avoid memory allocation from the heap class.

[yu-shipit]

Pull Request checklist

• The commit(s) message(s) follows the contribution guidelines ?
• Tests for the changes have been added (for bug fixes / features) ?
• Docs have been added / updated (for bug fixes / features) ?

🚀 Performance Optimization: Avoid Memory Allocation from Heap Class

Commit Details

• Commit Hash: a86d5bddb7fbebc4aea4dc29ae486e2adf7eb190
• Branch: integration (HEAD → integration, origin/integration)
• Author: yu-shipit 120689245+yu-shipit@users.noreply.github.com
• Co-authored by: Yu Zhang yzhang999272@apple.com
• Date: Fri May 8 13:05:46 2026 -0700

📋 Summary

This commit introduces a significant performance optimization by removing helper functions that caused unnecessary memory allocations and replacing them with inline NaN handling logic. This change improves performance in time-series aggregation functions while maintaining correct NaN handling behavior.

🔧 Changes Made

Primary Changes

1. ✅ Removed Helper Functions to Avoid Memory Allocation

Eliminated three helper functions from QueryUtils that were causing heap allocations:

• maxIgnoreNaN(a: Double, aTs: Long, b: Double, bTs: Long): (Double, Long)
• minIgnoreNaN(a: Double, aTs: Long, b: Double, bTs: Long): (Double, Long)
• lastIgnoreNaN(a: Double, aTs: Long, b: Double, bTs: Long): (Double, Long)

Why this improves performance:

• Each function call created tuple objects (Double, Long) on the heap
• In high-frequency aggregation operations, this caused significant GC pressure
• Removing these functions eliminates millions of temporary object allocations

2. ✅ Inlined NaN Handling Logic

Replaced function calls with direct inline logic in MinOverTimeChunkedFunctionD:

Before (heap allocation):

val (resultValue, resultTimestamp) = QueryUtils.minIgnoreNaN(min, minTimestamp, nextVal, nextTimestamp)
min = resultValue
minTimestamp = resultTimestamp

After (no allocation):

// Handle NaN values properly for timestamp functions
if (min != min) { // min is NaN
  min = nextVal
  minTimestamp = nextTimestamp
} else if (nextVal != nextVal) { // nextVal is NaN, keep current min
  // Do nothing - skip NaN values
} else if (nextVal <= min) {
  min = nextVal
  minTimestamp = nextTimestamp
}

BREAKING CHANGES

If this PR contains a breaking change, please describe the impact and migration
path for existing applications.
If not please remove this section.

Breaking changes may include:

• Any schema changes to any Cassandra tables
• The serialized format for Dataset and Column (see .toString methods)
• Over the wire formats for Akka messages / case classes
• Changes to the HTTP public API
• Changes to query parsing / PromQL parsing

Other information: