feat(health): add model and provider health monitoring

- Rename the original health monitor to endpoint health monitor
- Add tab navigation for endpoint, model, and provider health views
- Add model health monitor cards with availability, latency, first-byte latency, and 60-point history
- Add admin-only provider health monitor with collapsible active-provider sections
- Show per-provider model health cards after expanding a provider
- Add backend model health and provider health monitor payload builders
- Add admin endpoint for provider health monitoring
- Add provider-scoped usage breakdown filtering for per-provider model statistics
- Add frontend API types and request helpers for model/provider health data
- Add demo mock data for model and provider health monitoring
- Fix model health timeline time-unit handling so request history segments render correctly

Verification:
- cargo fmt
- npm run type-check
- npm run build
- cargo test -p aether-gateway health_models
- cargo test -p aether-gateway health_providers
- cargo test -p aether-gateway gateway_exposes_frontdoor_manifest_without_proxying_upstream
This commit is contained in:
AAEE86
2026-05-28 12:00:20 +08:00
parent 14ad6e9b75
commit 4ce056fe45
24 changed files with 1914 additions and 17 deletions
@@ -1,7 +1,7 @@
use super::extractors::{admin_health_key_id, admin_recover_key_id};
use crate::handlers::admin::request::{AdminAppState, AdminRequestContext};
use crate::handlers::admin::shared::query_param_value;
use crate::handlers::public::ApiFormatHealthMonitorOptions;
use crate::handlers::public::{ApiFormatHealthMonitorOptions, ModelHealthMonitorOptions};
use crate::GatewayError;
use axum::{
body::Body,
@@ -157,6 +157,80 @@ pub(super) async fn maybe_build_local_admin_endpoints_health_response(
return Ok(Some(Json(payload).into_response()));
}
if decision.route_family.as_deref() == Some("endpoints_health")
&& decision.route_kind.as_deref() == Some("health_models")
&& request_context.path() == "/api/admin/endpoints/health/models"
{
if !state.has_usage_data_reader() {
return Ok(Some(build_admin_endpoint_health_data_unavailable_response()));
}
let lookback_hours = query_param_value(request_context.query_string(), "lookback_hours")
.and_then(|value| value.parse::<u64>().ok())
.filter(|value| (1..=72).contains(value))
.unwrap_or(6);
let model_limit = query_param_value(request_context.query_string(), "model_limit")
.and_then(|value| value.parse::<usize>().ok())
.filter(|value| (1..=50).contains(value))
.unwrap_or(12);
let per_model_limit = query_param_value(request_context.query_string(), "per_model_limit")
.and_then(|value| value.parse::<usize>().ok())
.filter(|value| (10..=200).contains(value))
.unwrap_or(60);
let Some(payload) = state
.build_model_health_monitor_payload(
lookback_hours,
model_limit,
per_model_limit,
ModelHealthMonitorOptions {
include_provider_count: true,
},
)
.await
else {
return Ok(Some(build_admin_endpoint_health_data_unavailable_response()));
};
return Ok(Some(Json(payload).into_response()));
}
if decision.route_family.as_deref() == Some("endpoints_health")
&& decision.route_kind.as_deref() == Some("health_providers")
&& request_context.path() == "/api/admin/endpoints/health/providers"
{
if !state.has_provider_catalog_data_reader() || !state.has_usage_data_reader() {
return Ok(Some(build_admin_endpoint_health_data_unavailable_response()));
}
let lookback_hours = query_param_value(request_context.query_string(), "lookback_hours")
.and_then(|value| value.parse::<u64>().ok())
.filter(|value| (1..=72).contains(value))
.unwrap_or(6);
let provider_limit = query_param_value(request_context.query_string(), "provider_limit")
.and_then(|value| value.parse::<usize>().ok())
.filter(|value| (1..=100).contains(value))
.unwrap_or(50);
let per_provider_model_limit =
query_param_value(request_context.query_string(), "per_provider_model_limit")
.and_then(|value| value.parse::<usize>().ok())
.filter(|value| (1..=50).contains(value))
.unwrap_or(12);
let per_model_event_limit =
query_param_value(request_context.query_string(), "per_model_limit")
.and_then(|value| value.parse::<usize>().ok())
.filter(|value| (10..=200).contains(value))
.unwrap_or(100);
let Some(payload) = state
.build_provider_health_monitor_payload(
lookback_hours,
provider_limit,
per_provider_model_limit,
per_model_event_limit,
)
.await
else {
return Ok(Some(build_admin_endpoint_health_data_unavailable_response()));
};
return Ok(Some(Json(payload).into_response()));
}
if decision.route_family.as_deref() == Some("endpoints_health")
&& decision.route_kind.as_deref() == Some("health_status")
&& request_context.path() == "/api/admin/endpoints/health/status"