chenchenghao/sglang
1
0
Fork 0
Code Issues Pull Requests Actions 14 Packages Projects Releases Wiki Activity

[model-gateway] add OTEL integration to grpc router (#14671)

Browse Source
This commit is contained in:
Simo Lin
2025-12-08 14:45:05 -08:00
committed by GitHub
parent 6abb8051e8
commit edde5e5d40
6 changed files with 239 additions and 108 deletions
Download Patch File Download Diff File Expand all files Collapse all files
sgl-model-gateway/src/grpc_client/sglang_scheduler.rs
+14 -7
View File
@@ -12,12 +12,15 @@ use std::{
use tonic::{transport::Channel, Request, Streaming};
use tracing::{debug, warn};
use crate::protocols::{
chat::ChatCompletionRequest,
common::{ResponseFormat, StringOrArray, ToolChoice, ToolChoiceValue},
generate::GenerateRequest,
responses::ResponsesRequest,
sampling_params::SamplingParams as GenerateSamplingParams,
use crate::{
observability::otel_trace::inject_trace_context_grpc,
protocols::{
chat::ChatCompletionRequest,
common::{ResponseFormat, StringOrArray, ToolChoice, ToolChoiceValue},
generate::GenerateRequest,
responses::ResponsesRequest,
sampling_params::SamplingParams as GenerateSamplingParams,
},
};
// Include the generated protobuf code
@@ -163,7 +166,11 @@ impl SglangSchedulerClient {
) -> Result<AbortOnDropStream, Box<dyn std::error::Error + Send + Sync>> {
let request_id = req.request_id.clone();
let mut client = self.client.clone();
let request = Request::new(req);
let mut request = Request::new(req);
// Inject W3C trace context into gRPC metadata for distributed tracing
inject_trace_context_grpc(request.metadata_mut());
let response = client.generate(request).await?;
Ok(AbortOnDropStream::new(
sgl-model-gateway/src/grpc_client/vllm_engine.rs
+14 -7
View File
@@ -12,12 +12,15 @@ use std::{
use tonic::{transport::Channel, Request, Streaming};
use tracing::{debug, warn};
use crate::protocols::{
chat::ChatCompletionRequest,
common::{ResponseFormat, StringOrArray, ToolChoice, ToolChoiceValue},
generate::GenerateRequest,
responses::ResponsesRequest,
sampling_params::SamplingParams as GenerateSamplingParams,
use crate::{
observability::otel_trace::inject_trace_context_grpc,
protocols::{
chat::ChatCompletionRequest,
common::{ResponseFormat, StringOrArray, ToolChoice, ToolChoiceValue},
generate::GenerateRequest,
responses::ResponsesRequest,
sampling_params::SamplingParams as GenerateSamplingParams,
},
};
// Include the generated protobuf code
@@ -163,7 +166,11 @@ impl VllmEngineClient {
) -> Result<AbortOnDropStream, Box<dyn std::error::Error + Send + Sync>> {
let request_id = req.request_id.clone();
let mut client = self.client.clone();
let request = Request::new(req);
let mut request = Request::new(req);
// Inject W3C trace context into gRPC metadata for distributed tracing
inject_trace_context_grpc(request.metadata_mut());
let response = client.generate(request).await?;
Ok(AbortOnDropStream::new(
sgl-model-gateway/src/observability/otel_trace.rs
+30
View File
@@ -17,6 +17,7 @@ use opentelemetry_sdk::{
Resource,
};
use tokio::task::spawn_blocking;
use tonic::metadata::{MetadataKey, MetadataMap, MetadataValue};
use tracing::{Metadata, Subscriber};
use tracing_opentelemetry::{self, OpenTelemetrySpanExt};
use tracing_subscriber::{
@@ -240,3 +241,32 @@ pub fn inject_trace_context_http(headers: &mut HeaderMap) {
propagator.inject_context(&context, &mut HeaderInjector(headers));
});
}
/// Inject W3C trace context into gRPC metadata.
///
/// This propagates the current span context to downstream gRPC services.
