feat(handler): 新增ops管理接口和路由

- 添加告警静默管理接口
- 扩展错误日志查询和操作接口
- 新增重试和解决状态相关端点
- 完善错误日志记录功能

backend/internal/handler/admin/ops_alerts_handler.go

@@ -7,8 +7,10 @@ import (
 	"net/http"
 	"strconv"
 	"strings"
+	"time"
 
 	"github.com/Wei-Shaw/sub2api/internal/pkg/response"
+	"github.com/Wei-Shaw/sub2api/internal/server/middleware"
 	"github.com/Wei-Shaw/sub2api/internal/service"
 	"github.com/gin-gonic/gin"
 	"github.com/gin-gonic/gin/binding"
@@ -372,8 +374,135 @@ func (h *OpsHandler) DeleteAlertRule(c *gin.Context) {
 	response.Success(c, gin.H{"deleted": true})
 }
 
+// GetAlertEvent returns a single ops alert event.
+// GET /api/v1/admin/ops/alert-events/:id
+func (h *OpsHandler) GetAlertEvent(c *gin.Context) {
+	if h.opsService == nil {
+		response.Error(c, http.StatusServiceUnavailable, "Ops service not available")
+		return
+	}
+	if err := h.opsService.RequireMonitoringEnabled(c.Request.Context()); err != nil {
+		response.ErrorFrom(c, err)
+		return
+	}
+
+	id, err := strconv.ParseInt(c.Param("id"), 10, 64)
+	if err != nil || id <= 0 {
+		response.BadRequest(c, "Invalid event ID")
+		return
+	}
+
+	ev, err := h.opsService.GetAlertEventByID(c.Request.Context(), id)
+	if err != nil {
+		response.ErrorFrom(c, err)
+		return
+	}
+	response.Success(c, ev)
+}
+
+// UpdateAlertEventStatus updates an ops alert event status.
+// PUT /api/v1/admin/ops/alert-events/:id/status
+func (h *OpsHandler) UpdateAlertEventStatus(c *gin.Context) {
+	if h.opsService == nil {
+		response.Error(c, http.StatusServiceUnavailable, "Ops service not available")
+		return
+	}
+	if err := h.opsService.RequireMonitoringEnabled(c.Request.Context()); err != nil {
+		response.ErrorFrom(c, err)
+		return
+	}
+
+	id, err := strconv.ParseInt(c.Param("id"), 10, 64)
+	if err != nil || id <= 0 {
+		response.BadRequest(c, "Invalid event ID")
+		return
+	}
+
+	var payload struct {
+		Status string `json:"status"`
+	}
+	if err := c.ShouldBindJSON(&payload); err != nil {
+		response.BadRequest(c, "Invalid request body")
+		return
+	}
+	payload.Status = strings.TrimSpace(payload.Status)
+	if payload.Status == "" {
+		response.BadRequest(c, "Invalid status")
+		return
+	}
+	if payload.Status != service.OpsAlertStatusResolved && payload.Status != service.OpsAlertStatusManualResolved {
+		response.BadRequest(c, "Invalid status")
+		return
+	}
+
+	var resolvedAt *time.Time
+	if payload.Status == service.OpsAlertStatusResolved || payload.Status == service.OpsAlertStatusManualResolved {
+		now := time.Now().UTC()
+		resolvedAt = &now
+	}
+	if err := h.opsService.UpdateAlertEventStatus(c.Request.Context(), id, payload.Status, resolvedAt); err != nil {
+		response.ErrorFrom(c, err)
+		return
+	}
+	response.Success(c, gin.H{"updated": true})
+}
+
 // ListAlertEvents lists recent ops alert events.
 // GET /api/v1/admin/ops/alert-events
+// CreateAlertSilence creates a scoped silence for ops alerts.
+// POST /api/v1/admin/ops/alert-silences
+func (h *OpsHandler) CreateAlertSilence(c *gin.Context) {
+	if h.opsService == nil {
+		response.Error(c, http.StatusServiceUnavailable, "Ops service not available")
+		return
+	}
+	if err := h.opsService.RequireMonitoringEnabled(c.Request.Context()); err != nil {
+		response.ErrorFrom(c, err)
+		return
+	}
+
+	var payload struct {
+		RuleID   int64   `json:"rule_id"`
+		Platform string  `json:"platform"`
+		GroupID  *int64  `json:"group_id"`
+		Region   *string `json:"region"`
+		Until    string  `json:"until"`
+		Reason   string  `json:"reason"`
+	}
+	if err := c.ShouldBindJSON(&payload); err != nil {
+		response.BadRequest(c, "Invalid request body")
+		return
+	}
+	until, err := time.Parse(time.RFC3339, strings.TrimSpace(payload.Until))
+	if err != nil {
+		response.BadRequest(c, "Invalid until")
+		return
+	}
+
+	createdBy := (*int64)(nil)
+	if subject, ok := middleware.GetAuthSubjectFromContext(c); ok {
+		uid := subject.UserID
+		createdBy = &uid
+	}
+
+	silence := &service.OpsAlertSilence{
+		RuleID:    payload.RuleID,
+		Platform:  strings.TrimSpace(payload.Platform),
+		GroupID:   payload.GroupID,
+		Region:    payload.Region,
+		Until:     until,
+		Reason:    strings.TrimSpace(payload.Reason),
+		CreatedBy: createdBy,
+	}
+
+	created, err := h.opsService.CreateAlertSilence(c.Request.Context(), silence)
+	if err != nil {
+		response.ErrorFrom(c, err)
+		return
+	}
+	response.Success(c, created)
+}
+
 func (h *OpsHandler) ListAlertEvents(c *gin.Context) {
 	if h.opsService == nil {
 		response.Error(c, http.StatusServiceUnavailable, "Ops service not available")
@@ -384,7 +513,7 @@ func (h *OpsHandler) ListAlertEvents(c *gin.Context) {
 		return
 	}
 
