mirror of
https://github.com/BerriAI/litellm.git
synced 2026-09-14 23:21:35 +00:00
Review feedback: the access check called _get_container_owner then _get_container_team_id, two separately-evictable reads where the second was only assumed to be a cache hit off the first. Merge them into _get_container_owner_and_team so a miss on either triggers exactly one refetch that repopulates both. Add a regression test asserting a cold-cache access check issues a single query, and document that the teammate list-cache freshness is a TTL-bounded, view-only tradeoff that never affects the access decision. |
||
|---|---|---|
| .. | ||
| __init__.py | ||
| test_azure_container_transformation.py | ||
| test_container_api.py | ||
| test_container_handler_url.py | ||
| test_container_integration.py | ||
| test_container_proxy_ownership.py | ||
| test_container_regional_api_base.py | ||
| test_container_transformation.py | ||
| test_container_utils.py | ||