Revert "Merge PR #504"

This reverts commit d216a9e219, reversing
changes made to 21e82abd54.

apps/aether-gateway/src/ai_serving/planner/runtime_miss.rs

@@ -73,10 +73,6 @@ impl AiRuntimeMissDiagnosticPort for GatewayRuntimeMissDiagnosticPort<'_> {
             candidate_count: None,
             skipped_candidate_count: None,
             skip_reasons: std::collections::BTreeMap::new(),
-            provider_hint_id: None,
-            provider_hint_name: None,
-            endpoint_hint_id: None,
-            endpoint_hint_api_format: None,
         }
     }
 

apps/aether-gateway/src/dispatch/pool_scheduler.rs

@@ -428,20 +428,6 @@ impl<'a> PoolKeyCursor<'a> {
         if record_runtime_miss_diagnostic {
             self.runtime_miss_trace_id = Some(trace_id.to_string());
             self.record_runtime_miss_diagnostic = true;
-            let provider_id = self.group.candidate.provider_id.clone();
-            let provider_name = self.group.transport.provider.name.clone();
-            let endpoint_id = self.group.candidate.endpoint_id.clone();
-            let endpoint_api_format = self.group.provider_api_format.clone();
-            self.state
-                .app()
-                .mutate_local_execution_runtime_miss_diagnostic(trace_id, move |diagnostic| {
-                    diagnostic.provider_hint_id.get_or_insert(provider_id);
-                    diagnostic.provider_hint_name.get_or_insert(provider_name);
-                    diagnostic.endpoint_hint_id.get_or_insert(endpoint_id);
-                    diagnostic
-                        .endpoint_hint_api_format
-                        .get_or_insert(endpoint_api_format);
-                });
         }
         self
     }
@@ -2572,14 +2558,6 @@ mod tests {
         assert_eq!(diagnostic.reason, "all_candidates_skipped");
         assert_eq!(diagnostic.skipped_candidate_count, Some(1));
         assert_eq!(diagnostic.skip_reasons.get("pool_cooldown"), Some(&1));
-        assert_eq!(
-            diagnostic.provider_hint_id.as_deref(),
-            Some("provider-pool")
-        );
-        assert_eq!(
-            diagnostic.provider_hint_name.as_deref(),
-            Some("provider-pool")
-        );
     }
 
     #[tokio::test]

apps/aether-gateway/src/executor/outcome.rs

@@ -274,7 +274,6 @@ pub(crate) async fn record_failed_usage_for_exhausted_request(
         diagnostic,
         None,
         None,
-        0,
     );
     data.request_metadata = Some(Value::Object(request_metadata));
 
@@ -413,7 +412,6 @@ pub(crate) async fn record_failed_usage_for_runtime_miss_request(
         diagnostic,
         decision.and_then(|value| value.route_family.as_deref()),
         decision.and_then(|value| value.route_kind.as_deref()),
-        context.persisted_candidate_count(),
     );
     data.request_metadata =
         (!request_metadata.is_empty()).then_some(Value::Object(request_metadata));
@@ -985,7 +983,6 @@ fn apply_runtime_miss_usage_routing(
     diagnostic: Option<&LocalExecutionRuntimeMissDiagnostic>,
     route_family_fallback: Option<&str>,
     route_kind_fallback: Option<&str>,
-    persisted_candidate_count: usize,
 ) {
     data.candidate_id = data
         .candidate_id
@@ -1024,68 +1021,7 @@ fn apply_runtime_miss_usage_routing(
         .planner_kind
         .clone()
         .or_else(|| trimmed_non_empty(diagnostic.and_then(|value| value.plan_kind.as_deref())));
-    if let Some(diagnostic) = diagnostic {
+    let _ = request_metadata;
-        insert_runtime_miss_diagnostic_metadata(
-            request_metadata,
-            diagnostic,
-            persisted_candidate_count,
-        );
-    }
-}
-
-fn insert_runtime_miss_diagnostic_metadata(
-    request_metadata: &mut Map<String, Value>,
-    diagnostic: &LocalExecutionRuntimeMissDiagnostic,
-    persisted_candidate_count: usize,
-) {
-    let mut runtime_miss = Map::new();
-    if let Some(count) = diagnostic.candidate_count {
-        runtime_miss.insert("candidate_count".to_string(), json!(count));
-    }
-    runtime_miss.insert(
-        "persisted_candidate_count".to_string(),
-        json!(persisted_candidate_count),
-    );
-    if let Some(count) = diagnostic.skipped_candidate_count {
-        runtime_miss.insert("skipped_candidate_count".to_string(), json!(count));
-    }
-    if !diagnostic.skip_reasons.is_empty() {
-        runtime_miss.insert("skip_reasons".to_string(), json!(diagnostic.skip_reasons));
-    }
-    if let Some(requested_model) = trimmed_non_empty(diagnostic.requested_model.as_deref()) {
-        runtime_miss.insert(
-            "requested_model".to_string(),
-            Value::String(requested_model),
-        );
-    }
-
-    let mut provider_hint = Map::new();
-    if let Some(provider_id) = trimmed_provider_hint(diagnostic.provider_hint_id.as_deref()) {
-        provider_hint.insert("id".to_string(), Value::String(provider_id));
-    }
-    if let Some(provider_name) = trimmed_provider_hint(diagnostic.provider_hint_name.as_deref()) {
-        provider_hint.insert("name".to_string(), Value::String(provider_name));
-    }
-    if !provider_hint.is_empty() {
-        runtime_miss.insert("provider_hint".to_string(), Value::Object(provider_hint));
-    }
-
-    let mut endpoint_hint = Map::new();
-    if let Some(endpoint_id) = trimmed_non_empty(diagnostic.endpoint_hint_id.as_deref()) {
-        endpoint_hint.insert("id".to_string(), Value::String(endpoint_id));
-    }
-    if let Some(endpoint_api_format) =
-        trimmed_non_empty(diagnostic.endpoint_hint_api_format.as_deref())
-    {
-        endpoint_hint.insert("api_format".to_string(), Value::String(endpoint_api_format));
-    }
-    if !endpoint_hint.is_empty() {
-        runtime_miss.insert("endpoint_hint".to_string(), Value::Object(endpoint_hint));
-    }
-
-    if !runtime_miss.is_empty() {
-        request_metadata.insert("runtime_miss".to_string(), Value::Object(runtime_miss));
-    }
 }
 
