feishu config add defaultUserGroup (#3076)

Co-authored-by: jie210 <huangjie@flashcat.com>

alert/dispatch/dispatch.go

@@ -253,7 +253,7 @@ func HandleEventPipeline(pipelineConfigs []models.PipelineConfig, eventOrigin, e
 		// 统一使用工作流引擎执行（兼容线性模式和工作流模式）
 		triggerCtx := &models.WorkflowTriggerContext{
 			Mode:      models.TriggerModeEvent,
-			TriggerBy: from,
+			TriggerBy: from + "_" + strconv.FormatInt(id, 10),
 		}
 
 		resultEvent, result, err := workflowEngine.Execute(eventPipeline, event, triggerCtx)

alert/eval/eval.go

@@ -844,7 +844,7 @@ func (arw *AlertRuleWorker) GetHostAnomalyPoint(ruleConfig string) ([]models.Ano
 				}
 				m["ident"] = target.Ident
 
-				lst = append(lst, models.NewAnomalyPoint(trigger.Type, m, now, float64(now-target.UpdateAt), trigger.Severity))
+				lst = append(lst, models.NewAnomalyPoint(trigger.Type, m, now, float64(now-target.BeatTime), trigger.Severity))
 			}
 		case "offset":
 			idents, exists := arw.Processor.TargetsOfAlertRuleCache.Get(arw.Processor.EngineName, arw.Rule.Id)
@@ -873,7 +873,7 @@ func (arw *AlertRuleWorker) GetHostAnomalyPoint(ruleConfig string) ([]models.Ano
 					continue
 				}
 				if target, exists := targetMap[ident]; exists {
-					if now-target.UpdateAt > 120 {
+					if now-target.BeatTime > 120 {
 						// means this target is not a active host, do not check offset
 						continue
 					}
@@ -1614,11 +1614,15 @@ func (arw *AlertRuleWorker) GetAnomalyPoint(rule *models.AlertRule, dsId int64)
 							continue
 						}
 
-						switch v.(type) {
-						case float64:
-							values += fmt.Sprintf("%s:%.3f ", k, v)
-						case string:
-							values += fmt.Sprintf("%s:%s ", k, v)
+						if u, exists := valuesUnitMap[k]; exists { // 配置了单位，优先用配置了单位的值
+							values += fmt.Sprintf("%s:%s ", k, u.Text)
+						} else {
+							switch v.(type) {
+							case float64:
+								values += fmt.Sprintf("%s:%.3f ", k, v)
+							case string:
+								values += fmt.Sprintf("%s:%s ", k, v)
+							}
 						}
 					}
 

alert/pipeline/engine/engine.go

@@ -60,11 +60,11 @@ func (e *WorkflowEngine) Execute(pipeline *models.EventPipeline, event *models.A
 }
 
 func (e *WorkflowEngine) initWorkflowContext(pipeline *models.EventPipeline, event *models.AlertCurEvent, triggerCtx *models.WorkflowTriggerContext) *models.WorkflowContext {
-	// 合并环境变量
-	env := pipeline.GetEnvMap()
-	if triggerCtx != nil && triggerCtx.EnvOverrides != nil {
-		for k, v := range triggerCtx.EnvOverrides {
-			env[k] = v
+	// 合并输入参数
+	inputs := pipeline.GetInputsMap()
+	if triggerCtx != nil && triggerCtx.InputsOverrides != nil {
+		for k, v := range triggerCtx.InputsOverrides {
+			inputs[k] = v
 		}
 	}
 
@@ -84,7 +84,7 @@ func (e *WorkflowEngine) initWorkflowContext(pipeline *models.EventPipeline, eve
 
 	return &models.WorkflowContext{
 		Event:    event,
-		Env:      env,
+		Inputs:   inputs,
 		Vars:     make(map[string]interface{}), // 初始化空的 Vars，供节点间传递数据
 		Metadata: metadata,
 		Stream:   stream,
@@ -182,7 +182,6 @@ func (e *WorkflowEngine) executeDAG(nodeMap map[string]*models.WorkflowNode, con
 				result.Status = models.ExecutionStatusFailed
 				result.ErrorNode = nodeID
 				result.Message = fmt.Sprintf("node %s failed: %s", node.Name, nodeResult.Error)
-				return result
 			}
 		}
 