-	limit := 100
+	limit := 20
 	if raw := strings.TrimSpace(c.Query("limit")); raw != "" {
 		n, err := strconv.Atoi(raw)
 		if err != nil || n <= 0 {
@@ -400,6 +529,43 @@ func (h *OpsHandler) ListAlertEvents(c *gin.Context) {
 		Severity: strings.TrimSpace(c.Query("severity")),
 	}
 
+	if v := strings.TrimSpace(c.Query("email_sent")); v != "" {
+		vv := strings.ToLower(v)
+		switch vv {
+		case "true", "1":
+			b := true
+			filter.EmailSent = &b
+		case "false", "0":
+			b := false
+			filter.EmailSent = &b
+		default:
+			response.BadRequest(c, "Invalid email_sent")
+			return
+		}
+	}
+
+	// Cursor pagination
+	if rawTS := strings.TrimSpace(c.Query("before_fired_at")); rawTS != "" {
+		ts, err := time.Parse(time.RFC3339Nano, rawTS)
+		if err != nil {
+			if t2, err2 := time.Parse(time.RFC3339, rawTS); err2 == nil {
+				ts = t2
+			} else {
+				response.BadRequest(c, "Invalid before_fired_at")
+				return
+			}
+		}
+		filter.BeforeFiredAt = &ts
+	}
+	if rawID := strings.TrimSpace(c.Query("before_id")); rawID != "" {
+		id, err := strconv.ParseInt(rawID, 10, 64)
+		if err != nil || id <= 0 {
+			response.BadRequest(c, "Invalid before_id")
+			return
+		}
+		filter.BeforeID = &id
+	}
+
 	// Optional global filter support (platform/group/time range).
 	if platform := strings.TrimSpace(c.Query("platform")); platform != "" {
 		filter.Platform = platform

backend/internal/handler/admin/ops_handler.go

@@ -80,6 +80,25 @@ func (h *OpsHandler) GetErrorLogs(c *gin.Context) {
 	if phase := strings.TrimSpace(c.Query("phase")); phase != "" {
 		filter.Phase = phase
 	}
+	if owner := strings.TrimSpace(c.Query("error_owner")); owner != "" {
+		filter.Owner = owner
+	}
+	if source := strings.TrimSpace(c.Query("error_source")); source != "" {
+		filter.Source = source
+	}
+	if v := strings.TrimSpace(c.Query("resolved")); v != "" {
+		switch strings.ToLower(v) {
+		case "1", "true", "yes":
+			b := true
+			filter.Resolved = &b
+		case "0", "false", "no":
+			b := false
+			filter.Resolved = &b
+		default:
+			response.BadRequest(c, "Invalid resolved")
+			return
+		}
+	}
 	if q := strings.TrimSpace(c.Query("q")); q != "" {
 		filter.Query = q
 	}
@@ -242,6 +261,11 @@ func (h *OpsHandler) ListRequestDetails(c *gin.Context) {
 type opsRetryRequest struct {
 	Mode            string `json:"mode"`
 	PinnedAccountID *int64 `json:"pinned_account_id"`
+	Force           bool   `json:"force"`
+}
+
+type opsResolveRequest struct {
+	Resolved bool `json:"resolved"`
 }
 
