From 37c044fb4ba240b97fa33f0f6405d45fdcbad81f Mon Sep 17 00:00:00 2001 From: SsageParuders Date: Sat, 14 Mar 2026 22:47:44 +0800 Subject: [PATCH 1/8] =?UTF-8?q?fix:=20Bedrock=20=E8=B4=A6=E6=88=B7?= =?UTF-8?q?=E9=85=8D=E9=A2=9D=E9=99=90=E5=88=B6=E4=B8=8D=E7=94=9F=E6=95=88?= =?UTF-8?q?=EF=BC=8C=E9=85=8D=E9=A2=9D=E8=AE=A1=E6=95=B0=E5=99=A8=E5=A7=8B?= =?UTF-8?q?=E7=BB=88=E4=B8=BA=20$0.00?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit applyUsageBillingEffects() 中配额更新条件仅检查了 AccountTypeAPIKey， 遗漏了 AccountTypeBedrock，导致 Bedrock 账户的配额计数器永远不递增。 扩展条件以同时支持 APIKey 和 Bedrock 类型。 同时在前端账户筛选下拉框中添加 AWS Bedrock 选项。 --- backend/internal/repository/usage_billing_repo.go | 2 +- frontend/src/components/admin/account/AccountTableFilters.vue | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/backend/internal/repository/usage_billing_repo.go b/backend/internal/repository/usage_billing_repo.go index b13cfeb8..b4c76da5 100644 --- a/backend/internal/repository/usage_billing_repo.go +++ b/backend/internal/repository/usage_billing_repo.go @@ -132,7 +132,7 @@ func (r *usageBillingRepository) applyUsageBillingEffects(ctx context.Context, t } } - if cmd.AccountQuotaCost > 0 && strings.EqualFold(cmd.AccountType, service.AccountTypeAPIKey) { + if cmd.AccountQuotaCost > 0 && (strings.EqualFold(cmd.AccountType, service.AccountTypeAPIKey) || strings.EqualFold(cmd.AccountType, service.AccountTypeBedrock)) { if err := incrementUsageBillingAccountQuota(ctx, tx, cmd.AccountID, cmd.AccountQuotaCost); err != nil { return err } diff --git a/frontend/src/components/admin/account/AccountTableFilters.vue b/frontend/src/components/admin/account/AccountTableFilters.vue index abffbaa2..d8068336 100644 --- a/frontend/src/components/admin/account/AccountTableFilters.vue +++ b/frontend/src/components/admin/account/AccountTableFilters.vue @@ -24,7 +24,7 @@ const updateType = (value: string | number | boolean | null) => { emit('update:f const updateStatus = (value: string | number | boolean | null) => { emit('update:filters', { ...props.filters, status: value }) } const updateGroup = (value: string | number | boolean | null) => { emit('update:filters', { ...props.filters, group: value }) } const pOpts = computed(() => [{ value: '', label: t('admin.accounts.allPlatforms') }, { value: 'anthropic', label: 'Anthropic' }, { value: 'openai', label: 'OpenAI' }, { value: 'gemini', label: 'Gemini' }, { value: 'antigravity', label: 'Antigravity' }, { value: 'sora', label: 'Sora' }]) -const tOpts = computed(() => [{ value: '', label: t('admin.accounts.allTypes') }, { value: 'oauth', label: t('admin.accounts.oauthType') }, { value: 'setup-token', label: t('admin.accounts.setupToken') }, { value: 'apikey', label: t('admin.accounts.apiKey') }]) +const tOpts = computed(() => [{ value: '', label: t('admin.accounts.allTypes') }, { value: 'oauth', label: t('admin.accounts.oauthType') }, { value: 'setup-token', label: t('admin.accounts.setupToken') }, { value: 'apikey', label: t('admin.accounts.apiKey') }, { value: 'bedrock', label: 'AWS Bedrock' }]) const sOpts = computed(() => [{ value: '', label: t('admin.accounts.allStatus') }, { value: 'active', label: t('admin.accounts.status.active') }, { value: 'inactive', label: t('admin.accounts.status.inactive') }, { value: 'error', label: t('admin.accounts.status.error') }, { value: 'rate_limited', label: t('admin.accounts.status.rateLimited') }, { value: 'temp_unschedulable', label: t('admin.accounts.status.tempUnschedulable') }]) const gOpts = computed(() => [{ value: '', label: t('admin.accounts.allGroups') }, ...(props.groups || []).map(g => ({ value: String(g.id), label: g.name }))]) From 0e237326314b091d2c455e43ba56dad69f2106ae Mon Sep 17 00:00:00 2001 From: Elysia <1628615876@qq.com> Date: Sat, 14 Mar 2026 22:49:23 +0800 Subject: [PATCH 2/8] =?UTF-8?q?fix(gateway):=20=E9=98=B2=E6=AD=A2=E6=B5=81?= =?UTF-8?q?=E5=BC=8F=20failover=20=E6=8B=BC=E6=8E=A5=E8=85=90=E5=8C=96?= =?UTF-8?q?=E5=AF=BC=E8=87=B4=E5=AE=A2=E6=88=B7=E7=AB=AF=E6=94=B6=E5=88=B0?= =?UTF-8?q?=E5=8F=8C=20message=5Fstart?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit 当上游在 SSE 流中途返回 event:error 时，handleStreamingResponse 已将 部分 SSE 事件写入客户端，但原先的 failover 逻辑仍会切换到下一个账号 并写入完整流，导致客户端收到两个 message_start 进而产生 400 错误。 修复方案：在每次 Forward 调用前记录 c.Writer.Size()，若 Forward 返回 UpstreamFailoverError 后 writer 字节数增加，说明 SSE 内容已不可撤销地 发送给客户端，此时直接调用 handleFailoverExhausted 发送 SSE error 事件 终止流，而非继续 failover。 Ping-only 场景不受影响：slot 等待期的 ping 字节在 Forward 前后相等， 正常 failover 流程照常进行。 Co-Authored-By: Claude Sonnet 4.6 --- backend/internal/handler/gateway_handler.go | 14 ++++++++++++++ 1 file changed, 14 insertions(+) diff --git a/backend/internal/handler/gateway_handler.go b/backend/internal/handler/gateway_handler.go index 676ba0e1..a2fd3196 100644 --- a/backend/internal/handler/gateway_handler.go +++ b/backend/internal/handler/gateway_handler.go @@ -391,6 +391,8 @@ func (h *GatewayHandler) Messages(c *gin.Context) { if fs.SwitchCount > 0 { requestCtx = service.WithAccountSwitchCount(requestCtx, fs.SwitchCount, h.metadataBridgeEnabled()) } + // 记录 Forward 前已写入字节数，Forward 后若增加则说明 SSE 内容已发，禁止 failover + writerSizeBeforeForward := c.Writer.Size() if account.Platform == service.PlatformAntigravity { result, err = h.antigravityGatewayService.ForwardGemini(requestCtx, c, account, reqModel, "generateContent", reqStream, body, hasBoundSession) } else { @@ -402,6 +404,11 @@ func (h *GatewayHandler) Messages(c *gin.Context) { if err != nil { var failoverErr *service.UpstreamFailoverError if errors.As(err, &failoverErr) { + // 流式内容已写入客户端，无法撤销，禁止 failover 以防止流拼接腐化 + if c.Writer.Size() != writerSizeBeforeForward { + h.handleFailoverExhausted(c, failoverErr, account.Platform, true) + return + } action := fs.HandleFailoverError(c.Request.Context(), h.gatewayService, account.ID, account.Platform, failoverErr) switch action { case FailoverContinue: @@ -637,6 +644,8 @@ func (h *GatewayHandler) Messages(c *gin.Context) { if