@@ -371,10 +370,8 @@ func (e *WorkflowEngine) saveExecutionRecord(pipeline *models.EventPipeline, wfC
 		logger.Errorf("workflow: failed to set node results: pipeline_id=%d, error=%v", pipeline.ID, err)
 	}
 
-	secretKeys := pipeline.GetSecretKeys()
-	sanitizedEnv := wfCtx.SanitizedEnv(secretKeys)
-	if err := execution.SetEnvSnapshot(sanitizedEnv); err != nil {
-		logger.Errorf("workflow: failed to set env snapshot: pipeline_id=%d, error=%v", pipeline.ID, err)
+	if err := execution.SetInputsSnapshot(wfCtx.Inputs); err != nil {
+		logger.Errorf("workflow: failed to set inputs snapshot: pipeline_id=%d, error=%v", pipeline.ID, err)
 	}
 
 	if err := models.CreateEventPipelineExecution(e.ctx, execution); err != nil {

alert/pipeline/processor/aisummary/ai_summary.go

@@ -114,7 +114,7 @@ func (c *AISummaryConfig) initHTTPClient() error {
 func (c *AISummaryConfig) prepareEventInfo(wfCtx *models.WorkflowContext) (string, error) {
 	var defs = []string{
 		"{{$event := .Event}}",
-		"{{$env := .Env}}",
+		"{{$inputs := .Inputs}}",
 	}
 
 	text := strings.Join(append(defs, c.PromptTemplate), "")

alert/pipeline/processor/aisummary/ai_summary_test.go

@@ -44,8 +44,8 @@ func TestAISummaryConfig_Process(t *testing.T) {
 
 	// 创建 WorkflowContext
 	wfCtx := &models.WorkflowContext{
-		Event: event,
-		Env:   map[string]string{},
+		Event:  event,
+		Inputs: map[string]string{},
 	}
 
 	// 测试模板处理

alert/pipeline/processor/callback/callback.go

@@ -11,6 +11,7 @@ import (
 	"time"
 
 	"github.com/ccfos/nightingale/v6/alert/pipeline/processor/common"
+	"github.com/ccfos/nightingale/v6/alert/pipeline/processor/utils"
 	"github.com/ccfos/nightingale/v6/models"
 	"github.com/ccfos/nightingale/v6/pkg/ctx"
 	"github.com/toolkits/pkg/logger"
@@ -71,12 +72,17 @@ func (c *CallbackConfig) Process(ctx *ctx.Context, wfCtx *models.WorkflowContext
 		headers[k] = v
 	}
 
+	url, err := utils.TplRender(wfCtx, c.URL)
+	if err != nil {
+		return wfCtx, "", fmt.Errorf("failed to render url template: %v processor: %v", err, c)
+	}
+
 	body, err := json.Marshal(event)
 	if err != nil {
 		return wfCtx, "", fmt.Errorf("failed to marshal event: %v processor: %v", err, c)
 	}
 
-	req, err := http.NewRequest("POST", c.URL, strings.NewReader(string(body)))
+	req, err := http.NewRequest("POST", url, strings.NewReader(string(body)))
 	if err != nil {
 		return wfCtx, "", fmt.Errorf("failed to create request: %v processor: %v", err, c)
 	}

alert/pipeline/processor/eventdrop/event_drop.go

@@ -36,7 +36,7 @@ func (c *EventDropConfig) Process(ctx *ctx.Context, wfCtx *models.WorkflowContex
 		"{{ $event := .Event }}",
 		"{{ $labels := .Event.TagsMap }}",
 		"{{ $value := .Event.TriggerValue }}",
-		"{{ $env := .Env }}",
+		"{{ $inputs := .Inputs }}",
 	}
 
 	text := strings.Join(append(defs, c.Content), "")

alert/pipeline/processor/logic/if.go

@@ -148,7 +148,7 @@ func (c *IfConfig) evaluateExpressionCondition(wfCtx *models.WorkflowContext) (b
 		"{{ $event := .Event }}",
 		"{{ $labels := .Event.TagsMap }}",
 		"{{ $value := .Event.TriggerValue }}",
-		"{{ $env := .Env }}",
+		"{{ $inputs := .Inputs }}",
 	}
 
 	text := strings.Join(append(defs, c.Condition), "")