 fn trimmed_non_empty(value: Option<&str>) -> Option<String> {
@@ -1095,15 +1031,6 @@ fn trimmed_non_empty(value: Option<&str>) -> Option<String> {
         .map(ToOwned::to_owned)
 }
 
-fn trimmed_provider_hint(value: Option<&str>) -> Option<String> {
-    trimmed_non_empty(value).filter(|value| {
-        !matches!(
-            value.to_ascii_lowercase().as_str(),
-            "unknown" | "unknow" | "pending"
-        )
-    })
-}
-
 #[cfg(test)]
 mod tests {
     use super::{
@@ -1118,7 +1045,6 @@ mod tests {
     };
     use aether_usage_runtime::UsageEventData;
     use serde_json::{json, Map, Value};
-    use std::collections::BTreeMap;
 
     #[test]
     fn local_execution_client_error_message_is_client_friendly() {
@@ -1143,7 +1069,7 @@ mod tests {
     }
 
     #[test]
-    fn runtime_miss_routing_records_compact_diagnostic_metadata() {
+    fn runtime_miss_routing_moves_to_typed_usage_fields_and_keeps_metadata_lightweight() {
         let mut data = UsageEventData::default();
         let mut request_metadata =
             Map::from_iter([("trace_id".to_string(), Value::String("trace-1".to_string()))]);
@@ -1160,19 +1086,10 @@ mod tests {
                 route_family: Some("claude".to_string()),
                 route_kind: Some("cli".to_string()),
                 plan_kind: Some("claude_cli_sync".to_string()),
-                requested_model: Some("gpt-5".to_string()),
-                candidate_count: Some(1),
-                skipped_candidate_count: Some(4),
-                skip_reasons: BTreeMap::from([("pool_cooldown".to_string(), 4)]),
-                provider_hint_id: Some("provider-google-api".to_string()),
-                provider_hint_name: Some("Google API".to_string()),
-                endpoint_hint_id: Some("endpoint-gemini".to_string()),
-                endpoint_hint_api_format: Some("gemini:generate_content".to_string()),
                 ..LocalExecutionRuntimeMissDiagnostic::default()
             }),
             None,
             None,
-            0,
         );
 
         assert_eq!(data.candidate_id.as_deref(), Some("cand-1"));
@@ -1192,24 +1109,7 @@ mod tests {
         assert_eq!(
             Value::Object(request_metadata),
             json!({
-                "trace_id": "trace-1",
+                "trace_id": "trace-1"
-                "runtime_miss": {
-                    "candidate_count": 1,
-                    "persisted_candidate_count": 0,
-                    "skipped_candidate_count": 4,
-                    "skip_reasons": {
-                        "pool_cooldown": 4
-                    },
-                    "provider_hint": {
-                        "id": "provider-google-api",
-                        "name": "Google API"
-                    },
-                    "endpoint_hint": {
-                        "id": "endpoint-gemini",
-                        "api_format": "gemini:generate_content"
-                    },
-                    "requested_model": "gpt-5"
-                }
             })
         );
     }

