mirror of
https://github.com/BerriAI/litellm.git
synced 2026-10-12 03:48:56 +00:00
ccf866c801
* fix(router): await budget redis pipeline before sync reads * refactor(router): remove superseded Redis flush helper * fix(router): preserve concurrent spend during Redis synchronization * fix(router): type per-key spend totals without loop Final bindings * fix(router): log Redis failures before cancellable cleanup * fix(router): finalize Redis batches before propagating cancellation * test(router): reproduce cancellation while Redis cleanup is blocked * test: align budget hotpath checks with awaited Redis flush * fix(budgets): finish Redis flush after cancellation while queued * test(budgets): consolidate Redis regressions in mapped tests * fix(budgets): coalesce failed Redis increments by key * test(budgets): assert spend behavior instead of batch state * perf(router): sum pending budget spend by key once per sync --------- Co-authored-by: Emerson Gomes <emerson.gomes@thalesgroup.com> Co-authored-by: mateo-berri <277851410+mateo-berri@users.noreply.github.com> |
||
|---|---|---|
| .. | ||
| a2a_protocol | ||
| anthropic_interface | ||
| batches | ||
| caching | ||
| chat_completions | ||
| completion_extras | ||
| compression | ||
| endpoints | ||
| enterprise | ||
| gateway | ||
| integration_support | ||
| integrations | ||
| litellm_core_utils | ||
| litellm_proxy_extras | ||
| llms | ||
| messages | ||
| models | ||
| ocr | ||
| passthrough | ||
| proxy | ||
| rag | ||
| realtime_api | ||
| repositories | ||
| responses | ||
| router_strategy | ||
| router_utils | ||
| rust_bridge | ||
| sandbox | ||
| skills | ||
| test_router | ||
| types | ||
| videos | ||
| __init__.py | ||
| AGENTS.md | ||
| conftest.py | ||
| test_package_layout.py | ||
| test_socket_policy.py | ||