fix(account): prevent quota-exceeded API key/Bedrock accounts from being scheduled

Add quota exceeded check to IsSchedulable() and refactor
shouldClearStickySession to delegate to IsSchedulable(), eliminating
duplicated logic and fixing missed overload/rate-limit/expired checks.
Frontend displays quota exceeded status independently via quota fields.

backend/internal/service/gateway_service.go

@@ -435,26 +435,19 @@ func prefetchedStickyAccountIDFromContext(ctx context.Context, groupID *int64) i
 }
 
 // shouldClearStickySession 检查账号是否处于不可调度状态，需要清理粘性会话绑定。
-// 当账号状态为错误、禁用、不可调度、处于临时不可调度期间，
-// 或请求的模型处于限流状态时，返回 true。
-// 这确保后续请求不会继续使用不可用的账号。
+// 委托 IsSchedulable() 判断账号级可调度性（状态、配额、过载、限流等），
+// 额外检查模型级限流。
 //
 // shouldClearStickySession checks if an account is in an unschedulable state
 // and the sticky session binding should be cleared.
-// Returns true when account status is error/disabled, schedulable is false,
-// within temporary unschedulable period, or the requested model is rate-limited.
-// This ensures subsequent requests won't continue using unavailable accounts.
+// Delegates to IsSchedulable() for account-level checks, plus model-level rate limiting.
 func shouldClearStickySession(account *Account, requestedModel string) bool {
 	if account == nil {
 		return false
 	}
-	if account.Status == StatusError || account.Status == StatusDisabled || !account.Schedulable {
+	if !account.IsSchedulable() {
 		return true
 	}
-	if account.TempUnschedulableUntil != nil && time.Now().Before(*account.TempUnschedulableUntil) {
-		return true
-	}
-	// 检查模型限流和 scope 限流，有限流即清除粘性会话
 	if remaining := account.GetRateLimitRemainingTimeWithContext(context.Background(), requestedModel); remaining > 0 {
 		return true
 	}