fs.SwitchCount > 0 { requestCtx = service.WithAccountSwitchCount(requestCtx, fs.SwitchCount, h.metadataBridgeEnabled()) } + // 记录 Forward 前已写入字节数，Forward 后若增加则说明 SSE 内容已发，禁止 failover + writerSizeBeforeForward := c.Writer.Size() if account.Platform == service.PlatformAntigravity && account.Type != service.AccountTypeAPIKey { result, err = h.antigravityGatewayService.Forward(requestCtx, c, account, body, hasBoundSession) } else { @@ -706,6 +715,11 @@ func (h *GatewayHandler) Messages(c *gin.Context) { } var failoverErr *service.UpstreamFailoverError if errors.As(err, &failoverErr) { + // 流式内容已写入客户端，无法撤销，禁止 failover 以防止流拼接腐化 + if c.Writer.Size() != writerSizeBeforeForward { + h.handleFailoverExhausted(c, failoverErr, account.Platform, true) + return + } action := fs.HandleFailoverError(c.Request.Context(), h.gatewayService, account.ID, account.Platform, failoverErr) switch action { case FailoverContinue: From eefab159581819c6a048ccd70dad51cfdcd4af7a Mon Sep 17 00:00:00 2001 From: Ethan0x0000 <3352979663@qq.com> Date: Sun, 15 Mar 2026 11:26:42 +0800 Subject: [PATCH 3/8] =?UTF-8?q?feat:=20=E5=AE=8C=E5=96=84=E4=BD=BF?= =?UTF-8?q?=E7=94=A8=E8=AE=B0=E5=BD=95=E7=AB=AF=E7=82=B9=E5=8F=AF=E8=A7=82?= =?UTF-8?q?=E6=B5=8B=E6=80=A7=E4=B8=8E=E5=88=86=E5=B8=83=E7=BB=9F=E8=AE=A1?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit 将入站、上游与路径三类端点分布统一到使用记录页的一致化卡片交互中，并补齐端点元数据与统计链路，提升排障与流量分析效率。 --- backend/internal/handler/dto/mappers.go | 2 + .../handler/dto/mappers_usage_test.go | 12 + backend/internal/handler/dto/types.go | 4 + .../handler/openai_chat_completions.go | 18 +- ...nai_gateway_endpoint_normalization_test.go | 57 ++++ .../handler/openai_gateway_handler.go | 69 +++++ .../pkg/usagestats/usage_log_types.go | 38 ++- backend/internal/repository/usage_log_repo.go | 250 ++++++++++++++++- .../usage_log_repo_request_type_test.go | 10 + .../internal/service/account_usage_service.go | 2 + .../openai_gateway_record_usage_test.go | 35 +++ .../service/openai_gateway_service.go | 13 +- backend/internal/service/usage_log.go | 4 + .../074_add_usage_log_endpoints.sql | 5 + frontend/src/api/admin/usage.ts | 4 + .../components/account/AccountStatsModal.vue | 13 + .../admin/account/AccountStatsModal.vue | 13 + .../src/components/admin/usage/UsageTable.vue | 14 + .../charts/EndpointDistributionChart.vue | 257 ++++++++++++++++++ frontend/src/i18n/locales/en.ts | 7 + frontend/src/i18n/locales/zh.ts | 7 + frontend/src/types/index.ts | 12 + frontend/src/views/admin/UsageView.vue | 55 +++- frontend/src/views/user/UsageView.vue | 14 + 24 files changed, 878 insertions(+), 37 deletions(-) create mode 100644 backend/internal/handler/openai_gateway_endpoint_normalization_test.go create mode 100644 backend/migrations/074_add_usage_log_endpoints.sql create mode 100644 frontend/src/components/charts/EndpointDistributionChart.vue diff --git a/backend/internal/handler/dto/mappers.go b/backend/internal/handler/dto/mappers.go index 3706f725..8e5f23e7 100644 --- a/backend/internal/handler/dto/mappers.go +++ b/backend/internal/handler/dto/mappers.go @@ -523,6 +523,8 @@ func usageLogFromServiceUser(l *service.UsageLog) UsageLog { Model: l.Model, ServiceTier: l.ServiceTier, ReasoningEffort: l.ReasoningEffort, + InboundEndpoint: l.InboundEndpoint, + UpstreamEndpoint: l.UpstreamEndpoint, GroupID: l.GroupID, SubscriptionID: l.SubscriptionID, InputTokens: l.InputTokens, diff --git a/backend/internal/handler/dto/mappers_usage_test.go b/backend/internal/handler/dto/mappers_usage_test.go index ea408ecb..e4031970 100644 --- a/backend/internal/handler/dto/mappers_usage_test.go +++ b/backend/internal/handler/dto/mappers_usage_test.go @@ -76,10 +76,14 @@ func TestUsageLogFromService_IncludesServiceTierForUserAndAdmin(t *testing.T) { t.Parallel() serviceTier := "priority" + inboundEndpoint := "/v1/chat/completions" + upstreamEndpoint := "/v1/responses" log := &service.UsageLog{ RequestID: "req_3", Model: "gpt-5.4", ServiceTier: &serviceTier, + InboundEndpoint: &inboundEndpoint, + UpstreamEndpoint: &upstreamEndpoint, AccountRateMultiplier: f64Ptr(1.5), } @@ -88,8 +92,16 @@ func TestUsageLogFromService_IncludesServiceTierForUserAndAdmin(t *testing.T) { require.NotNil(t, userDTO.ServiceTier) require.Equal(t, serviceTier, *userDTO.ServiceTier) + require.NotNil(t, userDTO.InboundEndpoint) + require.Equal(t, inboundEndpoint, *userDTO.InboundEndpoint) + require.NotNil(t, userDTO.UpstreamEndpoint) + require.Equal(t, upstreamEndpoint, *userDTO.UpstreamEndpoint) require.NotNil(t, adminDTO.ServiceTier) require.Equal(t, serviceTier, *adminDTO.ServiceTier) + require.NotNil(t, adminDTO.InboundEndpoint) + require.Equal(t, inboundEndpoint, *adminDTO.InboundEndpoint) + require.NotNil(t, adminDTO.UpstreamEndpoint) + require.Equal(t, upstreamEndpoint, *adminDTO.UpstreamEndpoint) require.NotNil(t, adminDTO.AccountRateMultiplier) require.InDelta(t, 1.5, *adminDTO.AccountRateMultiplier, 1e-12) } diff --git a/backend/internal/handler/dto/types.go b/backend/internal/handler/dto/types.go index 3708eed5..20e83973 100644 --- a/backend/internal/handler/dto/types.go +++ b/backend/internal/handler/dto/types.go @@ -337,6 +337,10 @@ type UsageLog struct { // ReasoningEffort is the request's reasoning effort level (OpenAI Responses API). // nil means not provided / not applicable. ReasoningEffort *string `json:"reasoning_effort,omitempty"` + // InboundEndpoint is the client-facing API endpoint path, e.g. /v1/chat/completions. + InboundEndpoint *string `json:"inbound_endpoint,omitempty"` + // UpstreamEndpoint is the normalized upstream endpoint path, e.g. /v1/responses. + UpstreamEndpoint *string `json:"upstream_endpoint,omitempty"` GroupID *int64 `json:"group_id"` SubscriptionID *int64 `json:"subscription_id"` diff --git a/backend/internal/handler/openai_chat_completions.go b/backend/internal/handler/openai_chat_completions.go index 724376e3..82b11c10 100644 --- a/backend/internal/handler/openai_chat_completions.go +++ b/backend/internal/handler/openai_chat_completions.go @@ -256,14 +256,16 @@ func (h *OpenAIGatewayHandler) ChatCompletions(c *gin.Context) { h.submitUsageRecordTask(func(ctx context.Context) { if err := h.gatewayService.RecordUsage(ctx, &service.OpenAIRecordUsageInput{ - Result: result, - APIKey: apiKey, - User: apiKey.User, - Account: account, - Subscription: subscription, - UserAgent: userAgent, - IPAddress: clientIP, - APIKeyService: h.apiKeyService, + Result: result, + APIKey: apiKey, + User: apiKey.User, + Account: account, + Subscription: subscription, + InboundEndpoint: normalizedOpenAIInboundEndpoint(c, openAIInboundEndpointChatCompletions), + UpstreamEndpoint: normalizedOpenAIUpstreamEndpoint(c, openAIUpstreamEndpointResponses), + UserAgent: userAgent, + IPAddress: clientIP, + APIKeyService: h.apiKeyService, }); err != nil { logger.L().With( zap.String("component", "handler.openai_gateway.chat_completions"), diff --git a/backend/internal/handler/openai_gateway_endpoint_normalization_test.go b/backend/internal/handler/openai_gateway_endpoint_normalization_test.go new file mode 100644 index 00000000..6a055272 --- /dev/null +++ b/backend/internal/handler/openai_gateway_endpoint_normalization_test.go @@ -0,0 +1,57 @@ +package handler + +import ( + "net/http" + "net/http/httptest" + "testing" + + "github.com/gin-gonic/gin" + "github.com/stretchr/testify/require" +) + +func TestNormalizedOpenAIUpstreamEndpoint(t *testing.T) { + gin.SetMode(gin.TestMode) + + tests := []struct { + name string + path string + fallback string + want string + }{ + { + name: "responses root maps to responses upstream", + path: "/v1/responses", + fallback: openAIUpstreamEndpointResponses, + want: "/v1/responses", + }, + { + name: "responses compact keeps compact suffix", + path: "/openai/v1/responses/compact", + fallback: openAIUpstreamEndpointResponses, + want: "/v1/responses/compact", + }, + { + name: "responses nested suffix preserved", + path: "/openai/v1/responses/compact/detail", + fallback: openAIUpstreamEndpointResponses, + want: "/v1/responses/compact/detail", + }, + { + name: "non responses path uses fallback", + path: "/v1/messages", + fallback: openAIUpstreamEndpointResponses, + want: "/v1/responses", + }, + } + + for _, tt := range tests { + t.Run(tt.name, func(t *testing.T) { + rec := httptest.NewRecorder() + c, _ := gin.CreateTestContext(rec) + c.Request = httptest.NewRequest(http.MethodPost, tt.path, nil) + + got := normalizedOpenAIUpstreamEndpoint(c, tt.fallback) + require.Equal(t, tt.want, got) + }) + } +} diff --git a/backend/internal/handler/openai_gateway_handler.go b/backend/internal/handler/openai_gateway_handler.go index 87b0d0d6..b2aa5c50 100644 --- a/backend/internal/handler/openai_gateway_handler.go +++ b/backend/internal/handler/openai_gateway_handler.go @@ -37,6 +37,13 @@ type OpenAIGatewayHandler struct { cfg *config.Config } +const ( + openAIInboundEndpointResponses = "/v1/responses" + openAIInboundEndpointMessages = "/v1/messages" + openAIInboundEndpointChatCompletions = "/v1/chat/completions" + openAIUpstreamEndpointResponses = "/v1/responses" +) + // NewOpenAIGatewayHandler creates a new OpenAIGatewayHandler func NewOpenAIGatewayHandler( gatewayService *service.OpenAIGatewayService, @@ -362,6 +369,8 @@ func (h *OpenAIGatewayHandler) Responses(c *gin.Context) { User: apiKey.User, Account: account, Subscription: subscription, + InboundEndpoint: normalizedOpenAIInboundEndpoint(c, openAIInboundEndpointResponses), + UpstreamEndpoint: normalizedOpenAIUpstreamEndpoint(c, openAIUpstreamEndpointResponses), UserAgent: userAgent, IPAddress: clientIP, RequestPayloadHash: requestPayloadHash, @@ -738,6 +747,8 @@ func (h *OpenAIGatewayHandler) Messages(c *gin.Context) { User: apiKey.User, Account: account, Subscription: subscription, + InboundEndpoint: normalizedOpenAIInboundEndpoint(c, openAIInboundEndpointMessages), + UpstreamEndpoint: normalizedOpenAIUpstreamEndpoint(c, openAIUpstreamEndpointResponses), UserAgent: userAgent, IPAddress: clientIP, RequestPayloadHash: requestPayloadHash, @@ -1235,6 +1246,8 @@ func (h *OpenAIGatewayHandler) ResponsesWebSocket(c *gin.Context) { User: apiKey.User, Account: account, Subscription: subscription, + InboundEndpoint: normalizedOpenAIInboundEndpoint(c, openAIInboundEndpointResponses), + UpstreamEndpoint: normalizedOpenAIUpstreamEndpoint(c, openAIUpstreamEndpointResponses), UserAgent: userAgent, IPAddress: clientIP, RequestPayloadHash: service.HashUsageRequestPayload(firstMessage), @@ -1530,6 +1543,62 @@ func openAIWSIngressFallbackSessionSeed(userID, apiKeyID int64, groupID *int64) return fmt.Sprintf("openai_ws_ingress:%d:%d:%d", gid, userID, apiKeyID) } +func normalizedOpenAIInboundEndpoint(c *gin.Context, fallback string) string { + path := strings.TrimSpace(fallback) + if c != nil { + if fullPath := strings.TrimSpace(c.FullPath()); fullPath != "" { + path = fullPath + } else if c.Request != nil && c.Request.URL != nil { + if requestPath := strings.TrimSpace(c.Request.URL.Path); requestPath != "" { + path = requestPath + } + } + } + + switch { + case strings.Contains(path, openAIInboundEndpointChatCompletions): + return openAIInboundEndpointChatCompletions + case strings.Contains(path, openAIInboundEndpointMessages): + return openAIInboundEndpointMessages + case strings.Contains(path, openAIInboundEndpointResponses): + return openAIInboundEndpointResponses + default: + return path + } +} + +func normalizedOpenAIUpstreamEndpoint(c *gin.Context, fallback string) string { + base := strings.TrimSpace(fallback) + if base == "" { + base = openAIUpstreamEndpointResponses + } + base = strings.TrimRight(base, "/") + + if c == nil || c.Request == nil || c.Request.URL == nil { + return base + } + + path := strings.TrimRight(strings.TrimSpace(c.Request.URL.Path), "/") + if path == "" { + return base + } + + idx := strings.LastIndex(path, "/responses") + if idx < 0 { + return base + } + + suffix := strings.TrimSpace(path[idx+len("/responses"):]) + if suffix == "" || suffix == "/" { + return base + } + if !strings.HasPrefix(suffix, "/") { + return base + } + + return base + suffix +} + func isOpenAIWSUpgradeRequest(r *http.Request) bool { if r == nil { return false diff --git a/backend/internal/pkg/usagestats/usage_log_types.go b/backend/internal/pkg/usagestats/usage_log_types.go index 55a049d3..e9a5cae5 100644 --- a/backend/internal/pkg/usagestats/usage_log_types.go +++ b/backend/internal/pkg/usagestats/usage_log_types.go @@ -81,6 +81,15 @@ type ModelStat struct { ActualCost float64 `json:"actual_cost"` // 实际扣除 } +// EndpointStat represents usage statistics for a single request endpoint. +type EndpointStat struct { + Endpoint string `json:"endpoint"` + Requests int64 `json:"requests"` + TotalTokens int64 `json:"total_tokens"` + Cost float64 `json:"cost"` // 标准计费 + ActualCost float64 `json:"actual_cost"` // 实际扣除 +} + // GroupStat represents usage statistics for a single group type GroupStat struct { GroupID int64 `json:"group_id"` @@ -179,15 +188,18 @@ type UsageLogFilters struct { // UsageStats represents usage statistics type UsageStats struct { - TotalRequests int64 `json:"total_requests"` - TotalInputTokens int64 `json:"total_input_tokens"` - TotalOutputTokens int64 `json:"total_output_tokens"` - TotalCacheTokens int64 `json:"total_cache_tokens"` - TotalTokens int64 `json:"total_tokens"` - TotalCost float64 `json:"total_cost"` - TotalActualCost float64 `json:"total_actual_cost"` - TotalAccountCost *float64 `json:"total_account_cost,omitempty"` - AverageDurationMs float64 `json:"average_duration_ms"` + TotalRequests int64 `json:"total_requests"` + TotalInputTokens int64 `json:"total_input_tokens"` + TotalOutputTokens int64 `json:"total_output_tokens"` + TotalCacheTokens int64 `json:"total_cache_tokens"` + TotalTokens int64 `json:"total_tokens"` + TotalCost float64 `json:"total_cost"` + TotalActualCost float64 `json:"total_actual_cost"` + TotalAccountCost *float64 `json:"total_account_cost,omitempty"` + AverageDurationMs float64 `json:"average_duration_ms"` + Endpoints []EndpointStat `json:"endpoints,omitempty"` + UpstreamEndpoints []EndpointStat `json:"upstream_endpoints,omitempty"` + EndpointPaths []EndpointStat `json:"endpoint_paths,omitempty"` } // BatchUserUsageStats represents usage stats for a single user @@ -254,7 +266,9 @@ type AccountUsageSummary struct { // AccountUsageStatsResponse represents the full usage statistics response for an account type AccountUsageStatsResponse struct { - History []AccountUsageHistory `json:"history"` - Summary AccountUsageSummary `json:"summary"` - Models []ModelStat `json:"models"` + History []AccountUsageHistory `json:"history"` + Summary AccountUsageSummary `json:"summary"` + Models []ModelStat `json:"models"` + Endpoints []EndpointStat `json:"endpoints"` + UpstreamEndpoints []EndpointStat `json:"upstream_endpoints"` } diff --git a/backend/internal/repository/usage_log_repo.go b/backend/internal/repository/usage_log_repo.go index 845f2cf0..cc949db2 100644 --- a/backend/internal/repository/usage_log_repo.go +++ b/backend/internal/repository/usage_log_repo.go @@ -28,7 +28,7 @@ import ( gocache "github.com/patrickmn/go-cache" ) -const usageLogSelectColumns = "id, user_id, api_key_id, account_id, request_id, model, group_id, subscription_id, input_tokens, output_tokens, cache_creation_tokens, cache_read_tokens, cache_creation_5m_tokens, cache_creation_1h_tokens, input_cost, output_cost, cache_creation_cost, cache_read_cost, total_cost, actual_cost, rate_multiplier, account_rate_multiplier, billing_type, request_type, stream, openai_ws_mode, duration_ms, first_token_ms, user_agent, ip_address, image_count, image_size, media_type, service_tier, reasoning_effort, cache_ttl_overridden, created_at" +const usageLogSelectColumns = "id, user_id, api_key_id, account_id, request_id, model, group_id, subscription_id, input_tokens, output_tokens, cache_creation_tokens, cache_read_tokens, cache_creation_5m_tokens, cache_creation_1h_tokens, input_cost, output_cost, cache_creation_cost, cache_read_cost, total_cost, actual_cost, rate_multiplier, account_rate_multiplier, billing_type, request_type, stream, openai_ws_mode, duration_ms, first_token_ms, user_agent, ip_address, image_count, image_size, media_type, service_tier, reasoning_effort, inbound_endpoint, upstream_endpoint, cache_ttl_overridden, created_at" var usageLogInsertArgTypes = [...]string{ "bigint", @@ -65,6 +65,8 @@ var usageLogInsertArgTypes = [...]string{ "text", "text", "text", + "text", + "text", "boolean", "timestamptz", } @@ -304,6 +306,8 @@ func (r *usageLogRepository) createSingle(ctx context.Context, sqlq sqlExecutor, media_type, service_tier, reasoning_effort, + inbound_endpoint, + upstream_endpoint, cache_ttl_overridden, created_at ) VALUES ( @@ -312,7 +316,7 @@ func (r *usageLogRepository) createSingle(ctx context.Context, sqlq sqlExecutor, $8, $9, $10, $11, $12, $13, $14, $15, $16, $17, $18, $19, - $20, $21, $22, $23, $24, $25, $26, $27, $28, $29, $30, $31, $32, $33, $34, $35, $36 + $20, $21, $22, $23, $24, $25, $26, $27, $28, $29, $30, $31, $32, $33, $34, $35, $36, $37, $38 ) ON CONFLICT (request_id, api_key_id) DO NOTHING RETURNING id, created_at @@ -732,11 +736,13 @@ func buildUsageLogBatchInsertQuery(keys []string, preparedByKey map[string]usage media_type, service_tier, reasoning_effort, + inbound_endpoint, + upstream_endpoint, cache_ttl_overridden, created_at ) AS (VALUES `) - args := make([]any, 0, len(keys)*37) + args := make([]any, 0, len(keys)*38) argPos := 1 for idx, key := range keys { if idx > 0 { @@ -799,6 +805,8 @@ func buildUsageLogBatchInsertQuery(keys []string, preparedByKey map[string]usage media_type, service_tier, reasoning_effort, + inbound_endpoint, + upstream_endpoint, cache_ttl_overridden, created_at ) @@ -837,6 +845,8 @@ func buildUsageLogBatchInsertQuery(keys []string, preparedByKey map[string]usage