 // RetryErrorRequest retries a failed request using stored request_body.
@@ -278,6 +302,8 @@ func (h *OpsHandler) RetryErrorRequest(c *gin.Context) {
 		req.Mode = service.OpsRetryModeClient
 	}
 
+	// Force flag is currently a UI-level acknowledgement. Server may still enforce safety constraints.
+	_ = req.Force
 	result, err := h.opsService.RetryError(c.Request.Context(), subject.UserID, id, req.Mode, req.PinnedAccountID)
 	if err != nil {
 		response.ErrorFrom(c, err)
@@ -287,6 +313,81 @@ func (h *OpsHandler) RetryErrorRequest(c *gin.Context) {
 	response.Success(c, result)
 }
 
+// ListRetryAttempts lists retry attempts for an error log.
+// GET /api/v1/admin/ops/errors/:id/retries
+func (h *OpsHandler) ListRetryAttempts(c *gin.Context) {
+	if h.opsService == nil {
+		response.Error(c, http.StatusServiceUnavailable, "Ops service not available")
+		return
+	}
+	if err := h.opsService.RequireMonitoringEnabled(c.Request.Context()); err != nil {
+		response.ErrorFrom(c, err)
+		return
+	}
+
+	idStr := strings.TrimSpace(c.Param("id"))
+	id, err := strconv.ParseInt(idStr, 10, 64)
+	if err != nil || id <= 0 {
+		response.BadRequest(c, "Invalid error id")
+		return
+	}
+
+	limit := 50
+	if v := strings.TrimSpace(c.Query("limit")); v != "" {
+		n, err := strconv.Atoi(v)
+		if err != nil || n <= 0 {
+			response.BadRequest(c, "Invalid limit")
+			return
+		}
+		limit = n
+	}
+
+	items, err := h.opsService.ListRetryAttemptsByErrorID(c.Request.Context(), id, limit)
+	if err != nil {
+		response.ErrorFrom(c, err)
+		return
+	}
+	response.Success(c, items)
+}
+
+// UpdateErrorResolution allows manual resolve/unresolve.
+// PUT /api/v1/admin/ops/errors/:id/resolve
+func (h *OpsHandler) UpdateErrorResolution(c *gin.Context) {
+	if h.opsService == nil {
+		response.Error(c, http.StatusServiceUnavailable, "Ops service not available")
+		return
+	}
+	if err := h.opsService.RequireMonitoringEnabled(c.Request.Context()); err != nil {
+		response.ErrorFrom(c, err)
+		return
+	}
+
+	subject, ok := middleware.GetAuthSubjectFromContext(c)
+	if !ok || subject.UserID <= 0 {
+		response.Error(c, http.StatusUnauthorized, "Unauthorized")
+		return
+	}
+
+	idStr := strings.TrimSpace(c.Param("id"))
+	id, err := strconv.ParseInt(idStr, 10, 64)
+	if err != nil || id <= 0 {
+		response.BadRequest(c, "Invalid error id")
+		return
+	}
+
+	var req opsResolveRequest
+	if err := c.ShouldBindJSON(&req); err != nil {
+		response.BadRequest(c, "Invalid request: "+err.Error())
+		return
+	}
+	uid := subject.UserID
+	if err := h.opsService.UpdateErrorResolution(c.Request.Context(), id, req.Resolved, &uid, nil); err != nil {
+		response.ErrorFrom(c, err)
+		return
+	}
+	response.Success(c, gin.H{"ok": true})
+}
+
 func parseOpsTimeRange(c *gin.Context, defaultRange string) (time.Time, time.Time, error) {
 	startStr := strings.TrimSpace(c.Query("start_time"))
 	endStr := strings.TrimSpace(c.Query("end_time"))
@@ -358,6 +459,10 @@ func parseOpsDuration(v string) (time.Duration, bool) {
 		return 6 * time.Hour, true
 	case "24h":
 		return 24 * time.Hour, true
+	case "7d":
+		return 7 * 24 * time.Hour, true
+	case "30d":
+		return 30 * 24 * time.Hour, true
 	default:
 		return 0, false
 	}