alert/pipeline/processor/logic/switch.go

@@ -175,7 +175,7 @@ func (c *SwitchConfig) evaluateExpressionCondition(condition string, wfCtx *mode
 		"{{ $event := .Event }}",
 		"{{ $labels := .Event.TagsMap }}",
 		"{{ $value := .Event.TriggerValue }}",
-		"{{ $env := .Env }}",
+		"{{ $inputs := .Inputs }}",
 	}
 
 	text := strings.Join(append(defs, condition), "")

alert/pipeline/processor/utils/utils.go

@@ -0,0 +1,32 @@
+package utils
+
+import (
+	"bytes"
+	"fmt"
+	"strings"
+	"text/template"
+
+	"github.com/ccfos/nightingale/v6/models"
+	"github.com/ccfos/nightingale/v6/pkg/tplx"
+)
+
+func TplRender(wfCtx *models.WorkflowContext, content string) (string, error) {
+	var defs = []string{
+		"{{ $event := .Event }}",
+		"{{ $labels := .Event.TagsMap }}",
+		"{{ $value := .Event.TriggerValue }}",
+		"{{ $inputs := .Inputs }}",
+	}
+	text := strings.Join(append(defs, content), "")
+	tpl, err := template.New("tpl").Funcs(tplx.TemplateFuncMap).Parse(text)
+	if err != nil {
+		return "", fmt.Errorf("failed to parse template: %v", err)
+	}
+
+	var body bytes.Buffer
+	if err = tpl.Execute(&body, wfCtx); err != nil {
+		return "", fmt.Errorf("failed to execute template: %v", err)
+	}
+
+	return strings.TrimSpace(body.String()), nil
+}

center/integration/init.go

@@ -271,10 +271,8 @@ func Init(ctx *ctx.Context, builtinIntegrationsDir string) {
 				}
 
 				for _, metric := range metrics {
-					if metric.UUID == 0 {
-						time.Sleep(time.Microsecond)
-						metric.UUID = time.Now().UnixMicro()
-					}
+					time.Sleep(time.Microsecond)
+					metric.UUID = time.Now().UnixMicro()
 					metric.ID = metric.UUID
 					metric.CreatedBy = SYSTEM
 					metric.UpdatedBy = SYSTEM

center/metas/metas.go

@@ -118,7 +118,7 @@ func (s *Set) updateTargets(m map[string]models.HostMeta) error {
 	}
 
 	start := time.Now()
-	err := storage.MSet(context.Background(), s.redis, newMap)
+	err := storage.MSet(context.Background(), s.redis, newMap, 7*24*time.Hour)
 	if err != nil {
 		cstats.RedisOperationLatency.WithLabelValues("mset_target_meta", "fail").Observe(time.Since(start).Seconds())
 		return err
@@ -127,7 +127,7 @@ func (s *Set) updateTargets(m map[string]models.HostMeta) error {
 	}
 
 	if len(extendMap) > 0 {
-		err = storage.MSet(context.Background(), s.redis, extendMap)
+		err = storage.MSet(context.Background(), s.redis, extendMap, 7*24*time.Hour)
 		if err != nil {
 			cstats.RedisOperationLatency.WithLabelValues("mset_target_extend", "fail").Observe(time.Since(start).Seconds())
 			return err

center/router/router.go

@@ -251,10 +251,12 @@ func (rt *Router) Config(r *gin.Engine) {
 		pages.GET("/auth/redirect/cas", rt.loginRedirectCas)
 		pages.GET("/auth/redirect/oauth", rt.loginRedirectOAuth)
 		pages.GET("/auth/redirect/dingtalk", rt.loginRedirectDingTalk)
+		pages.GET("/auth/redirect/feishu", rt.loginRedirectFeiShu)
 		pages.GET("/auth/callback", rt.loginCallback)
 		pages.GET("/auth/callback/cas", rt.loginCallbackCas)
 		pages.GET("/auth/callback/oauth", rt.loginCallbackOAuth)
 		pages.GET("/auth/callback/dingtalk", rt.loginCallbackDingTalk)
+		pages.GET("/auth/callback/feishu", rt.loginCallbackFeiShu)
 		pages.GET("/auth/perms", rt.allPerms)
 