media_type, service_tier, reasoning_effort, + inbound_endpoint, + upstream_endpoint, cache_ttl_overridden, created_at FROM input @@ -915,11 +925,13 @@ func buildUsageLogBestEffortInsertQuery(preparedList []usageLogInsertPrepared) ( media_type, service_tier, reasoning_effort, + inbound_endpoint, + upstream_endpoint, cache_ttl_overridden, created_at ) AS (VALUES `) - args := make([]any, 0, len(preparedList)*36) + args := make([]any, 0, len(preparedList)*38) argPos := 1 for idx, prepared := range preparedList { if idx > 0 { @@ -979,6 +991,8 @@ func buildUsageLogBestEffortInsertQuery(preparedList []usageLogInsertPrepared) ( media_type, service_tier, reasoning_effort, + inbound_endpoint, + upstream_endpoint, cache_ttl_overridden, created_at ) @@ -1017,6 +1031,8 @@ func buildUsageLogBestEffortInsertQuery(preparedList []usageLogInsertPrepared) ( media_type, service_tier, reasoning_effort, + inbound_endpoint, + upstream_endpoint, cache_ttl_overridden, created_at FROM input @@ -1063,6 +1079,8 @@ func execUsageLogInsertNoResult(ctx context.Context, sqlq sqlExecutor, prepared media_type, service_tier, reasoning_effort, + inbound_endpoint, + upstream_endpoint, cache_ttl_overridden, created_at ) VALUES ( @@ -1071,7 +1089,7 @@ func execUsageLogInsertNoResult(ctx context.Context, sqlq sqlExecutor, prepared $8, $9, $10, $11, $12, $13, $14, $15, $16, $17, $18, $19, - $20, $21, $22, $23, $24, $25, $26, $27, $28, $29, $30, $31, $32, $33, $34, $35, $36 + $20, $21, $22, $23, $24, $25, $26, $27, $28, $29, $30, $31, $32, $33, $34, $35, $36, $37, $38 ) ON CONFLICT (request_id, api_key_id) DO NOTHING `, prepared.args...) @@ -1101,6 +1119,8 @@ func prepareUsageLogInsert(log *service.UsageLog) usageLogInsertPrepared { mediaType := nullString(log.MediaType) serviceTier := nullString(log.ServiceTier) reasoningEffort := nullString(log.ReasoningEffort) + inboundEndpoint := nullString(log.InboundEndpoint) + upstreamEndpoint := nullString(log.UpstreamEndpoint) var requestIDArg any if requestID != "" { @@ -1147,6 +1167,8 @@ func prepareUsageLogInsert(log *service.UsageLog) usageLogInsertPrepared { mediaType, serviceTier, reasoningEffort, + inboundEndpoint, + upstreamEndpoint, log.CacheTTLOverridden, createdAt, }, @@ -2505,7 +2527,7 @@ func (r *usageLogRepository) ListWithFilters(ctx context.Context, params paginat args = append(args, *filters.StartTime) } if filters.EndTime != nil { - conditions = append(conditions, fmt.Sprintf("created_at <= $%d", len(args)+1)) + conditions = append(conditions, fmt.Sprintf("created_at < $%d", len(args)+1)) args = append(args, *filters.EndTime) } @@ -3040,7 +3062,7 @@ func (r *usageLogRepository) GetStatsWithFilters(ctx context.Context, filters Us args = append(args, *filters.StartTime) } if filters.EndTime != nil { - conditions = append(conditions, fmt.Sprintf("created_at <= $%d", len(args)+1)) + conditions = append(conditions, fmt.Sprintf("created_at < $%d", len(args)+1)) args = append(args, *filters.EndTime) } @@ -3080,6 +3102,35 @@ func (r *usageLogRepository) GetStatsWithFilters(ctx context.Context, filters Us stats.TotalAccountCost = &totalAccountCost } stats.TotalTokens = stats.TotalInputTokens + stats.TotalOutputTokens + stats.TotalCacheTokens + + start := time.Unix(0, 0).UTC() + if filters.StartTime != nil { + start = *filters.StartTime + } + end := time.Now().UTC() + if filters.EndTime != nil { + end = *filters.EndTime + } + + endpoints, endpointErr := r.GetEndpointStatsWithFilters(ctx, start, end, filters.UserID, filters.APIKeyID, filters.AccountID, filters.GroupID, filters.Model, filters.RequestType, filters.Stream, filters.BillingType) + if endpointErr != nil { + logger.LegacyPrintf("repository.usage_log", "GetEndpointStatsWithFilters failed in GetStatsWithFilters: %v", endpointErr) + endpoints = []EndpointStat{} + } + upstreamEndpoints, upstreamEndpointErr := r.GetUpstreamEndpointStatsWithFilters(ctx, start, end, filters.UserID, filters.APIKeyID, filters.AccountID, filters.GroupID, filters.Model, filters.RequestType, filters.Stream, filters.BillingType) + if upstreamEndpointErr != nil { + logger.LegacyPrintf("repository.usage_log", "GetUpstreamEndpointStatsWithFilters failed in GetStatsWithFilters: %v", upstreamEndpointErr) + upstreamEndpoints = []EndpointStat{} + } + endpointPaths, endpointPathErr := r.getEndpointPathStatsWithFilters(ctx, start, end, filters.UserID, filters.APIKeyID, filters.AccountID, filters.GroupID, filters.Model, filters.RequestType, filters.Stream, filters.BillingType) + if endpointPathErr != nil { + logger.LegacyPrintf("repository.usage_log", "getEndpointPathStatsWithFilters failed in GetStatsWithFilters: %v", endpointPathErr) + endpointPaths = []EndpointStat{} + } + stats.Endpoints = endpoints + stats.UpstreamEndpoints = upstreamEndpoints + stats.EndpointPaths = endpointPaths + return stats, nil } @@ -3092,6 +3143,163 @@ type AccountUsageSummary = usagestats.AccountUsageSummary // AccountUsageStatsResponse represents the full usage statistics response for an account type AccountUsageStatsResponse = usagestats.AccountUsageStatsResponse +// EndpointStat represents endpoint usage statistics row. +type EndpointStat = usagestats.EndpointStat + +func (r *usageLogRepository) getEndpointStatsByColumnWithFilters(ctx context.Context, endpointColumn string, startTime, endTime time.Time, userID, apiKeyID, accountID, groupID int64, model string, requestType *int16, stream *bool, billingType *int8) (results []EndpointStat, err error) { + actualCostExpr := "COALESCE(SUM(actual_cost), 0) as actual_cost" + if accountID > 0 && userID == 0 && apiKeyID == 0 { + actualCostExpr = "COALESCE(SUM(total_cost * COALESCE(account_rate_multiplier, 1)), 0) as actual_cost" + } + + query := fmt.Sprintf(` + SELECT + COALESCE(NULLIF(TRIM(%s), ''), 'unknown') AS endpoint, + COUNT(*) AS requests, + COALESCE(SUM(input_tokens + output_tokens + cache_creation_tokens + cache_read_tokens), 0) AS total_tokens, + COALESCE(SUM(total_cost), 