Super tiny move error.rs (#14944)

sgl-model-gateway/src/observability/metrics.rs

@@ -318,8 +318,7 @@ impl RouterMetrics {
     }
 
     pub fn record_request_duration(duration: Duration) {
-        histogram!("sgl_router_request_duration_seconds")
+        histogram!("sgl_router_request_duration_seconds").record(duration.as_secs_f64());
-        .record(duration.as_secs_f64());
     }
 
     pub fn record_request_error(route: &str, error_type: &str) {

sgl-model-gateway/src/routers/grpc/common/response_collection.rs

@@ -5,8 +5,9 @@
 
 use axum::response::Response;
 
-use crate::routers::grpc::{
+use crate::routers::{
-    context::ExecutionResult, error, proto_wrapper::ProtoGenerateComplete, utils,
+    error,
+    grpc::{context::ExecutionResult, proto_wrapper::ProtoGenerateComplete, utils},
 };
 
 /// Collect and merge responses from execution result

sgl-model-gateway/src/routers/grpc/common/responses/utils.rs

@@ -18,7 +18,7 @@ use crate::{
         responses::{ResponseTool, ResponseToolType, ResponsesRequest, ResponsesResponse},
     },
     routers::{
-        grpc::error,
+        error,
         openai::{conversations::persist_conversation_items, mcp::ensure_request_mcp_client},
     },
 };

sgl-model-gateway/src/routers/grpc/common/stages/client_acquisition.rs

@@ -5,9 +5,12 @@ use axum::response::Response;
 use tracing::error;
 
 use super::PipelineStage;
-use crate::routers::grpc::{
+use crate::routers::{
-    context::{ClientSelection, RequestContext, WorkerSelection},
+    error,
-    error, utils,
+    grpc::{
+        context::{ClientSelection, RequestContext, WorkerSelection},
+        utils,
+    },
 };
 
 /// Client acquisition stage: Get gRPC clients from selected workers

sgl-model-gateway/src/routers/grpc/common/stages/dispatch_metadata.rs

@@ -7,9 +7,9 @@ use axum::response::Response;
 use tracing::error;
 
 use super::PipelineStage;
-use crate::routers::grpc::{
+use crate::routers::{
-    context::{DispatchMetadata, RequestContext, RequestType, WorkerSelection},
     error,
+    grpc::context::{DispatchMetadata, RequestContext, RequestType, WorkerSelection},
 };
 
 /// Dispatch metadata stage: Prepare metadata for dispatch

sgl-model-gateway/src/routers/grpc/common/stages/request_execution.rs

@@ -5,10 +5,12 @@ use axum::response::Response;
 use tracing::{error, info_span, Instrument};
 
 use super::PipelineStage;
-use crate::routers::grpc::{
+use crate::routers::{
-    context::{ClientSelection, ExecutionResult, RequestContext},
     error,
-    proto_wrapper::{ProtoGenerateRequest, ProtoStream},
+    grpc::{
+        context::{ClientSelection, ExecutionResult, RequestContext},
+        proto_wrapper::{ProtoGenerateRequest, ProtoStream},
+    },
 };
 
 type StreamResult = Result<ProtoStream, Box<dyn std::error::Error + Send + Sync>>;

sgl-model-gateway/src/routers/grpc/common/stages/worker_selection.rs

@@ -10,9 +10,9 @@ use super::PipelineStage;
 use crate::{
     core::{ConnectionMode, Worker, WorkerRegistry, WorkerType},
     policies::PolicyRegistry,
-    routers::grpc::{
+    routers::{
-        context::{RequestContext, WorkerSelection},
         error,
+        grpc::context::{RequestContext, WorkerSelection},
     },
 };
 

sgl-model-gateway/src/routers/grpc/harmony/processor.rs

@@ -16,10 +16,12 @@ use crate::{
             ResponseStatus, ResponseUsage, ResponsesRequest, ResponsesResponse, ResponsesUsage,
         },
     },
-    routers::grpc::{
+    routers::{
-        common::{response_collection, response_formatting},
-        context::{DispatchMetadata, ExecutionResult},
         error,
+        grpc::{
+            common::{response_collection, response_formatting},
+            context::{DispatchMetadata, ExecutionResult},
+        },
     },
 };
 