/// Does nothing if OTEL is not enabled.
pub fn inject_trace_context_grpc(metadata: &mut MetadataMap) {
if !is_otel_enabled() {
return;
}
let context = tracing::Span::current().context();
struct MetadataInjector<'a>(&'a mut MetadataMap);
impl opentelemetry::propagation::Injector for MetadataInjector<'_> {
fn set(&mut self, key: &str, value: String) {
// gRPC metadata keys must be lowercase ASCII
if let Ok(metadata_key) = MetadataKey::from_bytes(key.to_lowercase().as_bytes()) {
if let Ok(metadata_value) = MetadataValue::try_from(&value) {
self.0.insert(metadata_key, metadata_value);
}
}
}
}
global::get_text_map_propagator(|propagator| {
propagator.inject_context(&context, &mut MetadataInjector(metadata));
});
}
sgl-model-gateway/src/policies/tree.rs
-72
View File
@@ -590,30 +590,6 @@ impl Tree {
.collect()
}
pub fn get_smallest_tenant(&self) -> String {
// Return a placeholder if there are no tenants
if self.tenant_char_count.is_empty() {
return "empty".to_string();
}
// Find the tenant with minimum char count
let mut min_tenant = None;
let mut min_count = usize::MAX;
for entry in self.tenant_char_count.iter() {
let tenant = entry.key();
let count = *entry.value();
if count < min_count {
min_count = count;
min_tenant = Some(tenant.clone());
}
}
// Return the found tenant or "empty" if somehow none was found
min_tenant.unwrap_or_else(|| "empty".to_string())
}
#[allow(dead_code)]
pub fn get_used_size_per_tenant(&self) -> HashMap<String, usize> {
// perform a DFS to traverse all nodes and calculate the total size used by each tenant
@@ -728,54 +704,6 @@ mod tests {
use super::*;
#[test]
fn test_get_smallest_tenant() {
let tree = Tree::new();
assert_eq!(tree.get_smallest_tenant(), "empty");
// Insert data for tenant1 - "ap" + "icot" = 6 chars
tree.insert("ap", "tenant1");
tree.insert("icot", "tenant1");
// Insert data for tenant2 - "cat" = 3 chars
tree.insert("cat", "tenant2");
assert_eq!(
tree.get_smallest_tenant(),
"tenant2",
"Expected tenant2 to be smallest with 3 characters."
);
// Insert overlapping data for tenant3 and tenant4 to test equal counts
// tenant3: "do" = 2 chars
// tenant4: "hi" = 2 chars
tree.insert("do", "tenant3");
tree.insert("hi", "tenant4");
let smallest = tree.get_smallest_tenant();
assert!(
smallest == "tenant3" || smallest == "tenant4",
"Expected either tenant3 or tenant4 (both have 2 characters), got {}",
smallest
);
// Add more text to tenant4 to make it larger
tree.insert("hello", "tenant4"); // Now tenant4 has "hi" + "hello" = 6 chars
// Now tenant3 should be smallest (2 chars vs 6 chars for tenant4)
assert_eq!(
tree.get_smallest_tenant(),
"tenant3",
"Expected tenant3 to be smallest with 2 characters"
);
tree.evict_tenant_by_size(3); // This should evict tenants with more than 3 chars
let post_eviction_smallest = tree.get_smallest_tenant();
println!("Smallest tenant after eviction: {}", post_eviction_smallest);
}
#[test]
fn test_tenant_char_count() {
let tree = Tree::new();
sgl-model-gateway/src/routers/grpc/common/stages/request_execution.rs
+34 -6
View File
@@ -2,7 +2,7 @@
use async_trait::async_trait;
use axum::response::Response;
use tracing::error;
use tracing::{error, info_span, Instrument};
use super::PipelineStage;
use crate::routers::grpc::{
@@ -18,6 +18,7 @@ pub struct RequestExecutionStage {
mode: ExecutionMode,
}
#[derive(Debug, Clone, Copy)]
pub enum ExecutionMode {
/// Regular mode: single worker execution
Single,
@@ -50,12 +51,39 @@ impl PipelineStage for RequestExecutionStage {
error::internal_error("Client acquisition not completed")
})?;
let result = match self.mode {
ExecutionMode::Single => self.execute_single(proto_request, clients).await?,
ExecutionMode::DualDispatch => {
self.execute_dual_dispatch(proto_request, clients).await?
// Extract dispatch metadata for tracing span
let request_id = ctx
.state
.dispatch
.as_ref()
.map(|d| d.request_id.as_str())
.unwrap_or("unknown");
let model = ctx
.state
.dispatch
.as_ref()
.map(|d| d.model.as_str())
.unwrap_or("unknown");
// Create OTEL span for gRPC request execution
let span = info_span!(
target: "sgl_model_gateway::otel-trace",
"grpc_generate",
request_id = %request_id,
model = %model,
mode = ?self.mode,
);
let result = async {
match self.mode {
ExecutionMode::Single => self.execute_single(proto_request, clients).await,
ExecutionMode::DualDispatch => {
self.execute_dual_dispatch(proto_request, clients).await
}
}
};
}
.instrument(span)
.await?;
// Store result in context for ResponseProcessingStage
ctx.state.response.execution_result = Some(result);