litellm/tests/e2e_harness/perf
Kerry Lu 3c0e08146d
test(perf): set every perf ceiling to 1.2x the lane median (#47590)
* test(perf): set every perf ceiling to 1.2x the lane median

* test(perf): derive the upgrade verdict test's samples from the case ceilings

* test(perf): keep memory_growth_megabytes at 100 and 600, the metric swings 60 MB between reruns

* test(perf): keep small litellm_overhead_seconds p99 at 1.2x the lane max, it is the 3rd slowest of 300 requests

---------

Co-authored-by: devin-ai-integration[bot] <158243242+devin-ai-integration[bot]@users.noreply.github.com>
2026-10-20 17:19:22 +00:00
..
setup test(perf): measure Admin UI journey latency in a browser on the 1TB SpendLogs restore (#47486) 2026-10-20 16:18:44 +00:00
test_run.py test(perf): qualify database upgrades with shared 5 RPS traffic (#46802) 2026-10-17 11:39:00 -07:00
test_upgrade.py test(perf): set every perf ceiling to 1.2x the lane median (#47590) 2026-10-20 17:19:22 +00:00