sgl-model-gateway/src/routers/grpc/harmony/responses.rs

@@ -54,15 +54,17 @@ use crate::{
             ResponsesUsage, StringOrContentParts,
         },
     },
-    routers::grpc::{
+    routers::{
-        common::responses::{
-            build_sse_response, ensure_mcp_connection, persist_response_if_needed,
-            streaming::{OutputItemType, ResponseStreamEventEmitter},
-        },
-        context::SharedComponents,
         error,
-        harmony::{processor::ResponsesIterationResult, streaming::HarmonyStreamingProcessor},
+        grpc::{
-        pipeline::RequestPipeline,
+            common::responses::{
+                build_sse_response, ensure_mcp_connection, persist_response_if_needed,
+                streaming::{OutputItemType, ResponseStreamEventEmitter},
+            },
+            context::SharedComponents,
+            harmony::{processor::ResponsesIterationResult, streaming::HarmonyStreamingProcessor},
+            pipeline::RequestPipeline,
+        },
     },
 };
 

sgl-model-gateway/src/routers/grpc/harmony/stages/preparation.rs

@@ -12,10 +12,13 @@ use crate::{
         common::{Tool, ToolChoice, ToolChoiceValue},
         responses::ResponsesRequest,
     },
-    routers::grpc::{
+    routers::{
-        common::{responses::utils::extract_tools_from_response_tools, stages::PipelineStage},
+        error,
-        context::{PreparationOutput, RequestContext, RequestType},
+        grpc::{
-        error, utils,
+            common::{responses::utils::extract_tools_from_response_tools, stages::PipelineStage},
+            context::{PreparationOutput, RequestContext, RequestType},
+            utils,
+        },
     },
 };
 

sgl-model-gateway/src/routers/grpc/harmony/stages/request_building.rs

@@ -5,11 +5,13 @@ use axum::response::Response;
 use tracing::{debug, error};
 use uuid::Uuid;
 
-use crate::routers::grpc::{
+use crate::routers::{
-    common::stages::{helpers, PipelineStage},
-    context::{ClientSelection, RequestContext, RequestType, WorkerSelection},
     error,
-    proto_wrapper::ProtoGenerateRequest,
+    grpc::{
+        common::stages::{helpers, PipelineStage},
+        context::{ClientSelection, RequestContext, RequestType, WorkerSelection},
+        proto_wrapper::ProtoGenerateRequest,
+    },
 };
 
 /// Harmony Request Building stage: Convert Harmony tokens to gRPC request

sgl-model-gateway/src/routers/grpc/harmony/stages/response_processing.rs

@@ -7,10 +7,12 @@ use axum::response::Response;
 use tracing::error;
 
 use super::super::{HarmonyResponseProcessor, HarmonyStreamingProcessor};
-use crate::routers::grpc::{
+use crate::routers::{
-    common::stages::PipelineStage,
-    context::{FinalResponse, RequestContext, RequestType},
     error,
+    grpc::{
+        common::stages::PipelineStage,
+        context::{FinalResponse, RequestContext, RequestType},
+    },
 };
 
 /// Harmony Response Processing stage: Parse and format Harmony responses

sgl-model-gateway/src/routers/grpc/mod.rs

@@ -5,7 +5,6 @@ use crate::{grpc_client::sglang_proto::MultimodalInputs, protocols::common::Stri
 pub mod client;
 pub mod common;
 pub mod context;
-pub mod error;
 pub mod harmony;
 pub mod pd_router;
 pub mod pipeline;

sgl-model-gateway/src/routers/grpc/pipeline.rs

@@ -11,7 +11,7 @@ use tracing::error;
 use super::{
     common::stages::*,
     context::*,
-    error, harmony,
+    harmony,
     regular::{processor, stages::*, streaming},
 };
 use crate::{
@@ -22,6 +22,7 @@ use crate::{
         generate::GenerateRequest,
     },
     reasoning_parser::ParserFactory as ReasoningParserFactory,
+    routers::error,
     tokenizer::traits::Tokenizer,
     tool_parser::ParserFactory as ToolParserFactory,
 };