apps/aether-gateway/src/state/types.rs

@@ -30,10 +30,6 @@ pub(crate) struct LocalExecutionRuntimeMissDiagnostic {
     pub(crate) candidate_count: Option<usize>,
     pub(crate) skipped_candidate_count: Option<usize>,
     pub(crate) skip_reasons: std::collections::BTreeMap<String, usize>,
-    pub(crate) provider_hint_id: Option<String>,
-    pub(crate) provider_hint_name: Option<String>,
-    pub(crate) endpoint_hint_id: Option<String>,
-    pub(crate) endpoint_hint_api_format: Option<String>,
 }
 
 impl LocalExecutionRuntimeMissDiagnostic {

crates/aether-admin/src/observability/usage.rs

@@ -492,60 +492,6 @@ fn admin_usage_extract_local_runtime_miss_reason_summary(message: &str) -> Optio
     (!summary.is_empty()).then(|| summary.to_string())
 }
 
-fn admin_usage_runtime_miss_metadata(
-    item: &StoredRequestUsageAudit,
-) -> Option<&serde_json::Map<String, Value>> {
-    item.request_metadata
-        .as_ref()
-        .and_then(Value::as_object)
-        .and_then(|metadata| metadata.get("runtime_miss"))
-        .and_then(Value::as_object)
-}
-
-fn admin_usage_json_value_has_content(value: &Value) -> bool {
-    match value {
-        Value::Null => false,
-        Value::String(value) => !value.trim().is_empty(),
-        Value::Array(value) => !value.is_empty(),
-        Value::Object(value) => !value.is_empty(),
-        Value::Bool(_) | Value::Number(_) => true,
-    }
-}
-
-fn maybe_insert_runtime_miss_field(
-    object: &mut serde_json::Map<String, Value>,
-    runtime_miss: &serde_json::Map<String, Value>,
-    key: &str,
-) {
-    let Some(value) = runtime_miss
-        .get(key)
-        .filter(|value| admin_usage_json_value_has_content(value))
-    else {
-        return;
-    };
-    object.insert(key.to_string(), value.clone());
-}
-
-fn admin_usage_insert_runtime_miss_fields(
-    object: &mut serde_json::Map<String, Value>,
-    item: &StoredRequestUsageAudit,
-) {
-    let Some(runtime_miss) = admin_usage_runtime_miss_metadata(item) else {
-        return;
-    };
-    for key in [
-        "candidate_count",
-        "persisted_candidate_count",
-        "skipped_candidate_count",
-        "skip_reasons",
-        "requested_model",
-        "provider_hint",
-        "endpoint_hint",
-    ] {
-        maybe_insert_runtime_miss_field(object, runtime_miss, key);
-    }
-}
-
 fn admin_usage_scheduling_failure_json(
     item: &StoredRequestUsageAudit,
     client_error: &Value,
@@ -568,38 +514,20 @@ fn admin_usage_scheduling_failure_json(
     let reason_summary =
         raw_message.and_then(admin_usage_extract_local_runtime_miss_reason_summary);
 
-    let mut object = serde_json::Map::new();
+    json!({
-    object.insert(
+        "source": "local_execution_runtime_miss",
-        "source".to_string(),
+        "reason": reason,
-        Value::String("local_execution_runtime_miss".to_string()),
+        "reason_label": admin_usage_local_runtime_miss_reason_label(reason),
-    );
+        "title": format!("本地调度失败：{}", admin_usage_local_runtime_miss_reason_label(reason)),
-    object.insert("reason".to_string(), Value::String(reason.to_string()));
+        "message": message,
-    object.insert(
+        "reason_summary": reason_summary,
-        "reason_label".to_string(),
+        "status_code": item.status_code,
-        Value::String(admin_usage_local_runtime_miss_reason_label(reason).to_string()),
+        "no_upstream_attempt": item.candidate_id.is_none()
-    );
+            && item.provider_api_key_id.is_none()
-    object.insert(
+            && item.provider_request_headers.is_none()
-        "title".to_string(),
+            && item.provider_request_body.is_none()
-        Value::String(format!(
+            && item.provider_request_body_ref.is_none(),
-            "本地调度失败：{}",
+    })
-            admin_usage_local_runtime_miss_reason_label(reason)
-        )),
-    );
-    object.insert("message".to_string(), json!(message));
-    object.insert("reason_summary".to_string(), json!(reason_summary));
-    object.insert("status_code".to_string(), json!(item.status_code));
-    object.insert(
-        "no_upstream_attempt".to_string(),
-        json!(
-            item.candidate_id.is_none()
-                && item.provider_api_key_id.is_none()
-                && item.provider_request_headers.is_none()
-                && item.provider_request_body.is_none()
-                && item.provider_request_body_ref.is_none()
-        ),
-    );
-    admin_usage_insert_runtime_miss_fields(&mut object, item);
-    Value::Object(object)
 }
 
