i/Aether
1
0
Fork 0
mirror of https://github.com/fawney19/Aether.git synced 2026-10-11 11:49:50 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix(gateway): add provider key concurrency and cache affinity modes

Browse Source
This commit is contained in:
ZheFox
2026-09-01 08:06:58 +08:00
parent 36daba7a34
commit 9631b229b3
26 changed files with 1176 additions and 154 deletions
Download Patch File Download Diff File Expand all files Collapse all files
apps/aether-gateway/src/handlers/proxy/websocket/responses/admission.rs
+12 -9
View File
@@ -11,7 +11,7 @@ use aether_contracts::ExecutionPlan;
use crate::execution_runtime::acquire_upstream_execution_gate;
use crate::provider_pool_demand::{
acquire_provider_pool_in_flight_guard, ProviderPoolInFlightGuard,
acquire_provider_pool_execution_guard, ProviderPoolInFlightAdmission, ProviderPoolInFlightGuard,
};
use crate::upstream_admission::UpstreamTargetAdmissionPermit;
use crate::{AppState, GatewayError};
@@ -41,14 +41,17 @@ impl ResponsesWebSocketTurnAdmission {
return Err(error);
}
};
let provider_pool = acquire_provider_pool_in_flight_guard(
state.runtime_state.clone(),
&plan.provider_id,
&plan.request_id,
plan.candidate_id.as_deref(),
&plan.key_id,
)
.await;
let provider_pool = match acquire_provider_pool_execution_guard(state, plan).await? {
ProviderPoolInFlightAdmission::Acquired(guard) => guard,
ProviderPoolInFlightAdmission::Saturated { limit } => {
drop(upstream_target);
drop(upstream_execution);
return Err(GatewayError::Client {
status: http::StatusCode::TOO_MANY_REQUESTS,
message: format!("上游账号并发已达上限 ({limit})"),
});
}
};
Ok(Self {
upstream_execution,