[performance] Regression in CompileMemoryUsage: +103.7% slower #22277
Description
📊 Performance Regression Detected
Benchmark: BenchmarkCompileMemoryUsage
Current Performance: 13,105,483 ns/op (≈13.1ms)
Historical Average: 6,433,478 ns/op (≈6.4ms)
Change: +103.7% slower
Pattern: Sharp single-day jump (was 6,256,424 ns/op on 2026-03-21)
📈 Detailed Performance Metrics
Performance Comparison
| Date | ns/op | Change |
|---|---|---|
| 2026-03-15 | 6,447,942 | baseline |
| 2026-03-16 | 6,125,299 | -5.0% |
| 2026-03-17 | 6,157,707 | -4.5% |
| 2026-03-18 | 6,842,977 | +6.1% |
| 2026-03-19 | 6,579,200 | +2.0% |
| 2026-03-20 | 6,624,798 | +2.7% |
| 2026-03-21 | 6,256,424 | -3.0% |
| 2026-03-22 | 13,105,483 | +103.7% 🚨 |
- Memory: 5,382,410 B/op
- Allocations: 74,964 allocs/op
- Historical Data Points: 7
- 7-day Average: 6,433,478 ns/op
Baseline Targets
- Memory benchmark target: <500ms ✅ (within absolute target but doubled from baseline)
💡 Recommended Actions
- Review recent commits from the last 24 hours affecting workflow compilation or memory management
- Run
make bench-memoryto generate detailed memory profiles - Check for memory leaks or increased allocation patterns using
go tool pprof -alloc_space - Compare
BenchmarkCompileMemoryUsagetrace with a known-good revision - Note:
CompileComplexWorkflowalso regressed on the same run — likely the same root cause
📋 Additional Context
- Run ID: §23405229136
- Date: 2026-03-22
- Workflow: Daily CLI Performance
- Related regression:
CompileComplexWorkflowalso shows +113.9% regression on the same run - Contrast: Several benchmarks showed major improvements on this same run (CMW -51%, Validation -55.5%), suggesting a possible reorganization of work between compilation phases
References:
Generated by Daily CLI Performance Agent · ◷
- expires on Mar 24, 2026, 2:41 PM UTC
Metadata
Metadata
Type
Fields
Give feedbackNo fields configured for issues without a type.