 		pages.GET("/metrics/desc", rt.metricsDescGetFile)
@@ -389,8 +391,8 @@ func (rt *Router) Config(r *gin.Engine) {
 		pages.GET("/busi-group/:id/recording-rules", rt.auth(), rt.user(), rt.perm("/recording-rules"), rt.recordingRuleGets)
 		pages.POST("/busi-group/:id/recording-rules", rt.auth(), rt.user(), rt.perm("/recording-rules/add"), rt.bgrw(), rt.recordingRuleAddByFE)
 		pages.DELETE("/busi-group/:id/recording-rules", rt.auth(), rt.user(), rt.perm("/recording-rules/del"), rt.bgrw(), rt.recordingRuleDel)
-		pages.PUT("/busi-group/:id/recording-rule/:rrid", rt.auth(), rt.user(), rt.perm("/recording-rules/put"), rt.bgrw(), rt.recordingRulePutByFE)
 		pages.GET("/recording-rule/:rrid", rt.auth(), rt.user(), rt.perm("/recording-rules"), rt.recordingRuleGet)
+		pages.PUT("/recording-rule/:rrid", rt.auth(), rt.user(), rt.perm("/recording-rules"), rt.recordingRulePutByFE)
 		pages.PUT("/busi-group/:id/recording-rules/fields", rt.auth(), rt.user(), rt.perm("/recording-rules/put"), rt.recordingRulePutFields)
 
 		pages.GET("/busi-groups/alert-mutes", rt.auth(), rt.user(), rt.perm("/alert-mutes"), rt.alertMuteGetsByGids)
@@ -705,6 +707,7 @@ func (rt *Router) Config(r *gin.Engine) {
 			service.GET("/event-pipelines", rt.eventPipelinesListByService)
 			service.POST("/event-pipeline/:id/trigger", rt.triggerEventPipelineByService)
 			service.POST("/event-pipeline/:id/stream", rt.streamEventPipelineByService)
+			service.POST("/event-pipeline-execution", rt.eventPipelineExecutionAdd)
 
 			// 手机号加密存储配置接口
 			service.POST("/users/phone/encrypt", rt.usersPhoneEncrypt)

center/router/router_datasource.go

@@ -276,7 +276,7 @@ func (rt *Router) datasourceUpsert(c *gin.Context) {
 		}
 		err = req.Add(rt.Ctx)
 	} else {
-		err = req.Update(rt.Ctx, "name", "identifier", "description", "cluster_name", "settings", "http", "auth", "updated_by", "updated_at", "is_default")
+		err = req.Update(rt.Ctx, "name", "identifier", "description", "cluster_name", "settings", "http", "auth", "updated_by", "updated_at", "is_default", "weight")
 	}
 
 	Render(c, nil, err)

center/router/router_event_pipeline.go

@@ -164,7 +164,7 @@ func (rt *Router) tryRunEventPipeline(c *gin.Context) {
 	var f struct {
 		EventId        int64                `json:"event_id"`
 		PipelineConfig models.EventPipeline `json:"pipeline_config"`
-		EnvVariables   map[string]string    `json:"env_variables,omitempty"`
+		InputVariables map[string]string    `json:"input_variables,omitempty"`
 	}
 
 	ginx.BindJSON(c, &f)
@@ -182,9 +182,9 @@ func (rt *Router) tryRunEventPipeline(c *gin.Context) {
 	workflowEngine := engine.NewWorkflowEngine(rt.Ctx)
 
 	triggerCtx := &models.WorkflowTriggerContext{
-		Mode:         models.TriggerModeAPI,
-		TriggerBy:    me.Username,
-		EnvOverrides: f.EnvVariables,
+		Mode:            models.TriggerModeAPI,
+		TriggerBy:       me.Username,
+		InputsOverrides: f.InputVariables,
 	}
 
 	resultEvent, result, err := workflowEngine.Execute(&f.PipelineConfig, event, triggerCtx)
@@ -302,8 +302,8 @@ func (rt *Router) eventPipelinesListByService(c *gin.Context) {
 type EventPipelineRequest struct {
 	// 事件数据（可选，如果不传则使用空事件）
 	Event *models.AlertCurEvent `json:"event,omitempty"`
-	// 环境变量覆盖
-	EnvOverrides map[string]string `json:"env_overrides,omitempty"`
+	// 输入参数覆盖
+	InputsOverrides map[string]string `json:"inputs_overrides,omitempty"`
 
 	Username string `json:"username,omitempty"`
 }
@@ -321,20 +321,15 @@ func (rt *Router) executePipelineTrigger(pipeline *models.EventPipeline, req *Ev
 		}
 	}
 
