Skip to content

README.md

Latest commit

 

History

History
86 lines (64 loc) · 3.07 KB

README.md

File metadata and controls

86 lines (64 loc) · 3.07 KB

AsyncSeq Performance Benchmarks

This project contains systematic performance benchmarks for FSharp.Control.AsyncSeq using BenchmarkDotNet.

Purpose

These benchmarks provide:

  • Baseline Performance Metrics: Establish current performance characteristics
  • Regression Detection: Detect performance regressions in future changes
  • Optimization Validation: Measure impact of performance improvements
  • Issue Verification: Verify fixes for known performance issues (#35, #57, etc.)

Benchmark Suites

Core Operations (core)

Tests fundamental AsyncSeq operations:

  • unfoldAsync - Core sequence generation
  • replicate - Simple constant generation
  • mapAsync - Common transformation operation
  • chooseAsync - Filtering with async projection

Append Operations (append)

Tests append functionality (previously had memory leaks):

  • ChainedAppends - Sequential append operations
  • MultipleAppends - Batch append operations

Builder Patterns (builder)

Tests computation expression performance (previously O(n²)):

  • RecursiveAsyncSeq - Recursive yield! patterns
  • UnfoldAsyncEquivalent - Optimized equivalent using unfoldAsync

Usage

Build and Run

# Build benchmarks
dotnet build tests/FSharp.Control.AsyncSeq.Benchmarks -c Release

# Run all benchmark suites
dotnet run --project tests/FSharp.Control.AsyncSeq.Benchmarks -c Release

# Run specific suite
dotnet run --project tests/FSharp.Control.AsyncSeq.Benchmarks -c Release -- core
dotnet run --project tests/FSharp.Control.AsyncSeq.Benchmarks -c Release -- append
dotnet run --project tests/FSharp.Control.AsyncSeq.Benchmarks -c Release -- builder

Interpreting Results

BenchmarkDotNet provides detailed metrics including:

  • Mean: Average execution time
  • Error: Standard error of measurements
  • StdDev: Standard deviation
  • Gen0/Gen1/Gen2: Garbage collection counts
  • Allocated: Memory allocations per operation

Key Performance Indicators

Watch for:

  • Linear scaling: Execution time should scale linearly with element count
  • Low GC pressure: Minimal Gen0/Gen1 collections for streaming operations
  • Constant memory: No memory leaks in long-running operations
  • No O(n²) patterns: Performance should not degrade quadratically

Integration with Performance Plan

These benchmarks support the performance improvement plan phases:

Round 1 (Foundation):

  • ✅ Systematic benchmarking infrastructure (this project)
  • ✅ Baseline performance documentation
  • 🔄 Performance regression detection

Round 2-4:

  • Use these benchmarks to measure optimization impact
  • Add new benchmarks for additional performance improvements
  • Validate performance targets and success metrics

Known Issues Being Monitored

  • Issue #35: Memory leaks in append operations (fixed in PR #193)
  • Issue #57: O(n²) performance in recursive patterns (fixed in 2016)
  • Issue #50: Excessive append function calls
  • Issues #74, #95: Parallelism inefficiencies

Results from these benchmarks help verify these issues remain resolved and detect any regressions.