backend/internal/handler/ops_error_logger.go

@@ -832,28 +832,30 @@ func normalizeOpsErrorType(errType string, code string) string {
 
 func classifyOpsPhase(errType, message, code string) string {
 	msg := strings.ToLower(message)
+	// Standardized phases: request|auth|routing|upstream|network|internal
+	// Map billing/concurrency/response => request; scheduling => routing.
 	switch strings.TrimSpace(code) {
 	case "INSUFFICIENT_BALANCE", "USAGE_LIMIT_EXCEEDED", "SUBSCRIPTION_NOT_FOUND", "SUBSCRIPTION_INVALID":
-		return "billing"
+		return "request"
 	}
 
 	switch errType {
 	case "authentication_error":
 		return "auth"
 	case "billing_error", "subscription_error":
-		return "billing"
+		return "request"
 	case "rate_limit_error":
 		if strings.Contains(msg, "concurrency") || strings.Contains(msg, "pending") || strings.Contains(msg, "queue") {
-			return "concurrency"
+			return "request"
 		}
 		return "upstream"
 	case "invalid_request_error":
-		return "response"
+		return "request"
 	case "upstream_error", "overloaded_error":
 		return "upstream"
 	case "api_error":
 		if strings.Contains(msg, "no available accounts") {
-			return "scheduling"
+			return "routing"
 		}
 		return "internal"
 	default:
@@ -914,34 +916,38 @@ func classifyOpsIsBusinessLimited(errType, phase, code string, status int, messa
 }
 
 func classifyOpsErrorOwner(phase string, message string) string {
+	// Standardized owners: client|provider|platform
 	switch phase {
 	case "upstream", "network":
 		return "provider"
-	case "billing", "concurrency", "auth", "response":
+	case "request", "auth":
 		return "client"
+	case "routing", "internal":
+		return "platform"
 	default:
 		if strings.Contains(strings.ToLower(message), "upstream") {
 			return "provider"
 		}
-		return "sub2api"
+		return "platform"
 	}
 }
 
 func classifyOpsErrorSource(phase string, message string) string {
+	// Standardized sources: client_request|upstream_http|gateway
 	switch phase {
 	case "upstream":
 		return "upstream_http"
 	case "network":
-		return "upstream_network"
-	case "billing":
-		return "billing"
-	case "concurrency":
-		return "concurrency"
+		return "gateway"
+	case "request", "auth":
+		return "client_request"
+	case "routing", "internal":
+		return "gateway"
 	default:
 		if strings.Contains(strings.ToLower(message), "upstream") {
 			return "upstream_http"
 		}
-		return "internal"
+		return "gateway"
 	}
 }
 

backend/internal/server/routes/admin.go

@@ -81,6 +81,9 @@ func registerOpsRoutes(admin *gin.RouterGroup, h *handler.Handlers) {
 		ops.PUT("/alert-rules/:id", h.Admin.Ops.UpdateAlertRule)
 		ops.DELETE("/alert-rules/:id", h.Admin.Ops.DeleteAlertRule)
 		ops.GET("/alert-events", h.Admin.Ops.ListAlertEvents)
+		ops.GET("/alert-events/:id", h.Admin.Ops.GetAlertEvent)
+		ops.PUT("/alert-events/:id/status", h.Admin.Ops.UpdateAlertEventStatus)
+		ops.POST("/alert-silences", h.Admin.Ops.CreateAlertSilence)
 
 		// Email notification config (DB-backed)
 		ops.GET("/email-notification/config", h.Admin.Ops.GetEmailNotificationConfig)
@@ -113,7 +116,9 @@ func registerOpsRoutes(admin *gin.RouterGroup, h *handler.Handlers) {
 		// Error logs (MVP-1)
 		ops.GET("/errors", h.Admin.Ops.GetErrorLogs)
 		ops.GET("/errors/:id", h.Admin.Ops.GetErrorLogByID)
+		ops.GET("/errors/:id/retries", h.Admin.Ops.ListRetryAttempts)
 		ops.POST("/errors/:id/retry", h.Admin.Ops.RetryErrorRequest)
+		ops.PUT("/errors/:id/resolve", h.Admin.Ops.UpdateErrorResolution)
 
 		// Request drilldown (success + error)
 		ops.GET("/requests", h.Admin.Ops.ListRequestDetails)