 fn admin_usage_extract_local_execution_request_mode(message: &str) -> Option<&str> {
@@ -1234,7 +1162,6 @@ fn admin_usage_active_request_json(
         "request_path": admin_usage_metadata_string(item, "request_path"),
         "request_path_and_query": admin_usage_metadata_string(item, "request_path_and_query"),
         "has_fallback": admin_usage_has_fallback(item),
-        "scheduling_failure": admin_usage_scheduling_failure_json(item, &Value::Null),
     });
     if let Some(api_format) = item.api_format.as_ref() {
         value["api_format"] = json!(api_format);
@@ -1324,7 +1251,6 @@ pub fn admin_usage_record_json(
         "api_key_name": api_key_name,
         "provider_key_name": provider_key_name,
         "model_version": Value::Null,
-        "scheduling_failure": admin_usage_scheduling_failure_json(item, &Value::Null),
     });
     let object = payload
         .as_object_mut()
@@ -3295,122 +3221,6 @@ mod tests {
         assert_eq!(payload["scheduling_failure"]["no_upstream_attempt"], true);
     }
 
-    #[test]
-    fn detail_payload_exposes_runtime_miss_context_inside_scheduling_failure() {
-        let message = "找到 1 个支持模型 gemma-4-31b-it 的候选提供商，但本次同步请求全部不可用：provider_quota_blocked 1 次（原因代码: all_candidates_skipped）";
-        let item = StoredRequestUsageAudit {
-            provider_name: "unknown".to_string(),
-            provider_id: None,
-            provider_endpoint_id: None,
-            provider_api_key_id: None,
-            provider_request_headers: None,
-            provider_request_body: None,
-            provider_request_body_ref: None,
-            candidate_id: None,
-            execution_path: Some("local_execution_runtime_miss".to_string()),
-            local_execution_runtime_miss_reason: Some("all_candidates_skipped".to_string()),
-            request_metadata: Some(json!({
-                "runtime_miss": {
-                    "candidate_count": 1,
-                    "persisted_candidate_count": 0,
-                    "skipped_candidate_count": 1,
-                    "skip_reasons": {
-                        "provider_quota_blocked": 1
-                    },
-                    "requested_model": "gemma-4-31b-it",
-                    "provider_hint": {
-                        "id": "provider-google-api",
-                        "name": "Google API"
-                    },
-                    "endpoint_hint": {
-                        "id": "endpoint-gemini",
-                        "api_format": "gemini:generate_content"
-                    }
-                }
-            })),
-            ..sample_usage("failed", Some(503), Some(message))
-        };
-
-        let payload = build_admin_usage_detail_payload(
-            &item,
-            &BTreeMap::new(),
-            &BTreeMap::new(),
-            false,
-            false,
-            None,
-            false,
-            Some(json!({"model": "gemma-4-31b-it"})),
-            &BTreeMap::new(),
-        );
-
-        assert_eq!(
-            payload["scheduling_failure"]["provider_hint"]["name"],
-            "Google API"
-        );
-        assert_eq!(
-            payload["scheduling_failure"]["endpoint_hint"]["api_format"],
-            "gemini:generate_content"
-        );
-        assert_eq!(
-            payload["scheduling_failure"]["requested_model"],
-            "gemma-4-31b-it"
-        );
-        assert_eq!(payload["scheduling_failure"]["candidate_count"], 1);
-        assert_eq!(
-            payload["scheduling_failure"]["persisted_candidate_count"],
-            0
-        );
-        assert_eq!(payload["scheduling_failure"]["skipped_candidate_count"], 1);
-        assert_eq!(
-            payload["scheduling_failure"]["skip_reasons"]["provider_quota_blocked"],
-            1
-        );
-    }
-
-    #[test]
-    fn record_json_exposes_scheduling_failure_for_local_runtime_miss_records() {
-        let message = "没有可用提供商支持模型 gemma-4-31b-it 的同步请求。请检查模型映射、端点启用状态和 API Key 权限（原因代码: candidate_list_empty）";
-        let item = StoredRequestUsageAudit {
-            provider_name: "unknown".to_string(),
-            provider_id: None,
-            provider_endpoint_id: None,
-            provider_api_key_id: None,
-            provider_request_headers: None,
-            provider_request_body: None,
-            provider_request_body_ref: None,
-            candidate_id: None,
-            execution_path: Some("local_execution_runtime_miss".to_string()),
-            local_execution_runtime_miss_reason: Some("candidate_list_empty".to_string()),
-            request_metadata: Some(json!({
-                "runtime_miss": {
-                    "candidate_count": 0,
-                    "persisted_candidate_count": 0,
-                    "requested_model": "gemma-4-31b-it"
-                }
-            })),
-            ..sample_usage("failed", Some(503), Some(message))
-        };
-
-        let payload = admin_usage_record_json(
-            &item,
-            &BTreeMap::new(),
-            &BTreeMap::new(),
-            false,
-            false,
-            None,
-        );
-
-        assert_eq!(
-            payload["scheduling_failure"]["title"],
-            "本地调度失败：没有可调度候选"
-        );
-        assert_eq!(
-            payload["scheduling_failure"]["requested_model"],
-            "gemma-4-31b-it"
-        );
-        assert_eq!(payload["scheduling_failure"]["no_upstream_attempt"], true);
-    }
-
     #[test]
     fn detail_payload_preserves_legacy_body_capture_metadata_keys() {
         let item = StoredRequestUsageAudit {

