diff --git a/tests/code_coverage_tests/test_provider_cache.py b/tests/code_coverage_tests/test_provider_cache.py index ac1ccd5692b..c24ba8d6221 100644 --- a/tests/code_coverage_tests/test_provider_cache.py +++ b/tests/code_coverage_tests/test_provider_cache.py @@ -33,7 +33,13 @@ from provider_cache import ( successful_response, ) from provider_cache_redis import PUBLISH, RedisCommands, RedisResponseStore, configured_cache, redis_store -from provider_cache_routing import LIVE_PROVIDER_REQUIRED, route_cache_model +from provider_cache_routing import ( + BEDROCK_CROSS_REGION_PREFIX, + BEDROCK_EDGE_MODELS, + LIVE_PROVIDER_REQUIRED, + bedrock_region, + route_cache_model, +) from fixture_mode import SESSION_TEST_KEY from provider_edge import EDGE_MOUNTS, configured_cache_backend, resolve_mount, start_provider_edge from provider_edge_bedrock import bedrock_signer @@ -856,6 +862,31 @@ def test_bedrock_deployments_the_edge_must_not_touch_keep_their_direct_route(par assert routed is params +@pytest.mark.parametrize("declared,expected", [ + (None, "us-east-1"), + ("us-west-2", "us-west-2"), + ("eu-west-1", "eu-west-1"), + ("os.environ/AWS_REGION", "us-east-1"), + ("os.environ/ANY_OTHER_NAME", "us-east-1"), +]) +def test_a_region_only_the_proxy_can_resolve_falls_back_to_the_default_mount( + declared: str | None, expected: str, +) -> None: + """A declared literal region is the one the deployment meant. A region the + proxy resolves from its own environment is one the run pod cannot see, and + the default mount answers it.""" + assert bedrock_region(declared) == expected + + +def test_every_model_on_the_edge_allowlist_is_a_cross_region_profile() -> None: + """Answering an env-referenced region with the default mount is only correct + for a profile that fans out across the US regions and is reachable from any + of them. A single-region model on this list would be sent to a region it may + not exist in, so the list is where that is caught.""" + assert BEDROCK_EDGE_MODELS + assert all(model.startswith(BEDROCK_CROSS_REGION_PREFIX) for model in BEDROCK_EDGE_MODELS) + + @pytest.mark.parametrize("mode", ["batch", "realtime", "image_generation"]) def test_a_bedrock_deployment_with_a_mode_keeps_its_direct_route(mode: ModelMode) -> None: params: Final = LiteLLMParamsBody(model="bedrock/us.anthropic.claude-haiku-4-5-20251001-v1:0") diff --git a/tests/e2e/provider_cache_routing.py b/tests/e2e/provider_cache_routing.py index 97e05344423..f9775a2b152 100644 --- a/tests/e2e/provider_cache_routing.py +++ b/tests/e2e/provider_cache_routing.py @@ -21,22 +21,18 @@ BEDROCK_EDGE_MODELS: Final = frozenset( ENV_REFERENCE_PREFIX: Final = "os.environ/" -def bedrock_region(declared: str | None, model: str) -> str | None: - """The region whose edge mount a deployment belongs to, or None when the - harness cannot know it. +def bedrock_region(declared: str | None) -> str: + """The region whose edge mount a deployment belongs to. - Most Bedrock deployments declare `os.environ/AWS_REGION`, which the proxy - resolves from its own environment. The run pod does not share that - environment, so the harness genuinely does not know the region. A `us.` - inference profile fans out across the US regions and is reachable from any - of them, so the default entry point is correct for those whatever the proxy - resolved; anything else keeps its direct path rather than being sent to a - region the model may not exist in.""" - if declared is None: + Most Bedrock deployments declare `os.environ/AWS_REGION`, which only the + proxy can resolve from its own environment; the run pod does not share it. + Answering those with the default mount is correct because every model on the + edge allowlist is a `us.` inference profile, which fans out across the US + regions and is reachable from any of them. That invariant is enforced on the + allowlist itself rather than re-checked per call.""" + if declared is None or declared.startswith(ENV_REFERENCE_PREFIX): return DEFAULT_BEDROCK_REGION - if not declared.startswith(ENV_REFERENCE_PREFIX): - return declared - return DEFAULT_BEDROCK_REGION if model.startswith(BEDROCK_CROSS_REGION_PREFIX) else None + return declared def bedrock_mount(params: LiteLLMParamsBody) -> str | None: @@ -51,8 +47,7 @@ def bedrock_mount(params: LiteLLMParamsBody) -> str | None: model: Final = route.partition("/")[2] or route if model not in BEDROCK_EDGE_MODELS: return None - region: Final = bedrock_region(params.aws_region_name, model) - return None if region is None else f"bedrock/{region}" + return f"bedrock/{bedrock_region(params.aws_region_name)}" def route_bedrock(