sgl-model-gateway/src/routers/grpc/regular/processor.rs

@@ -16,12 +16,14 @@ use crate::{
         generate::{GenerateMetaInfo, GenerateRequest, GenerateResponse},
     },
     reasoning_parser::ParserFactory as ReasoningParserFactory,
-    routers::grpc::{
+    routers::{
-        common::{response_collection, response_formatting},
-        context::{DispatchMetadata, ExecutionResult},
         error,
-        proto_wrapper::ProtoGenerateComplete,
+        grpc::{
-        utils,
+            common::{response_collection, response_formatting},
+            context::{DispatchMetadata, ExecutionResult},
+            proto_wrapper::ProtoGenerateComplete,
+            utils,
+        },
     },
     tokenizer::{
         stop::{SequenceDecoderOutput, StopSequenceDecoder},

sgl-model-gateway/src/routers/grpc/regular/responses/handlers.rs

@@ -64,12 +64,12 @@ use crate::{
             ResponsesUsage,
         },
     },
-    routers::grpc::{
+    routers::{
-        common::responses::{
+        error,
+        grpc::common::responses::{
             build_sse_response, ensure_mcp_connection, persist_response_if_needed,
             streaming::ResponseStreamEventEmitter,
         },
-        error,
     },
 };
 

sgl-model-gateway/src/routers/grpc/regular/responses/tool_loop.rs

@@ -34,9 +34,9 @@ use crate::{
             ResponsesResponse,
         },
     },
-    routers::grpc::{
+    routers::{
-        common::responses::streaming::{OutputItemType, ResponseStreamEventEmitter},
         error,
+        grpc::common::responses::streaming::{OutputItemType, ResponseStreamEventEmitter},
     },
 };
 

sgl-model-gateway/src/routers/grpc/regular/stages/chat/preparation.rs

@@ -8,10 +8,13 @@ use tracing::error;
 
 use crate::{
     protocols::chat::ChatCompletionRequest,
-    routers::grpc::{
+    routers::{
-        common::stages::PipelineStage,
+        error,
-        context::{PreparationOutput, RequestContext},
+        grpc::{
-        error, utils,
+            common::stages::PipelineStage,
+            context::{PreparationOutput, RequestContext},
+            utils,
+        },
     },
 };
 

sgl-model-gateway/src/routers/grpc/regular/stages/chat/request_building.rs

@@ -5,12 +5,14 @@ use axum::response::Response;
 use tracing::error;
 use uuid::Uuid;
 
-use crate::routers::grpc::{
+use crate::routers::{
-    client::GrpcClient,
-    common::stages::{helpers, PipelineStage},
-    context::{ClientSelection, RequestContext, WorkerSelection},
     error,
-    proto_wrapper::ProtoGenerateRequest,
+    grpc::{
+        client::GrpcClient,
+        common::stages::{helpers, PipelineStage},
+        context::{ClientSelection, RequestContext, WorkerSelection},
+        proto_wrapper::ProtoGenerateRequest,
+    },
 };
 
 /// Chat request building stage

sgl-model-gateway/src/routers/grpc/regular/stages/chat/response_processing.rs

@@ -9,11 +9,13 @@ use async_trait::async_trait;
 use axum::response::Response;
 use tracing::error;
 
-use crate::routers::grpc::{
+use crate::routers::{
-    common::stages::PipelineStage,
-    context::{FinalResponse, RequestContext},
     error,
-    regular::{processor, streaming},
+    grpc::{
+        common::stages::PipelineStage,
+        context::{FinalResponse, RequestContext},
+        regular::{processor, streaming},
+    },
 };
 
 /// Chat response processing stage

sgl-model-gateway/src/routers/grpc/regular/stages/generate/preparation.rs

@@ -8,10 +8,13 @@ use tracing::error;
 
 use crate::{
     protocols::{common::InputIds, generate::GenerateRequest},
-    routers::grpc::{
+    routers::{
-        common::stages::PipelineStage,
+        error,
-        context::{PreparationOutput, RequestContext},
+        grpc::{
-        error, utils,
+            common::stages::PipelineStage,
+            context::{PreparationOutput, RequestContext},
+            utils,
+        },
     },
     tokenizer::traits::Tokenizer,
 };

sgl-model-gateway/src/routers/grpc/regular/stages/generate/request_building.rs

@@ -5,12 +5,14 @@ use axum::response::Response;
 use tracing::error;
 use uuid::Uuid;
 