frontend/src/api/dashboard.ts

@@ -151,19 +151,6 @@ export interface RequestSchedulingFailure {
   reason_summary?: string | null
   status_code?: number | null
   no_upstream_attempt?: boolean | null
-  requested_model?: string | null
-  candidate_count?: number | null
-  persisted_candidate_count?: number | null
-  skipped_candidate_count?: number | null
-  skip_reasons?: Record<string, number> | null
-  provider_hint?: {
-    id?: string | null
-    name?: string | null
-  } | null
-  endpoint_hint?: {
-    id?: string | null
-    api_format?: string | null
-  } | null
 }
 
 export interface RequestDetail {

frontend/src/features/usage/components/HorizontalRequestTimeline.vue

@@ -547,38 +547,6 @@
       </Card>
     </div>
 
-    <!-- Local Scheduling Failure State -->
-    <Card
-      v-else-if="schedulingFailureNotice"
-      class="border-red-200 dark:border-red-800"
-    >
-      <div class="p-4 space-y-2">
-        <div class="flex flex-wrap items-center gap-2">
-          <Badge variant="destructive">
-            调度失败
-          </Badge>
-          <h4 class="text-sm font-semibold text-red-950 dark:text-red-100">
-            {{ schedulingFailureNotice.title }}
-          </h4>
-        </div>
-        <p class="text-sm leading-6 text-red-900 dark:text-red-100">
-          {{ schedulingFailureNotice.message }}
-        </p>
-        <div
-          v-if="schedulingFailureNotice.meta.length > 0"
-          class="flex flex-wrap gap-1.5"
-        >
-          <span
-            v-for="item in schedulingFailureNotice.meta"
-            :key="item"
-            class="rounded-full border border-red-200 bg-white/70 px-2 py-0.5 text-[11px] font-mono text-red-700 dark:border-red-900 dark:bg-red-950/50 dark:text-red-200"
-          >
-            {{ item }}
-          </span>
-        </div>
-      </div>
-    </Card>
-
     <!-- Empty State -->
     <Card
       v-else
@@ -607,7 +575,6 @@ import { parseApiError } from '@/utils/errorParser'
 import { formatApiFormat } from '@/api/endpoints/types/api-format'
 import { useDarkMode } from '@/composables/useDarkMode'
 import { resolveTimelineFinalStatus } from '../utils/status'
-import type { RequestSchedulingFailure } from '@/api/dashboard'
 import {
   buildPoolGroupVisibleAttempts,
   buildPoolParticipatedCandidates,
@@ -670,8 +637,6 @@ const props = defineProps<{
   usageData?: UsageData | null
   /** 请求元数据（用于号池调度组装） */
   requestMetadata?: Record<string, unknown> | null
-  /** 本地调度失败摘要；用于没有 trace candidate 时替代空态 */
-  schedulingFailure?: RequestSchedulingFailure | null
   /** 已获取的追踪数据；传入时不再内部拉取 */
   traceData?: RequestTrace | null
 }>()
@@ -847,62 +812,6 @@ const computedFinalStatus = computed(() => {
   })
 })
 
