Buffer Size Constraints / Large Data Hang (Critical Blocking Issue)

[LukasGPC]

The buffer appears to only allow about 50K of data to be returned in a large query causing hangs and connection pooling polution:

From my internal document

• ✅ Found PRIMARY BOTTLENECK: Hardcoded 200-row fetch limit (sql.go:401)
  • Issue SQL Functions (UPPER/LOWER) + JOINs Cause Hang **Severity:** ⚠️ CRITICAL #6 (Large data hang): Hardcoded 200-row limit + no timeout

From my API Code:

1. Buffer Size Constants:
   • SQL Chunk Size: 31,904 bytes (message.go:101) - NOT the issue
   • Max Rows Per Fetch: 200 (sql.go:401) - PRIMARY BOTTLENECK
   • Query Timeout: 10 seconds (sql.go:400) - doesn't protect data fetching
   • List Item Boundary: 253 bytes (listitem.go:82) - already fixed

Working on returning a result set on my project table:

2. Root Cause of Issue SQL Functions (UPPER/LOWER) + JOINs Cause Hang **Severity:** ⚠️ CRITICAL #6 (Large data hang):
   • Product table: 8× VARCHAR(1024) = ~8KB per row
   • 200 rows × 8KB = 1.6MB theoretical
   • With 2 JOINs: 2-3MB actual
   • Server can't send oversized batch → hangs at row 9-15
   • Driver waits forever (no timeout) → infinite hang

Recommendations:

• sql.go:401 - Make maxRows configurable
• sql.go:129-143 - Add timeout to fetchMoreData()
• mod.go - Add maxRowsPerFetch field to Connection struct
• url.go - Parse maxRows from DSN

[LukasGPC]

This is my number 1 production issues, this issue is also documented in more detail in Issue #6 , Issue #8 and Issue #9