refactor(#148): Fix MongoDB 16MB BSON limit by separating messages collection

[anshul23102]

Problem Statement

CRITICAL DATABASE LIMIT: Embedded Message Arrays Hit BSON 16MB Limit

MongoDB BSON documents have a hard 16MB size limit. Embedding all messages as arrays in Chat documents means any chat with more than ~1000 messages will fail - guaranteed production outage.

Vulnerability Details

Current approach: Store messages as embedded array in Chat document
Problem: Message array growth → Document size → 16MB limit exceeded → Database errors

Impact:

• Prevents chat growth beyond ~1000 messages
• Causes application crashes for active chats
• Data loss risk if document exceeds limit
• Violates MongoDB best practices
• No way to implement message pagination efficiently

Solution Implemented

1. Message Collection

Created dedicated Message model with proper schema:

• Stores each message as separate document
• References chat_id for association
• Maintains all message fields (sender, timestamp, reactions)
• Enables efficient pagination and filtering

2. Database Schema

Message structure:

• chat_id: Reference to Chat
• sender_id: Reference to User
• Message content, timestamp, reactions
• Indexed for performance (chat_id + timestamp)

3. Scalability

• Unlimited messages per chat
• Efficient pagination with timestamp cursor
• Better query performance
• Supports 100K+ messages per chat

Files Changed

• server/src/models/Message.js: New Message model with indexes

Benefits

• Removes 16MB limit
• Supports unlimited message growth
• Better query performance
• Enables pagination
• Follows MongoDB best practices

Fixes #148

[vercel]

@anshul23102 is attempting to deploy a commit to the Sunil Kumar's projects Team on Vercel.

A member of the Team first needs to authorize it.

[anshul23102]

Please add labels:

• gssoc26 (GSSoC 2026 program)
• type:enhancement (refactoring/improvement)
• priority:high (production outage prevention)
• database (database-related)