-const nonEmptyNoticeString = (value: string | null | undefined): string | null => {
-  const trimmed = value?.trim()
-  return trimmed ? trimmed : null
-}
-
-const uniqueNoticeMeta = (values: Array<string | null | undefined>): string[] => {
-  return Array.from(new Set(values.map(value => value?.trim()).filter((value): value is string => Boolean(value))))
-}
-
-const isUnknownProviderHint = (value: string | null): boolean => {
-  const normalized = value?.trim().toLowerCase()
-  return !normalized || ['unknown', 'unknow', 'pending'].includes(normalized)
-}
-
-const schedulingFailureProviderHint = (failure: RequestSchedulingFailure): string | null => {
-  const name = nonEmptyNoticeString(failure.provider_hint?.name)
-  if (name && !isUnknownProviderHint(name)) return name
-
-  const id = nonEmptyNoticeString(failure.provider_hint?.id)
-  if (id && !isUnknownProviderHint(id)) return id
-
-  return null
-}
-
-const schedulingFailureEndpointHint = (failure: RequestSchedulingFailure): string | null => {
-  return nonEmptyNoticeString(failure.endpoint_hint?.api_format)
-    ?? nonEmptyNoticeString(failure.endpoint_hint?.id)
-}
-
-const schedulingFailureNotice = computed(() => {
-  const failure = props.schedulingFailure
-  if (!failure) return null
-
-  const title = nonEmptyNoticeString(failure.title) ?? '本地调度失败'
-  const message = nonEmptyNoticeString(failure.message)
-    ?? nonEmptyNoticeString(failure.reason_summary)
-    ?? nonEmptyNoticeString(failure.reason_label)
-    ?? nonEmptyNoticeString(failure.reason)
-    ?? '本地调度阶段没有选出可用上游提供商'
-
-  return {
-    title,
-    message,
-    meta: uniqueNoticeMeta([
-      schedulingFailureProviderHint(failure),
-      schedulingFailureEndpointHint(failure),
-      nonEmptyNoticeString(failure.requested_model),
-      nonEmptyNoticeString(failure.reason_summary),
-      nonEmptyNoticeString(failure.reason_label),
-      nonEmptyNoticeString(failure.reason),
-      typeof failure.status_code === 'number' ? `HTTP ${failure.status_code}` : null,
-      failure.no_upstream_attempt ? '未进入上游执行' : null,
-    ]),
-  }
-})
-
 const compareBySchedulingOrder = (a: CandidateRecord, b: CandidateRecord): number => {
   if (a.candidate_index !== b.candidate_index) {
     return a.candidate_index - b.candidate_index

frontend/src/features/usage/components/RequestDetailDrawer.vue

@@ -544,7 +544,6 @@
                   :request-status="detail.status"
                   :request-api-format="detail.api_format || null"
                   :request-metadata="traceRequestMetadata"
-                  :scheduling-failure="detail.scheduling_failure"
                   @trace-state="handleTraceState"
                 />
               </div>

frontend/src/features/usage/components/UsageRecordsTable.vue

@@ -584,18 +584,15 @@
           >
             <div class="flex min-w-0 items-center gap-1">
               <div class="flex min-w-0 flex-col text-xs gap-0.5">
+                <span class="truncate">{{ record.provider }}</span>
                 <span
-                  class="truncate"
+                  v-if="record.provider_key_name"
-                  :title="getRecordProviderTitle(record)"
-                >{{ getRecordProviderDisplay(record) }}</span>
-                <span
-                  v-if="getRecordProviderSecondaryText(record)"
                   class="text-muted-foreground truncate"
-                  :title="getRecordProviderSecondaryTitle(record)"
+                  :title="record.provider_key_name"
                 >
-                  {{ getRecordProviderSecondaryText(record) }}
+                  {{ record.provider_key_name }}
                   <span
-                    v-if="record.provider_key_name && record.rate_multiplier && record.rate_multiplier !== 1.0"
+                    v-if="record.rate_multiplier && record.rate_multiplier !== 1.0"
                     class="text-foreground/60"
                   >({{ record.rate_multiplier }}x)</span>
                 </span>
@@ -1177,57 +1174,6 @@ function getDisplayStatus(record: UsageRecord) {
   return resolveDisplayRequestStatus(record)
 }
 
-function nonEmptyDisplayString(value: string | null | undefined): string | null {
-  const trimmed = value?.trim()
-  return trimmed ? trimmed : null
-}
-
-function isUnknownProvider(value: string | null | undefined): boolean {
-  const normalized = value?.trim().toLowerCase()
-  return !normalized || ['unknown', 'unknow', 'pending'].includes(normalized)
-}
-
-function getRecordSchedulingProviderHint(record: UsageRecord): string | null {
-  const name = nonEmptyDisplayString(record.scheduling_failure?.provider_hint?.name)
-  if (name && !isUnknownProvider(name)) return name
-
-  const id = nonEmptyDisplayString(record.scheduling_failure?.provider_hint?.id)
-  if (id && !isUnknownProvider(id)) return id
-
-  return null
-}
-
-function getRecordProviderDisplay(record: UsageRecord): string {
-  const provider = nonEmptyDisplayString(record.provider)
-  if (provider && !isUnknownProvider(provider)) return provider
-
-  const providerHint = getRecordSchedulingProviderHint(record)
-  if (providerHint) return providerHint
-
-  if (record.scheduling_failure) return '未选定提供商'
-  return provider ?? 'unknown'
-}
-
-function getRecordProviderTitle(record: UsageRecord): string {
-  const schedulingMessage = nonEmptyDisplayString(record.scheduling_failure?.message)
-  const providerDisplay = getRecordProviderDisplay(record)
-  return schedulingMessage ? `${providerDisplay}\n${schedulingMessage}` : providerDisplay
-}
-
-function getRecordProviderSecondaryText(record: UsageRecord): string | null {
-  return nonEmptyDisplayString(record.provider_key_name)
-    ?? nonEmptyDisplayString(record.scheduling_failure?.reason_label)
-    ?? nonEmptyDisplayString(record.scheduling_failure?.reason_summary)
-    ?? nonEmptyDisplayString(record.scheduling_failure?.reason)
-}
-
-function getRecordProviderSecondaryTitle(record: UsageRecord): string | undefined {
-  return nonEmptyDisplayString(record.provider_key_name)
-    ?? nonEmptyDisplayString(record.scheduling_failure?.message)
-    ?? getRecordProviderSecondaryText(record)
-    ?? undefined
-}
-
 function getStreamModeLabel(record: UsageRecord): string {
   return formatUsageStreamLabel(record)
 }

