YOUR_RAILWAY_CONFIG.md

🚂 Your Railway Configuration - Ready to Deploy

Your Current Setup Analysis

Based on your existing Railway variables, your configuration uses:

• ✅ Database: test namespace and database (not production)
• ✅ Multiple AI Providers: Gemini (Google), Groq, Llama
• ✅ Worker Configuration: 5 concurrent tasks with retry logic
• ✅ Proper retry settings: Exponential jitter for reliability

STEP 1: Current Railway Variables (What You Already Have)

These are already set in your Railway deployment:

# Database - Already Configured ✅
SURREAL_URL=ws://localhost:8000/rpc              # ⚠️ NEEDS FIX (see below)
SURREAL_USER=root
SURREAL_PASSWORD=root
SURREAL_NAMESPACE=test
SURREAL_DATABASE=test

# API URL - Already Configured ✅
API_URL=http://localhost:5055                    # ⚠️ NEEDS UPDATE (see below)

# Worker & Retry - Already Configured ✅
SURREAL_COMMANDS_MAX_TASKS=5
SURREAL_COMMANDS_RETRY_ENABLED=true
SURREAL_COMMANDS_RETRY_MAX_ATTEMPTS=3
SURREAL_COMMANDS_RETRY_WAIT_STRATEGY=exponential_jitter
SURREAL_COMMANDS_RETRY_WAIT_MIN=1
SURREAL_COMMANDS_RETRY_WAIT_MAX=30

STEP 2: Critical Fixes Needed

Fix 1: Change localhost to 127.0.0.1 for SURREAL_URL

In Railway Dashboard → Variables:

❌ Current (WRONG):

SURREAL_URL=ws://localhost:8000/rpc

✅ Change to (CORRECT):

SURREAL_URL=ws://127.0.0.1:8000/rpc

Why? Railway's container networking requires 127.0.0.1 instead of localhost.

Fix 2: Add INTERNAL_API_URL

Add this new variable in Railway:

INTERNAL_API_URL=http://127.0.0.1:5055

Why? Next.js needs this for server-side API proxying.

Fix 3: Update API_URL to Your Railway Domain

After your first successful deploy:

❌ Current:

API_URL=http://localhost:5055

✅ Change to YOUR Railway domain:

API_URL=https://your-app-production-xxxx.up.railway.app

How to find your Railway domain:

1. Go to Railway Dashboard → Your Service
2. Look at the "Deployments" tab
3. Copy the domain (e.g., https://se-production-1234.up.railway.app)
4. Paste it as the API_URL value (without /api at the end)

STEP 3: Add Your AI API Keys

You mentioned using Gemini, Groq, and Llama. Add these variables:

For Google Gemini (Required)

GOOGLE_API_KEY=your_actual_gemini_api_key

Get it at: https://makersuite.google.com/app/apikey

For Groq (Required)

GROQ_API_KEY=your_actual_groq_api_key

Get it at: https://console.groq.com/keys

For Llama via Ollama (If applicable)

If you're running Ollama somewhere accessible:

OLLAMA_API_BASE=http://your-ollama-host:11434

OR if using Llama via Groq (most common):

• No extra configuration needed - Groq provides Llama models

Optional: Other Providers

If you want to add more providers later:

# OpenAI (optional)
OPENAI_API_KEY=sk-your_key

# Anthropic Claude (optional)
ANTHROPIC_API_KEY=sk-ant-your_key

# Mistral (optional)
MISTRAL_API_KEY=your_key

STEP 4: Complete Railway Variables List

Copy this EXACT configuration to Railway:

# ============================================
# DATABASE (Keep as-is)
# ============================================
SURREAL_URL=ws://127.0.0.1:8000/rpc
SURREAL_USER=root
SURREAL_PASSWORD=root
SURREAL_NAMESPACE=test
SURREAL_DATABASE=test

# ============================================
# API CONFIGURATION
# ============================================
INTERNAL_API_URL=http://127.0.0.1:5055
API_URL=https://YOUR-RAILWAY-DOMAIN.up.railway.app

# ============================================
# WORKER & RETRY (Keep as-is)
# ============================================
SURREAL_COMMANDS_MAX_TASKS=5
SURREAL_COMMANDS_RETRY_ENABLED=true
SURREAL_COMMANDS_RETRY_MAX_ATTEMPTS=3
SURREAL_COMMANDS_RETRY_WAIT_STRATEGY=exponential_jitter
SURREAL_COMMANDS_RETRY_WAIT_MIN=1
SURREAL_COMMANDS_RETRY_WAIT_MAX=30