0) as cost, + %s + FROM usage_logs + WHERE created_at >= $1 AND created_at < $2 + `, endpointColumn, actualCostExpr) + + args := []any{startTime, endTime} + if userID > 0 { + query += fmt.Sprintf(" AND user_id = $%d", len(args)+1) + args = append(args, userID) + } + if apiKeyID > 0 { + query += fmt.Sprintf(" AND api_key_id = $%d", len(args)+1) + args = append(args, apiKeyID) + } + if accountID > 0 { + query += fmt.Sprintf(" AND account_id = $%d", len(args)+1) + args = append(args, accountID) + } + if groupID > 0 { + query += fmt.Sprintf(" AND group_id = $%d", len(args)+1) + args = append(args, groupID) + } + if model != "" { + query += fmt.Sprintf(" AND model = $%d", len(args)+1) + args = append(args, model) + } + query, args = appendRequestTypeOrStreamQueryFilter(query, args, requestType, stream) + if billingType != nil { + query += fmt.Sprintf(" AND billing_type = $%d", len(args)+1) + args = append(args, int16(*billingType)) + } + query += " GROUP BY endpoint ORDER BY requests DESC" + + rows, err := r.sql.QueryContext(ctx, query, args...) + if err != nil { + return nil, err + } + defer func() { + if closeErr := rows.Close(); closeErr != nil && err == nil { + err = closeErr + results = nil + } + }() + + results = make([]EndpointStat, 0) + for rows.Next() { + var row EndpointStat + if err := rows.Scan(&row.Endpoint, &row.Requests, &row.TotalTokens, &row.Cost, &row.ActualCost); err != nil { + return nil, err + } + results = append(results, row) + } + if err := rows.Err(); err != nil { + return nil, err + } + return results, nil +} + +func (r *usageLogRepository) getEndpointPathStatsWithFilters(ctx context.Context, startTime, endTime time.Time, userID, apiKeyID, accountID, groupID int64, model string, requestType *int16, stream *bool, billingType *int8) (results []EndpointStat, err error) { + actualCostExpr := "COALESCE(SUM(actual_cost), 0) as actual_cost" + if accountID > 0 && userID == 0 && apiKeyID == 0 { + actualCostExpr = "COALESCE(SUM(total_cost * COALESCE(account_rate_multiplier, 1)), 0) as actual_cost" + } + + query := fmt.Sprintf(` + SELECT + CONCAT( + COALESCE(NULLIF(TRIM(inbound_endpoint), ''), 'unknown'), + ' -> ', + COALESCE(NULLIF(TRIM(upstream_endpoint), ''), 'unknown') + ) AS endpoint, + COUNT(*) AS requests, + COALESCE(SUM(input_tokens + output_tokens + cache_creation_tokens + cache_read_tokens), 0) AS total_tokens, + COALESCE(SUM(total_cost), 0) as cost, + %s + FROM usage_logs + WHERE created_at >= $1 AND created_at < $2 + `, actualCostExpr) + + args := []any{startTime, endTime} + if userID > 0 { + query += fmt.Sprintf(" AND user_id = $%d", len(args)+1) + args = append(args, userID) + } + if apiKeyID > 0 { + query += fmt.Sprintf(" AND api_key_id = $%d", len(args)+1) + args = append(args, apiKeyID) + } + if accountID > 0 { + query += fmt.Sprintf(" AND account_id = $%d", len(args)+1) + args = append(args, accountID) + } + if groupID > 0 { + query += fmt.Sprintf(" AND group_id = $%d", len(args)+1) + args = append(args, groupID) + } + if model != "" { + query += fmt.Sprintf(" AND model = $%d", len(args)+1) + args = append(args, model) + } + query, args = appendRequestTypeOrStreamQueryFilter(query, args, requestType, stream) + if billingType != nil { + query += fmt.Sprintf(" AND billing_type = $%d", len(args)+1) + args = append(args, int16(*billingType)) + } + query += " GROUP BY endpoint ORDER BY requests DESC" + + rows, err := r.sql.QueryContext(ctx, query, args...) + if err != nil { + return nil, err + } + defer func() { + if closeErr := rows.Close(); closeErr != nil && err == nil { + err = closeErr + results = nil + } + }() + + results = make([]EndpointStat, 0) + for rows.Next() { + var row EndpointStat + if err := rows.Scan(&row.Endpoint, &row.Requests, &row.TotalTokens, &row.Cost, &row.ActualCost); err != nil { + return nil, err + } + results = append(results, row) + } + if err := rows.Err(); err != nil { + return nil, err + } + return results, nil +} + +// GetEndpointStatsWithFilters returns inbound endpoint statistics with optional filters. +func (r *usageLogRepository) GetEndpointStatsWithFilters(ctx context.Context, startTime, endTime time.Time, userID, apiKeyID, accountID, groupID int64, model string, requestType *int16, stream *bool, billingType *int8) ([]EndpointStat, error) { + return r.getEndpointStatsByColumnWithFilters(ctx, "inbound_endpoint", startTime, endTime, userID, apiKeyID, accountID, groupID, model, requestType, stream, billingType) +} + +// GetUpstreamEndpointStatsWithFilters returns upstream endpoint statistics with optional filters. +func (r *usageLogRepository) GetUpstreamEndpointStatsWithFilters(ctx context.Context, startTime, endTime time.Time, userID, apiKeyID, accountID, groupID int64, model string, requestType *int16, stream *bool, billingType *int8) ([]EndpointStat, error) { + return r.getEndpointStatsByColumnWithFilters(ctx, "upstream_endpoint", startTime, endTime, userID, apiKeyID, accountID, groupID, model, requestType, stream, billingType) +} + // GetAccountUsageStats returns comprehensive usage statistics for an account over a time range func (r *usageLogRepository) GetAccountUsageStats(ctx context.Context, accountID int64, startTime, endTime time.Time) (resp *AccountUsageStatsResponse, err error) { daysCount := int(endTime.Sub(startTime).Hours()/24) + 1 @@ -3254,11 +3462,23 @@ func (r *usageLogRepository) GetAccountUsageStats(ctx context.Context, accountID if err != nil { models = []ModelStat{} } + endpoints, endpointErr := r.GetEndpointStatsWithFilters(ctx, startTime, endTime, 0, 0, accountID, 0, "", nil, nil, nil) + if endpointErr != nil { + logger.LegacyPrintf("repository.usage_log", "GetEndpointStatsWithFilters failed in GetAccountUsageStats: %v", endpointErr) + endpoints = []EndpointStat{} + } + upstreamEndpoints, upstreamEndpointErr := r.GetUpstreamEndpointStatsWithFilters(ctx, startTime, endTime, 0, 0, accountID, 0, "", nil, nil, nil) + if upstreamEndpointErr != nil { + logger.LegacyPrintf("repository.usage_log", "GetUpstreamEndpointStatsWithFilters failed in GetAccountUsageStats: %v", upstreamEndpointErr) + upstreamEndpoints = []EndpointStat{} + } resp = &AccountUsageStatsResponse{ - History: history, - Summary: summary, - Models: models, + History: history, + Summary: summary, + Models: models, + Endpoints: endpoints, + UpstreamEndpoints: upstreamEndpoints, } return resp, nil } @@ -3541,6 +3761,8 @@ func scanUsageLog(scanner interface{ Scan(...any) error }) (*service.UsageLog, e mediaType sql.NullString serviceTier sql.NullString reasoningEffort sql.NullString + inboundEndpoint sql.NullString + upstreamEndpoint sql.NullString cacheTTLOverridden bool createdAt time.Time ) @@ -3581,6 +3803,8 @@ func scanUsageLog(scanner interface{ Scan(...any) error }) (*service.UsageLog, e &mediaType, &serviceTier, &reasoningEffort, + &inboundEndpoint, + &upstreamEndpoint, &cacheTTLOverridden, &createdAt, ); err != nil { @@ -3656,6 +3880,12 @@ func scanUsageLog(scanner interface{ Scan(...any) error }) (*service.UsageLog, e if reasoningEffort.Valid { log.ReasoningEffort = &reasoningEffort.String } + if inboundEndpoint.Valid { + log.InboundEndpoint = &inboundEndpoint.String + } + if upstreamEndpoint.Valid { + log.UpstreamEndpoint = &upstreamEndpoint.String + } return log, nil } diff --git a/backend/internal/repository/usage_log_repo_request_type_test.go b/backend/internal/repository/usage_log_repo_request_type_test.go index bcb23717..70b82bb0 100644 --- a/backend/internal/repository/usage_log_repo_request_type_test.go +++ b/backend/internal/repository/usage_log_repo_request_type_test.go @@ -73,6 +73,8 @@ func TestUsageLogRepositoryCreateSyncRequestTypeAndLegacyFields(t *testing.T) { sqlmock.AnyArg(), // media_type sqlmock.AnyArg(), // service_tier sqlmock.AnyArg(), // reasoning_effort + sqlmock.AnyArg(), // inbound_endpoint + sqlmock.AnyArg(), // upstream_endpoint log.CacheTTLOverridden, createdAt, ). @@ -141,6 +143,8 @@ func TestUsageLogRepositoryCreate_PersistsServiceTier(t *testing.T) { sqlmock.AnyArg(), serviceTier, sqlmock.AnyArg(), + sqlmock.AnyArg(), + sqlmock.AnyArg(), log.CacheTTLOverridden, createdAt, ). @@ -376,6 +380,8 @@ func TestScanUsageLogRequestTypeAndLegacyFallback(t *testing.T) { sql.NullString{}, sql.NullString{Valid: true, String: "priority"}, sql.NullString{}, + sql.NullString{}, + sql.NullString{}, false, now, }}) @@ -415,6 +421,8 @@ func TestScanUsageLogRequestTypeAndLegacyFallback(t *testing.T) { sql.NullString{}, sql.NullString{Valid: true, String: "flex"}, sql.NullString{}, + sql.NullString{}, + sql.NullString{}, false, now, }}) @@ -454,6 +462,8 @@ func TestScanUsageLogRequestTypeAndLegacyFallback(t *testing.T) { sql.NullString{}, sql.NullString{Valid: true, String: "priority"}, sql.NullString{}, + sql.NullString{}, + sql.NullString{}, false, now, }}) diff --git a/backend/internal/service/account_usage_service.go b/backend/internal/service/account_usage_service.go index d41e890a..b0a04d47 100644 --- a/backend/internal/service/account_usage_service.go +++ b/backend/internal/service/account_usage_service.go @@ -45,6 +45,8 @@ type UsageLogRepository interface { GetDashboardStats(ctx context.Context) (*usagestats.DashboardStats, error) GetUsageTrendWithFilters(ctx context.Context, startTime, endTime time.Time, granularity string, userID, apiKeyID, accountID, groupID int64, model string, requestType *int16, stream *bool, billingType *int8) ([]usagestats.TrendDataPoint, error) GetModelStatsWithFilters(ctx context.Context, startTime, endTime time.Time, userID, apiKeyID, accountID, groupID int64, requestType *int16, stream *bool, billingType *int8) ([]usagestats.ModelStat, error) + GetEndpointStatsWithFilters(ctx context.Context, startTime, endTime time.Time, userID, apiKeyID, accountID, groupID int64, model string, requestType *int16, stream *bool, billingType *int8) ([]usagestats.EndpointStat, error) + GetUpstreamEndpointStatsWithFilters(ctx context.Context, startTime, endTime time.Time, userID, apiKeyID, accountID, groupID int64, model string, requestType *int16, stream *bool, billingType *int8) ([]usagestats.EndpointStat, error) GetGroupStatsWithFilters(ctx context.Context, startTime, endTime time.Time, userID, apiKeyID, accountID, groupID int64, requestType *int16, stream *bool, billingType *int8) ([]usagestats.GroupStat, error) GetAPIKeyUsageTrend(ctx context.Context, startTime, endTime time.Time, granularity string, limit int) ([]usagestats.APIKeyUsageTrendPoint, error) GetUserUsageTrend(ctx context.Context, startTime, endTime time.Time, granularity string, limit int) ([]usagestats.UserUsageTrendPoint, error) diff --git a/backend/internal/service/openai_gateway_record_usage_test.go b/backend/internal/service/openai_gateway_record_usage_test.go index cd4d58fd..ada7d805 100644 --- a/backend/internal/service/openai_gateway_record_usage_test.go +++ b/backend/internal/service/openai_gateway_record_usage_test.go @@ -226,6 +226,41 @@ func TestOpenAIGatewayServiceRecordUsage_UsesUserSpecificGroupRate(t *testing.T) require.Equal(t, 1, userRepo.deductCalls) } +func TestOpenAIGatewayServiceRecordUsage_IncludesEndpointMetadata(t *testing.T) { + usageRepo := &openAIRecordUsageLogRepoStub{inserted: true} + userRepo := &openAIRecordUsageUserRepoStub{} + subRepo := &openAIRecordUsageSubRepoStub{} + rateRepo := &openAIUserGroupRateRepoStub{} + svc := newOpenAIRecordUsageServiceForTest(usageRepo, userRepo, subRepo, rateRepo) + + err := svc.RecordUsage(context.Background(), &OpenAIRecordUsageInput{ + Result: &OpenAIForwardResult{ + RequestID: "resp_endpoint_metadata", + Usage: OpenAIUsage{ + InputTokens: 8, + OutputTokens: 2, + }, + Model: "gpt-5.1", + Duration: time.Second, + }, + APIKey: &APIKey{ + ID: 1002, + Group: &Group{RateMultiplier: 1}, + }, + User: &User{ID: 2002}, + Account: &Account{ID: 3002}, + InboundEndpoint: " /v1/chat/completions ", + UpstreamEndpoint: " /v1/responses ", + }) + + require.NoError(t, err) + require.NotNil(t, usageRepo.lastLog) + require.NotNil(t, usageRepo.lastLog.InboundEndpoint) + require.Equal(t, "/v1/chat/completions", *usageRepo.lastLog.InboundEndpoint) + require.NotNil(t, usageRepo.lastLog.UpstreamEndpoint) + require.Equal(t, "/v1/responses", *usageRepo.lastLog.UpstreamEndpoint) +} + func TestOpenAIGatewayServiceRecordUsage_FallsBackToGroupDefaultRateOnResolverError(t *testing.T) { groupID := int64(12) groupRate := 1.6 diff --git a/backend/internal/service/openai_gateway_service.go b/backend/internal/service/openai_gateway_service.go index 313666dd..406344e3 100644 --- a/backend/internal/service/openai_gateway_service.go +++ b/backend/internal/service/openai_gateway_service.go @@ -4028,6 +4028,8 @@ type OpenAIRecordUsageInput struct { User *User Account *Account Subscription *UserSubscription + InboundEndpoint string + UpstreamEndpoint string UserAgent string // 请求的 User-Agent IPAddress string // 请求的客户端 IP 地址 RequestPayloadHash string @@ -4106,6 +4108,8 @@ func (s *OpenAIGatewayService) RecordUsage(ctx context.Context, input *OpenAIRec Model: billingModel, ServiceTier: result.ServiceTier, ReasoningEffort: result.ReasoningEffort, + InboundEndpoint: optionalTrimmedStringPtr(input.InboundEndpoint), + UpstreamEndpoint: optionalTrimmedStringPtr(input.UpstreamEndpoint), InputTokens: actualInputTokens, OutputTokens: result.Usage.OutputTokens, CacheCreationTokens: result.Usage.CacheCreationInputTokens, @@ -4125,7 +4129,6 @@ func (s *OpenAIGatewayService) RecordUsage(ctx context.Context, input *OpenAIRec FirstTokenMs: result.FirstTokenMs, CreatedAt: time.Now(), } - // 添加 UserAgent if input.UserAgent != "" { usageLog.UserAgent = &input.UserAgent @@ -4668,3 +4671,11 @@ func normalizeOpenAIReasoningEffort(raw string) string { return "" } } + +func optionalTrimmedStringPtr(raw string) *string { + trimmed := strings.TrimSpace(raw) + if trimmed == "" { + return nil + } + return &trimmed +} diff --git a/backend/internal/service/usage_log.go b/backend/internal/service/usage_log.go index a7464956..ef313a01 100644 --- a/backend/internal/service/usage_log.go +++ b/backend/internal/service/usage_log.go @@ -103,6 +103,10 @@ type UsageLog struct { // ReasoningEffort is the request's reasoning effort level (OpenAI Responses API), // e.g. "low" / "medium" / "high" / "xhigh". Nil means not provided / not applicable. ReasoningEffort *string + // InboundEndpoint is the client-facing API endpoint path, e.g. /v1/chat/completions. + InboundEndpoint *string + // UpstreamEndpoint is the normalized upstream endpoint path, e.g. /v1/responses. + UpstreamEndpoint *string GroupID *int64 SubscriptionID *int64 diff --git a/backend/migrations/074_add_usage_log_endpoints.sql b/backend/migrations/074_add_usage_log_endpoints.sql new file mode 100644 index 00000000..2a34e7c3 --- /dev/null +++ b/backend/migrations/074_add_usage_log_endpoints.sql @@ -0,0 +1,5 @@ +-- Add endpoint tracking fields to usage_logs. +-- inbound_endpoint: client-facing API route (e.g. /v1/chat/completions, /v1/messages, /v1/responses) +-- upstream_endpoint: normalized upstream route (e.g. /v1/responses) +ALTER TABLE usage_logs ADD COLUMN IF NOT EXISTS inbound_endpoint VARCHAR(128); +ALTER TABLE usage_logs ADD COLUMN IF NOT EXISTS upstream_endpoint VARCHAR(128); diff --git a/frontend/src/api/admin/usage.ts b/frontend/src/api/admin/usage.ts index 2d6212c5..bd7e3e57 100644 --- a/frontend/src/api/admin/usage.ts +++ b/frontend/src/api/admin/usage.ts @@ -5,6 +5,7 @@ import { apiClient } from '../client' import type { AdminUsageLog, UsageQueryParams, PaginatedResponse, UsageRequestType } from '@/types' +import type { EndpointStat } from '@/types' // ==================== Types ==================== @@ -18,6 +19,9 @@ export interface AdminUsageStatsResponse { total_actual_cost: number total_account_cost?: number average_duration_ms: number + endpoints?: EndpointStat[] + upstream_endpoints?: EndpointStat[] + endpoint_paths?: EndpointStat[] } export interface SimpleUser { diff --git a/frontend/src/components/account/AccountStatsModal.vue b/frontend/src/components/account/AccountStatsModal.vue index 7968fa8d..bb4edb97 100644 --- a/frontend/src/components/account/AccountStatsModal.vue +++ b/frontend/src/components/account/AccountStatsModal.vue @@ -446,6 +446,18 @@ + + + + @@ -489,6 +501,7 @@ import { Line } from 'vue-chartjs' import BaseDialog from '@/components/common/BaseDialog.vue' import LoadingSpinner from '@/components/common/LoadingSpinner.vue' import ModelDistributionChart from '@/components/charts/ModelDistributionChart.vue' +import EndpointDistributionChart from '@/components/charts/EndpointDistributionChart.vue' import Icon from '@/components/icons/Icon.vue' import { adminAPI } from '@/api/admin' import type { Account, AccountUsageStatsResponse } from '@/types' diff --git a/frontend/src/components/admin/account/AccountStatsModal.vue b/frontend/src/components/admin/account/AccountStatsModal.vue index 72a71d36..4dc84d5e 100644 --- a/frontend/src/components/admin/account/AccountStatsModal.vue +++ b/frontend/src/components/admin/account/AccountStatsModal.vue @@ -410,6 +410,18 @@ + + + + @@ -453,6 +465,7 @@ import { Line } from 'vue-chartjs' import BaseDialog from '@/components/common/BaseDialog.vue' import LoadingSpinner from '@/components/common/LoadingSpinner.vue' import ModelDistributionChart from '@/components/charts/ModelDistributionChart.vue' +import EndpointDistributionChart from '@/components/charts/EndpointDistributionChart.vue' import Icon from '@/components/icons/Icon.vue' import { adminAPI } from '@/api/admin' import type { Account, AccountUsageStatsResponse } from '@/types' diff --git a/frontend/src/components/admin/usage/UsageTable.vue b/frontend/src/components/admin/usage/UsageTable.vue index 72f7c010..aa6c2bbd 100644 --- a/frontend/src/components/admin/usage/UsageTable.vue +++ b/frontend/src/components/admin/usage/UsageTable.vue @@ -35,6 +35,19 @@ + +