-	// 校验必填环境变量
-	if err := pipeline.ValidateEnvVariables(req.EnvOverrides); err != nil {
-		return "", fmt.Errorf("env validation failed: %v", err)
-	}
-
 	// 生成执行ID
 	executionID := uuid.New().String()
 
 	// 创建触发上下文
 	triggerCtx := &models.WorkflowTriggerContext{
-		Mode:         models.TriggerModeAPI,
-		TriggerBy:    triggerBy,
-		EnvOverrides: req.EnvOverrides,
-		RequestID:    executionID,
+		Mode:            models.TriggerModeAPI,
+		TriggerBy:       triggerBy,
+		InputsOverrides: req.InputsOverrides,
+		RequestID:       executionID,
 	}
 
 	// 异步执行工作流
@@ -401,6 +396,7 @@ func (rt *Router) triggerEventPipelineByAPI(c *gin.Context) {
 }
 
 func (rt *Router) listAllEventPipelineExecutions(c *gin.Context) {
+	pipelineId := ginx.QueryInt64(c, "pipeline_id", 0)
 	pipelineName := ginx.QueryStr(c, "pipeline_name", "")
 	mode := ginx.QueryStr(c, "mode", "")
 	status := ginx.QueryStr(c, "status", "")
@@ -414,7 +410,7 @@ func (rt *Router) listAllEventPipelineExecutions(c *gin.Context) {
 		offset = 1
 	}
 
-	executions, total, err := models.ListAllEventPipelineExecutions(rt.Ctx, pipelineName, mode, status, limit, (offset-1)*limit)
+	executions, total, err := models.ListAllEventPipelineExecutions(rt.Ctx, pipelineId, pipelineName, mode, status, limit, (offset-1)*limit)
 	ginx.Dangerous(err)
 
 	ginx.NewRender(c).Data(gin.H{
@@ -509,11 +505,11 @@ func (rt *Router) streamEventPipeline(c *gin.Context) {
 	}
 
 	triggerCtx := &models.WorkflowTriggerContext{
-		Mode:         models.TriggerModeAPI,
-		TriggerBy:    me.Username,
-		EnvOverrides: f.EnvOverrides,
-		RequestID:    uuid.New().String(),
-		Stream:       true, // 流式端点强制启用流式输出
+		Mode:            models.TriggerModeAPI,
+		TriggerBy:       me.Username,
+		InputsOverrides: f.InputsOverrides,
+		RequestID:       uuid.New().String(),
+		Stream:          true, // 流式端点强制启用流式输出
 	}
 
 	workflowEngine := engine.NewWorkflowEngine(rt.Ctx)
@@ -604,11 +600,11 @@ func (rt *Router) streamEventPipelineByService(c *gin.Context) {
 	}
 
 	triggerCtx := &models.WorkflowTriggerContext{
-		Mode:         models.TriggerModeAPI,
-		TriggerBy:    f.Username,
-		EnvOverrides: f.EnvOverrides,
-		RequestID:    uuid.New().String(),
-		Stream:       true, // 流式端点强制启用流式输出
+		Mode:            models.TriggerModeAPI,
+		TriggerBy:       f.Username,
+		InputsOverrides: f.InputsOverrides,
+		RequestID:       uuid.New().String(),
+		Stream:          true, // 流式端点强制启用流式输出
 	}
 
 	workflowEngine := engine.NewWorkflowEngine(rt.Ctx)
@@ -625,3 +621,18 @@ func (rt *Router) streamEventPipelineByService(c *gin.Context) {
 
 	ginx.NewRender(c).Data(result, nil)
 }
+
+// eventPipelineExecutionAdd 接收 edge 节点同步的 Pipeline 执行记录
+func (rt *Router) eventPipelineExecutionAdd(c *gin.Context) {
+	var execution models.EventPipelineExecution
+	ginx.BindJSON(c, &execution)
+
+	if execution.ID == "" {
+		ginx.Bomb(http.StatusBadRequest, "id is required")
+	}
+	if execution.PipelineID <= 0 {
+		ginx.Bomb(http.StatusBadRequest, "pipeline_id is required")
+	}
+
+	ginx.NewRender(c).Message(models.DB(rt.Ctx).Create(&execution).Error)
+}

center/router/router_login.go

@@ -12,6 +12,7 @@ import (
 	"github.com/ccfos/nightingale/v6/models"
 	"github.com/ccfos/nightingale/v6/pkg/cas"
 	"github.com/ccfos/nightingale/v6/pkg/dingtalk"
+	"github.com/ccfos/nightingale/v6/pkg/feishu"
 	"github.com/ccfos/nightingale/v6/pkg/ldapx"
 	"github.com/ccfos/nightingale/v6/pkg/oauth2x"
 	"github.com/ccfos/nightingale/v6/pkg/oidcx"
@@ -519,6 +520,92 @@ func (rt *Router) loginCallbackDingTalk(c *gin.Context) {
 