frontend/src/features/usage/components/__tests__/HorizontalRequestTimeline.spec.ts

@@ -350,40 +350,6 @@ describe('HorizontalRequestTimeline', () => {
     expect(nodeDot?.classList.contains('status-success')).toBe(false)
   })
 
-  it('shows scheduling failure context instead of an empty trace state when no candidates exist', async () => {
-    const trace = buildTrace([])
-    trace.final_status = 'failed'
-
-    const root = mountTimeline(trace, {
-      requestStatus: 'failed',
-      overrideStatusCode: 503,
-      schedulingFailure: {
-        source: 'local_execution_runtime_miss',
-        reason: 'all_candidates_skipped',
-        reason_label: '所有候选均被跳过',
-        title: '本地调度失败：所有候选均被跳过',
-        message: '没有可用提供商支持模型 gemma-4-31b-it 的同步请求',
-        status_code: 503,
-        no_upstream_attempt: true,
-        provider_hint: {
-          id: 'provider-google-api',
-          name: 'Google API',
-        },
-        endpoint_hint: {
-          id: 'endpoint-gemini',
-          api_format: 'gemini:generate_content',
-        },
-      },
-    })
-    await nextTick()
-
-    expect(root.textContent).toContain('本地调度失败：所有候选均被跳过')
-    expect(root.textContent).toContain('没有可用提供商支持模型 gemma-4-31b-it 的同步请求')
-    expect(root.textContent).toContain('Google API')
-    expect(root.textContent).toContain('gemini:generate_content')
-    expect(root.textContent).not.toContain('暂无追踪数据')
-  })
-
   it('keeps emitted trace state active while the request lifecycle is still streaming', async () => {
     const onTraceState = vi.fn()
     const trace = buildTrace([

frontend/src/features/usage/components/__tests__/UsageRecordsTable.spec.ts

@@ -255,31 +255,6 @@ describe('UsageRecordsTable', () => {
     expect(root.textContent).not.toContain('等待中')
   })
 
-  it('uses scheduling failure provider hints instead of rendering unknown provider', () => {
-    const root = mountUsageRecordsTable([buildRecord({
-      provider: 'unknown',
-      status: 'failed',
-      status_code: 503,
-      scheduling_failure: {
-        source: 'local_execution_runtime_miss',
-        reason: 'all_candidates_skipped',
-        reason_label: '所有候选均被跳过',
-        title: '本地调度失败：所有候选均被跳过',
-        message: '没有可用提供商支持模型 gemma-4-31b-it 的同步请求',
-        status_code: 503,
-        no_upstream_attempt: true,
-        provider_hint: {
-          id: 'provider-google-api',
-          name: 'Google API',
-        },
-      },
-    } as Partial<UsageRecord>)])
-
-    expect(root.textContent).toContain('Google API')
-    expect(root.textContent).toContain('所有候选均被跳过')
-    expect(root.textContent).not.toContain('unknown')
-  })
-
   it('renders output TPS in the non-admin usage table', () => {
     const root = mountUsageRecordsTable([buildRecord()], { isAdmin: false })
 

frontend/src/features/usage/composables/useUsageData.ts

@@ -515,8 +515,7 @@ export function useUsageData(options: UseUsageDataOptions) {
           api_key_name: existing.api_key_name || record.api_key_name,
           provider_key_name: existing.provider_key_name || record.provider_key_name,
           rate_multiplier: existing.rate_multiplier ?? record.rate_multiplier,
-          target_model: existing.target_model || record.target_model,
+          target_model: existing.target_model || record.target_model
-          scheduling_failure: existing.scheduling_failure ?? record.scheduling_failure
         }
       }
 
