Throttling, Queue Backlogs
• Errors: Pod Restarts, Crash Loops

Cost Analytics

• Real-time GKE cluster cost estimates
• Cloud SQL instance costs
• Total infrastructure spend
• Daily/Monthly projections

Components Monitored

• 🔍 Crawler (Content Ingestion)
• 📡 Datasource Events Handler
• 🧠 Query Parser (NLP/ML)
• ⚡ Query Engine
• 🧬 Semantic Index (Qdrant Vector DB)
• 🔎 Keyword Index (Cloud SQL)
• 👥 User Data Layer
• 🌐 Load Balancer

🏗️ Architecture

┌─────────────────────────────────────┐
│        Next.js Frontend             │
│     (TypeScript + Tailwind)         │
│                                     │
│  - Dashboard UI                     │
│  - USE Metric Visualization         │
│  - Cost Charts                      │
│  - Component Health Cards           │
└──────────────┬──────────────────────┘
               │ REST API
               ↓
┌─────────────────────────────────────┐
│       FastAPI Backend               │
│          (Python)                   │
│                                     │
│  - Prometheus Client                │
│  - GCP Metrics Client               │
│  - Cost Estimator                   │
│  - Config Management                │
└──────────────┬──────────────────────┘
               │
               ↓
┌─────────────────────────────────────┐
│         GCP Services                │
│                                     │
│  - GKE Managed Prometheus           │
│  - Cloud Monitoring API             │
│  - Cloud SQL                        │
│  - Cloud Billing API                │
└─────────────────────────────────────┘

📁 Project Structure

glean-observability-dashboard/
├── frontend/              # Next.js application
│   ├── src/
│   │   ├── app/          # App router pages
│   │   ├── components/   # React components
│   │   ├── lib/          # Utilities
│   │   └── types/        # TypeScript types
│   └── package.json
│
├── backend/               # FastAPI application
│   ├── app/
│   │   ├── api/          # API endpoints
│   │   ├── clients/      # GCP/Prometheus clients
│   │   ├── models/       # Data models
│   │   └── main.py       # FastAPI app
│   └── requirements.txt
│
└── README.md

🚀 Quick Start

For teammates: See SETUP_GUIDE.md for complete setup instructions with GCP authentication.

For deployment: See DEPLOYMENT.md for production deployment in 10 minutes.

Fastest Local Setup (3 commands)

# 1. Authenticate with GCP
gcloud auth application-default login

# 2. Start backend (Terminal 1)
cd backend && python3 -m venv venv && source venv/bin/activate && pip install -r requirements.txt && uvicorn app.main:app --reload --port 8000

# 3. Start frontend (Terminal 2)
cd frontend && npm install && npm run dev

Open http://localhost:3000 and you're ready! 🎉

---

🌐 Production Deployment

Option 1: Automated (Recommended)

Click the deploy buttons at the top of this README for one-click deployment to Vercel and Cloud Run.

Option 2: Manual (Full Control)

See DEPLOYMENT.md for step-by-step instructions.

Summary:

1. Deploy backend to Cloud Run: gcloud run deploy --source .
2. Deploy frontend to Vercel: vercel --prod
3. Connect them with environment variables
4. Share with your team! 🎊

Deployed URLs:

• Frontend: https://glean-observability-dashboard.vercel.app
• Backend: https://glean-observability-api-xxx.run.app

---

🎨 UI Features

• ✅ Real-time Metrics: Auto-refresh with configurable intervals (30-300s)
• ✅ USE Methodology: Utilization, Saturation, Errors for all components
• ✅ Cost Analytics: GKE + Cloud SQL cost estimates and breakdowns
• ✅ Responsive Design: Mobile, tablet, and desktop optimized
• ✅ Time Range Selector: 1h, 6h, 24h, 7d, 30d historical data
• ✅ Component Cards: Expandable details with color-coded health status
• ✅ Interactive UI: Collapsible sections, hover states, loading indicators
• ✅ Professional Theme: Modern design with Tailwind CSS and Lucide icons

---

📊 API Endpoints

Core Endpoints

• GET /api/health - Health check
• GET /api/cluster/overview - Cluster metrics (nodes, CPU, memory, pods)
• GET /api/cluster/cost - Cost estimates (GKE + Cloud SQL)
• GET /api/components - List all monitored components
• GET /api/components/{name} - Detailed USE metrics for a component

Query Parameters

• project_id (required) - GCP project ID
• deployment_name (optional) - Deployment identifier
• lookback_hours (optional) - Historical data window (default: 24)

Example Request

curl "http://localhost:8000/api/cluster/overview?project_id=glean-support-sandbox&deployment_name=support-sandbox"

Full API docs: http://localhost:8000/docs (FastAPI auto-generated)

---

🔒 Security & Authentication

For Local Development

• Use gcloud auth application-default login (your Google account)
• No additional setup needed

For Production

• Backend: Service account with IAM roles:
  • roles/monitoring.viewer
  • roles/container.viewer
  • roles/cloudsql.viewer
• Frontend: Optional OAuth 2.0 or Vercel password protection
• CORS: Configured for Vercel domains
• Rate Limiting: Built into Cloud Run (80 concurrent requests)

See SETUP_GUIDE.md for detailed security configuration.

---

📈 Monitoring Best Practices

This dashboard implements Brendan Gregg's USE Method:

1. For every resource, check:

   • Utilization: How busy is it?
   • Saturation: Is there queuing?
   • Errors: Are there failures?

2. Color-coded thresholds:

   • 🟢 Green: Healthy
   • 🟡 Yellow: Warning
   • 🔴 Red: Critical

🤝 Contributing

This is an internal Glean tool. For improvements or bug fixes:

1. Fork the repository
2. Create a feature branch: git checkout -b feature/amazing-feature
3. Make your changes
4. Test thoroughly (see TESTING_GUIDE.md)
5. Commit: git commit -m 'Add amazing feature'
6. Push: git push origin feature/amazing-feature
7. Open a Pull Request

Development Workflow

# Make changes to backend
cd backend && source venv/bin/activate
# Edit code, then restart: uvicorn app.main:app --reload --port 8000

# Make changes to frontend
cd frontend
# Edit code - Next.js hot-reloads automatically

---

🐛 Troubleshooting

Common Issues

"Failed to fetch data"

• Check backend is running: curl http://localhost:8000/api/health
• Verify GCP auth: gcloud auth application-default print-access-token
• Check browser console for CORS errors

"Permission denied"

• Run: gcloud auth application-default login
• Ensure you have GCP project access

"No metrics available"

• Verify GKE cluster has Managed Prometheus enabled
• Check project ID is correct

See SETUP_GUIDE.md for detailed troubleshooting.

---

📝 License

Internal Glean tool - Not for external distribution.

---

🙏 Acknowledgments

• USE Method: Based on Brendan Gregg's USE methodology
• Built for: Glean infrastructure monitoring
• Powered by: Google Cloud Platform, Prometheus, FastAPI, Next.js
• Migration: Successfully migrated from Streamlit to Next.js for better UX

---

📞 Support

• Internal Slack: #glean-observability
• Issues: Open a GitHub issue
• Documentation: See linked guides above
• On-call: Check PagerDuty rotation

---

🎉 Success Metrics

After deploying, you should see:

• ✅ Cluster overview with health status
• ✅ Cost estimates for GKE and Cloud SQL
• ✅ 8 component cards with USE metrics
• ✅ Auto-refresh working
• ✅ Time range selection functional
• ✅ All metrics color-coded (green/yellow/red)

Happy monitoring! 📊🚀