 }
 
+func (rt *Router) loginRedirectFeiShu(c *gin.Context) {
+	redirect := ginx.QueryStr(c, "redirect", "/")
+
+	v, exists := c.Get("userid")
+	if exists {
+		userid := v.(int64)
+		user, err := models.UserGetById(rt.Ctx, userid)
+		ginx.Dangerous(err)
+		if user == nil {
+			ginx.Bomb(200, "user not found")
+		}
+
+		if user.Username != "" { // already login
+			ginx.NewRender(c).Data(redirect, nil)
+			return
+		}
+	}
+
+	if rt.Sso.FeiShu == nil || !rt.Sso.FeiShu.Enable {
+		ginx.NewRender(c).Data("", nil)
+		return
+	}
+
+	redirect, err := rt.Sso.FeiShu.Authorize(rt.Redis, redirect)
+	ginx.Dangerous(err)
+
+	ginx.NewRender(c).Data(redirect, err)
+}
+
+func (rt *Router) loginCallbackFeiShu(c *gin.Context) {
+	code := ginx.QueryStr(c, "code", "")
+	state := ginx.QueryStr(c, "state", "")
+
+	ret, err := rt.Sso.FeiShu.Callback(rt.Redis, c.Request.Context(), code, state)
+	if err != nil {
+		logger.Errorf("sso_callback FeiShu fail. code:%s, state:%s, get ret: %+v. error: %v", code, state, ret, err)
+		ginx.NewRender(c).Data(CallbackOutput{}, err)
+		return
+	}
+
+	user, err := models.UserGet(rt.Ctx, "username=?", ret.Username)
+	ginx.Dangerous(err)
+
+	if user != nil {
+		if rt.Sso.FeiShu != nil && rt.Sso.FeiShu.FeiShuConfig != nil && rt.Sso.FeiShu.FeiShuConfig.CoverAttributes {
+			updatedFields := user.UpdateSsoFields(feishu.SsoTypeName, ret.Nickname, ret.Phone, ret.Email)
+			ginx.Dangerous(user.Update(rt.Ctx, "update_at", updatedFields...))
+		}
+	} else {
+		user = new(models.User)
+		defaultRoles := []string{}
+		defaultUserGroups := []int64{}
+		if rt.Sso.FeiShu != nil && rt.Sso.FeiShu.FeiShuConfig != nil {
+			defaultRoles = rt.Sso.FeiShu.FeiShuConfig.DefaultRoles
+			defaultUserGroups = rt.Sso.FeiShu.FeiShuConfig.DefaultUserGroups
+		}
+
+		user.FullSsoFields(feishu.SsoTypeName, ret.Username, ret.Nickname, ret.Phone, ret.Email, defaultRoles)
+		ginx.Dangerous(user.Add(rt.Ctx))
+
+		if len(defaultUserGroups) > 0 {
+			ginx.Dangerous(user.UpdateUserGroup(rt.Ctx, defaultUserGroups))
+		}
+
+	}
+
+	// set user login state
+	userIdentity := fmt.Sprintf("%d-%s", user.Id, user.Username)
+	ts, err := rt.createTokens(rt.HTTP.JWTAuth.SigningKey, userIdentity)
+	ginx.Dangerous(err)
+	ginx.Dangerous(rt.createAuth(c.Request.Context(), userIdentity, ts))
+
+	redirect := "/"
+	if ret.Redirect != "/login" {
+		redirect = ret.Redirect
+	}
+
+	ginx.NewRender(c).Data(CallbackOutput{
+		Redirect:     redirect,
+		User:         user,
+		AccessToken:  ts.AccessToken,
+		RefreshToken: ts.RefreshToken,
+	}, nil)
+
+}
+
 func (rt *Router) loginCallbackOAuth(c *gin.Context) {
 	code := ginx.QueryStr(c, "code", "")
 	state := ginx.QueryStr(c, "state", "")
@@ -569,10 +656,11 @@ type SsoConfigOutput struct {
 	CasDisplayName      string `json:"casDisplayName"`
 	OauthDisplayName    string `json:"oauthDisplayName"`
 	DingTalkDisplayName string `json:"dingTalkDisplayName"`
+	FeiShuDisplayName   string `json:"feishuDisplayName"`
 }
 