@@ -524,8 +523,7 @@ export function useUsageData(options: UseUsageDataOptions) {
       if (protectProvider) {
         return {
           ...record,
-          provider: existing.provider,
+          provider: existing.provider
-          scheduling_failure: existing.scheduling_failure ?? record.scheduling_failure
         }
       }
 

frontend/src/features/usage/types.ts

@@ -1,5 +1,4 @@
 import type { ImageProgress } from '@/api/requestTrace'
-import type { RequestSchedulingFailure } from '@/api/dashboard'
 
 // 统计数据状态
 export interface UsageStatsState {
@@ -121,7 +120,6 @@ export interface UsageRecord {
   status_code?: number
   error_message?: string
   status?: RequestStatus  // 请求状态: pending, streaming, completed, failed
-  scheduling_failure?: RequestSchedulingFailure | null
   created_at: string
   has_fallback?: boolean
   has_retry?: boolean

frontend/src/features/usage/utils/__tests__/errorNotice.spec.ts

@@ -73,43 +73,6 @@ describe('request failure notice', () => {
     })
   })
 
-  it('includes scheduling failure provider and endpoint hints in metadata', () => {
-    const notice = resolveRequestFailureNotice(buildRequestDetail({
-      failure_summary: {
-        status_code: 503,
-        message: '没有可用提供商支持模型 gemma-4-31b-it 的同步请求',
-      },
-      scheduling_failure: {
-        source: 'local_execution_runtime_miss',
-        reason: 'all_candidates_skipped',
-        reason_label: '所有候选均被跳过',
-        title: '本地调度失败：所有候选均被跳过',
-        message: '没有可用提供商支持模型 gemma-4-31b-it 的同步请求',
-        status_code: 503,
-        no_upstream_attempt: true,
-        requested_model: 'gemma-4-31b-it',
-        provider_hint: {
-          id: 'provider-google-api',
-          name: 'Google API',
-        },
-        endpoint_hint: {
-          id: 'endpoint-gemini',
-          api_format: 'gemini:generate_content',
-        },
-      } as NonNullable<RequestDetail['scheduling_failure']>,
-    }))
-
-    expect(notice?.meta).toEqual([
-      'Google API',
-      'gemini:generate_content',
-      'gemma-4-31b-it',
-      '所有候选均被跳过',
-      'all_candidates_skipped',
-      'HTTP 503',
-      '未进入上游执行',
-    ])
-  })
-
   it('falls back to the failure summary for upstream failures', () => {
     const notice = resolveRequestFailureNotice(buildRequestDetail({
       failure_summary: {

frontend/src/features/usage/utils/errorNotice.ts

@@ -25,11 +25,6 @@ function uniqueMeta(values: Array<string | null | undefined>): string[] {
   return Array.from(new Set(values.map(value => value?.trim()).filter((value): value is string => Boolean(value))))
 }
 
-function isUnknownProviderHint(value: string | null): boolean {
-  const normalized = value?.trim().toLowerCase()
-  return !normalized || ['unknown', 'unknow', 'pending'].includes(normalized)
-}
-
 function schedulingFailureMessage(
   failure: RequestSchedulingFailure,
   fallbackDomain: RequestErrorDomain | null,
@@ -42,21 +37,6 @@ function schedulingFailureMessage(
     ?? nonEmptyString(failure.reason)
 }
 
-function schedulingFailureProviderHint(failure: RequestSchedulingFailure): string | null {
-  const name = nonEmptyString(failure.provider_hint?.name)
-  if (name && !isUnknownProviderHint(name)) return name
-
-  const id = nonEmptyString(failure.provider_hint?.id)
-  if (id && !isUnknownProviderHint(id)) return id
-
-  return null
-}
-
-function schedulingFailureEndpointHint(failure: RequestSchedulingFailure): string | null {
-  return nonEmptyString(failure.endpoint_hint?.api_format)
-    ?? nonEmptyString(failure.endpoint_hint?.id)
-}
-
 export function resolveRequestFailureNotice(detail: RequestDetail | null | undefined): RequestFailureNotice | null {
   if (!detail) return null
 
@@ -75,9 +55,6 @@ export function resolveRequestFailureNotice(detail: RequestDetail | null | undef
         message,
         isSchedulingFailure: true,
         meta: uniqueMeta([
-          schedulingFailureProviderHint(schedulingFailure),
-          schedulingFailureEndpointHint(schedulingFailure),
-          nonEmptyString(schedulingFailure.requested_model),
           nonEmptyString(schedulingFailure.reason_summary),
           nonEmptyString(schedulingFailure.reason_label),
           nonEmptyString(schedulingFailure.reason),