mirror of
https://github.com/BerriAI/litellm.git
synced 2026-09-15 23:31:29 +00:00
A ratio against the healthy phase cannot bound either metric. Once the Redis circuit breaker opens, a request skips Redis instead of waiting on its socket timeout, so the chaos phase can measure cheaper than the baseline it is compared against: local runs came in at 0.61x baseline p90 while a log-bytes ratio read 724x. Splitting Budget into RatioBudget and AbsoluteBudget lets RSS and CPU keep the ratio they need, since both are machine-shaped, while latency and log volume get the wall-clock ceiling a user actually cares about. Co-Authored-By: Claude Code <noreply@anthropic.com> |
||
|---|---|---|
| .. | ||
| conftest.py | ||
| load_client.py | ||
| locust_load.py | ||
| locustfile.py | ||
| phase_budget.py | ||
| proxy_usage.py | ||
| session_anomaly.py | ||
| test_locust_load.py | ||
| test_phase_budget.py | ||
| test_proxy_usage.py | ||
| test_redis_chaos_e2e.py | ||
| test_session_anomaly.py | ||
| test_weekly_session_anomaly_e2e.py | ||
| weekly_anomaly_config.yml | ||