docs: add connection pool exhaustion troubleshooting guide

[Rohan170603]

Problem

The connection pooling documentation explains what limits exist (max_client_conn, default_pool_size, max_connections) and how pooling works, but it doesn't provide a practical troubleshooting guide for developers who hit pool exhaustion in production.

Developers experiencing API 500s, timeouts at 28-32 seconds, and multiple endpoint failures have no clear path to diagnose and fix the issue. One developer reported this exact scenario with their Neon PostgreSQL connection pool.

Solution

Add a "Troubleshooting connection pool exhaustion" section that provides:

• Symptoms – Real error messages and failure patterns
• Step 1: Check connection count – SQL query + Monitoring page link
• Step 2: Determine which limit – Error-to-limit mapping table
• Step 3: Identify the cause – Common causes (connection leaks, pool size, serverless, multiple pools, compute size)
• Step 4: Fix common issues – Code examples, config table, Prisma guidance
• Prevention – 7 actionable best practices

Impact

• Developers can quickly diagnose and fix pool exhaustion issues
• Reduces support burden for connection-related problems
• Prevents production outages from connection limits

[vercel]

@Rohan170603 is attempting to deploy a commit to the neondatabase Team on Vercel.

A member of the Team first needs to authorize it.