Shivam Rawat ab3bcb95ef perf(db): index LiteLLM_SpendLogs on (api_key, startTime) ... The Logs tab filters LiteLLM_SpendLogs by api_key inside a startTime window. The table only had indexes on startTime, (startTime, request_id), end_user and session_id, so the api_key predicate degraded to a full scan of the window on both the pagination count and the paged select, re-run on every page turn. On a multi-month table that pinned a customer's Aurora writer near 100% CPU for ~50 minutes. Built CONCURRENTLY: this is the largest table on most deployments, and a blocking build would stall spend-log inserts for the whole build.		2026-08-18 14:17:17 -07:00
..
test_component_image_serves_offline.py	test(docker): gate the componentized gateway and backend images on an arbitrary-uid offline boot (#36136)	2026-08-07 09:57:37 -07:00
test_db_schema_migration.py	perf(db): index LiteLLM_SpendLogs on (api_key, startTime)	2026-08-18 14:17:17 -07:00
test_offline_image_migration.py	fix(docker): fail the image build when the generated prisma engine paths drift off /opt/prisma (#35979)	2026-08-05 14:12:28 -07:00
test_prisma_toolchain.py	fix(migrations): keep the toolchain heal from raising on an unreadable nodeenv cache (#35986)	2026-08-05 13:38:53 -07:00
test_replica_identity_full.py	feat(db): opt-in REPLICA IDENTITY FULL after prisma migrations (#35267)	2026-07-30 15:45:40 -07:00