UIE Multi-Region Backend

This repository provides a complete guide for setting up a globally distributed backend infrastructure for HubSpot UI Extensions, specifically designed to minimize latency for app cards. The setup consists of two main components:

1. A geo-sharded MongoDB Atlas cluster for distributed data storage
2. A Vercel serverless API deployment leveraging Edge Network regions

Purpose

This infrastructure is designed to complement the guidance provided in the HubSpot Developer Blog article "Minimizing Latency: A Guide to High‑Performance App Cards on HubSpot." The setup enables:

• Data storage closest to your users through MongoDB's geo-sharding
• Low-latency API responses via Vercel's Edge Network
• Globally distributed request handling
• Automatic scaling based on demand

Getting Started

The setup process is divided into two main parts:

1. MongoDB Atlas Setup

   • Installing required CLIs
   • Creating a geo-sharded cluster
   • Configuring network access and database users
   • Setting up collections with global writes
   • Testing data distribution

2. Vercel API Setup

   • Setting up Vercel account and CLI
   • Deploying the serverless API
   • Configuring environment variables
   • Testing the API endpoints

Cost Considerations

Please note that this setup involves costs:

• MongoDB Atlas uses M30 dedicated clusters
• Vercel may require a Pro/Enterprise plan for multiple regions
• Remember to pause or terminate clusters when not in use

Prerequisites

• MongoDB Atlas account
• Vercel account
• Node.js and npm installed
• Basic understanding of MongoDB and REST APIs

Next Steps

After completing the setup, you can:

1. Use the Vercel API endpoint in your HubSpot UI Extension
2. Implement the API calls using hubspot.fetch()
3. Monitor performance across different regions