zwitschi
cc11082ea7
CI / lint-test-build (push) Failing after 19s
- Added synthetic latency profiler scenarios and CLI scripts for baseline generation and regression checks. - Introduced latency baseline and threshold artifacts for CI enforcement. - Enhanced CI workflow with latency guardrail checks. - Updated documentation to include latency profiling commands and performance metrics. - Added unit tests for latency guardrail evaluation.
Performance Hardening
This folder contains latency profiling baselines and guardrail thresholds used in CI.
Scenarios
The profiler covers representative load patterns:
book_update_burst: rapid market-data deltas with moderate detection load.execution_spike: heavier detection/execution pressure.reconnect_storm: frequent reconnect/reset behavior.
Profiling Commands
Generate a fresh profile:
python scripts/profile_latency.py --iterations 600 --output ops/performance/latency_baseline.json
Check current performance against the baseline and thresholds:
python scripts/check_latency_regression.py \
--baseline ops/performance/latency_baseline.json \
--thresholds ops/performance/latency_thresholds.json \
--iterations 600
CI executes the same guardrail check.
Baseline Snapshot (2026-06-01)
Key end-to-end latency baselines from latency_baseline.json:
book_update_burst: p95 = 0.0132 ms, p99 = 0.0198 msexecution_spike: p95 = 0.0139 ms, p99 = 0.0177 msreconnect_storm: p95 = 0.0114 ms, p99 = 0.0134 ms
Optimization Note
MetricsCalculator.compute() was optimized to use DuckDB SQL aggregations and quantiles, reducing Python-side row scans.
Measured benchmark (scripts/benchmark_metrics_compute.py):
- Python scan baseline: 12.623 ms
- SQL aggregate implementation: 11.039 ms
- Speedup: 1.14x