-use crate::routers::grpc::{
+use crate::routers::{
-    client::GrpcClient,
-    common::stages::{helpers, PipelineStage},
-    context::{ClientSelection, RequestContext, WorkerSelection},
     error,
-    proto_wrapper::ProtoGenerateRequest,
+    grpc::{
+        client::GrpcClient,
+        common::stages::{helpers, PipelineStage},
+        context::{ClientSelection, RequestContext, WorkerSelection},
+        proto_wrapper::ProtoGenerateRequest,
+    },
 };
 
 /// Generate request building stage

sgl-model-gateway/src/routers/grpc/regular/stages/generate/response_processing.rs

@@ -6,11 +6,13 @@ use async_trait::async_trait;
 use axum::response::Response;
 use tracing::error;
 
-use crate::routers::grpc::{
+use crate::routers::{
-    common::stages::PipelineStage,
-    context::{FinalResponse, RequestContext},
     error,
-    regular::{processor, streaming},
+    grpc::{
+        common::stages::PipelineStage,
+        context::{FinalResponse, RequestContext},
+        regular::{processor, streaming},
+    },
 };
 
 /// Generate response processing stage

sgl-model-gateway/src/routers/grpc/regular/stages/preparation.rs

@@ -8,10 +8,12 @@ use axum::response::Response;
 use tracing::error;
 
 use super::{chat::ChatPreparationStage, generate::GeneratePreparationStage};
-use crate::routers::grpc::{
+use crate::routers::{
-    common::stages::PipelineStage,
-    context::{RequestContext, RequestType},
     error as grpc_error,
+    grpc::{
+        common::stages::PipelineStage,
+        context::{RequestContext, RequestType},
+    },
 };
 
 /// Preparation stage (delegates to endpoint-specific implementations)

sgl-model-gateway/src/routers/grpc/regular/stages/request_building.rs

@@ -5,10 +5,12 @@ use axum::response::Response;
 use tracing::error;
 
 use super::{chat::ChatRequestBuildingStage, generate::GenerateRequestBuildingStage};
-use crate::routers::grpc::{
+use crate::routers::{
-    common::stages::PipelineStage,
-    context::{RequestContext, RequestType},
     error as grpc_error,
+    grpc::{
+        common::stages::PipelineStage,
+        context::{RequestContext, RequestType},
+    },
 };
 
 /// Request building stage (delegates to endpoint-specific implementations)

sgl-model-gateway/src/routers/grpc/regular/stages/response_processing.rs

@@ -7,11 +7,13 @@ use axum::response::Response;
 use tracing::error;
 
 use super::{chat::ChatResponseProcessingStage, generate::GenerateResponseProcessingStage};
-use crate::routers::grpc::{
+use crate::routers::{
-    common::stages::PipelineStage,
-    context::{RequestContext, RequestType},
     error,
-    regular::{processor, streaming},
+    grpc::{
+        common::stages::PipelineStage,
+        context::{RequestContext, RequestType},
+        regular::{processor, streaming},
+    },
 };
 
 /// Response processing stage (delegates to endpoint-specific implementations)

sgl-model-gateway/src/routers/grpc/utils.rs

@@ -9,7 +9,6 @@ use uuid::Uuid;
 
 use super::{
     client::GrpcClient,
-    error,
     proto_wrapper::{ProtoGenerateComplete, ProtoStream},
     ProcessedMessages,
 };
@@ -28,7 +27,7 @@ use crate::{
         ParserFactory as ReasoningParserFactory, PooledParser as ReasoningPooledParser,
         ReasoningParser,
     },
-    routers::grpc::proto_wrapper::ProtoResponseVariant,
+    routers::{error, grpc::proto_wrapper::ProtoResponseVariant},
     tokenizer::{
         cache::CachedTokenizer,
         chat_template::{ChatTemplateContentFormat, ChatTemplateParams},

sgl-model-gateway/src/routers/mod.rs

@@ -21,6 +21,7 @@ use crate::protocols::{
 };
 
 pub mod conversations;
+pub mod error;
 pub mod factory;
 pub mod grpc;
 pub mod header_utils;