# ============================================
# AI PROVIDERS (ADD YOUR KEYS)
# ============================================
GOOGLE_API_KEY=your_actual_gemini_key
GROQ_API_KEY=your_actual_groq_key

# Optional: If using Ollama for Llama
# OLLAMA_API_BASE=http://your-ollama-host:11434

# Optional: Other providers
# OPENAI_API_KEY=sk-your_key
# ANTHROPIC_API_KEY=sk-ant-your_key

STEP 5: Deploy Order

A. Before Redeploying - Set These First:

1. ✅ Change SURREAL_URL to use 127.0.0.1
2. ✅ Add INTERNAL_API_URL=http://127.0.0.1:5055
3. ✅ Add GOOGLE_API_KEY (your Gemini key)
4. ✅ Add GROQ_API_KEY (your Groq key)
5. ⏸️ Keep API_URL as is for now (update after deploy)

B. Push Code Changes:

cd c:\sem6-real\studyrocket\notebookllm\open-notebook
git add .
git commit -m "Fix Railway deployment configuration"
git push origin main

C. After Successful Deploy:

1. ✅ Copy your Railway domain
2. ✅ Update API_URL to your Railway domain
3. ✅ Railway will auto-redeploy

STEP 6: Verification Checklist

After deployment completes, verify:

• Service shows "RUNNING" in Railway
• Check logs: "Application startup complete"
• Check logs: "Migrations completed successfully. Database is now at version 17"
• Visit https://your-domain.up.railway.app/ → Should load UI
• Visit https://your-domain.up.railway.app/api/health → Should return {"status":"ok"}
• Try creating a notebook in the UI
• Test AI features (chat, generation)

Common Issues Specific to Your Setup

Issue: "Database Connection Failed"

Cause: Using localhost instead of 127.0.0.1 Solution: Change SURREAL_URL=ws://127.0.0.1:8000/rpc

Issue: "Unable to Connect to API Server"

Cause: INTERNAL_API_URL not set or API_URL pointing to localhost Solution:

• Set INTERNAL_API_URL=http://127.0.0.1:5055
• Set API_URL=https://your-railway-domain.up.railway.app

Issue: "AI Model Not Available"

Cause: API keys not set or incorrect Solution:

• Verify GOOGLE_API_KEY is set correctly
• Verify GROQ_API_KEY is set correctly
• Check API key validity at provider dashboards

Issue: "Migrations Stuck at Version 14"

Cause: Code changes not deployed Solution:

• Ensure you pushed the latest code with migrations 15-17
• Check Railway logs for migration errors
• Verify all migration files exist in the repo

Model Configuration by Provider

Based on your setup, here's which models you can use:

Via Gemini (GOOGLE_API_KEY)

• ✅ gemini-pro - General purpose
• ✅ gemini-pro-vision - Image understanding
• ✅ gemini-1.5-pro - Long context (1M tokens)
• ✅ gemini-1.5-flash - Fast & efficient
• ✅ Text embeddings via Gemini

Via Groq (GROQ_API_KEY)

• ✅ llama-3.1-70b-versatile - Best Llama model
• ✅ llama-3.1-8b-instant - Fast Llama
• ✅ llama3-70b-8192 - Older Llama version
• ✅ mixtral-8x7b-32768 - Mixtral model
• ✅ gemma2-9b-it - Google's Gemma

Via Ollama (if configured)

• ✅ Any locally installed model
• ✅ llama3:latest, llama3.1:latest
• ✅ mistral:latest, mixtral:latest
• ✅ Custom models

Cost Estimation for Your Setup

Gemini (Google)

• Free Tier: 60 requests/minute
• Paid: $0.50 per 1M input tokens (very affordable)
• Best for: Long context, embeddings, general use

Groq

• Free Tier: Generous free tier
• Paid: Very competitive pricing
• Best for: Fast inference, Llama models

Total Monthly Cost (Estimated)

• Light use (testing): $0-5/month
• Medium use (regular): $10-30/month
• Heavy use (production): $50-100/month

Plus Railway hosting: ~$5-10/month

Next Steps

1. Update variables in Railway as shown above
2. Push code to GitHub
3. Wait for deploy (5-10 minutes)
4. Update API_URL with your Railway domain
5. Test all features with your AI providers

Support

If you encounter issues:

1. Check Railway logs: Dashboard → Deployments → View Logs
2. Look for specific error messages
3. Verify all environment variables are set
4. Test API keys at provider dashboards
5. Join Discord for help: https://discord.gg/37XJPXfz2w

---

Your setup is nearly perfect! Just make the three fixes above (127.0.0.1, INTERNAL_API_URL, and API_URL) and add your AI keys, then you're good to go! 🚀