 func (rt *Router) ssoConfigNameGet(c *gin.Context) {
-	var oidcDisplayName, casDisplayName, oauthDisplayName, dingTalkDisplayName string
+	var oidcDisplayName, casDisplayName, oauthDisplayName, dingTalkDisplayName, feiShuDisplayName string
 	if rt.Sso.OIDC != nil {
 		oidcDisplayName = rt.Sso.OIDC.GetDisplayName()
 	}
@@ -589,11 +677,16 @@ func (rt *Router) ssoConfigNameGet(c *gin.Context) {
 		dingTalkDisplayName = rt.Sso.DingTalk.GetDisplayName()
 	}
 
+	if rt.Sso.FeiShu != nil {
+		feiShuDisplayName = rt.Sso.FeiShu.GetDisplayName()
+	}
+
 	ginx.NewRender(c).Data(SsoConfigOutput{
 		OidcDisplayName:     oidcDisplayName,
 		CasDisplayName:      casDisplayName,
 		OauthDisplayName:    oauthDisplayName,
 		DingTalkDisplayName: dingTalkDisplayName,
+		FeiShuDisplayName:   feiShuDisplayName,
 	}, nil)
 }
 
@@ -608,6 +701,7 @@ func (rt *Router) ssoConfigGets(c *gin.Context) {
 
 	// TODO: dingTalkExist 为了兼容当前前端配置, 后期单点登陆统一调整后不在预先设置默认内容
 	dingTalkExist := false
+	feiShuExist := false
 	for _, config := range lst {
 		var ssoReqConfig models.SsoConfig
 		ssoReqConfig.Id = config.Id
@@ -618,6 +712,10 @@ func (rt *Router) ssoConfigGets(c *gin.Context) {
 			dingTalkExist = true
 			err := json.Unmarshal([]byte(config.Content), &ssoReqConfig.SettingJson)
 			ginx.Dangerous(err)
+		case feishu.SsoTypeName:
+			feiShuExist = true
+			err := json.Unmarshal([]byte(config.Content), &ssoReqConfig.SettingJson)
+			ginx.Dangerous(err)
 		default:
 			ssoReqConfig.Content = config.Content
 		}
@@ -630,6 +728,11 @@ func (rt *Router) ssoConfigGets(c *gin.Context) {
 		ssoConfig.Name = dingtalk.SsoTypeName
 		ssoConfigs = append(ssoConfigs, ssoConfig)
 	}
+	if !feiShuExist {
+		var ssoConfig models.SsoConfig
+		ssoConfig.Name = feishu.SsoTypeName
+		ssoConfigs = append(ssoConfigs, ssoConfig)
+	}
 
 	ginx.NewRender(c).Data(ssoConfigs, nil)
 }
@@ -657,6 +760,23 @@ func (rt *Router) ssoConfigUpdate(c *gin.Context) {
 			err = f.Update(rt.Ctx)
 		}
 		ginx.Dangerous(err)
+	case feishu.SsoTypeName:
+		f.Name = ssoConfig.Name
+		setting, err := json.Marshal(ssoConfig.SettingJson)
+		ginx.Dangerous(err)
+		f.Content = string(setting)
+		f.UpdateAt = time.Now().Unix()
+		sso, err := f.Query(rt.Ctx)
+		if !errors.Is(err, gorm.ErrRecordNotFound) {
+			ginx.Dangerous(err)
+		}
+		if errors.Is(err, gorm.ErrRecordNotFound) {
+			err = f.Create(rt.Ctx)
+		} else {
+			f.Id = sso.Id
+			err = f.Update(rt.Ctx)
+		}
+		ginx.Dangerous(err)
 	default:
 		f.Id = ssoConfig.Id
 		f.Name = ssoConfig.Name
@@ -695,6 +815,14 @@ func (rt *Router) ssoConfigUpdate(c *gin.Context) {
 			rt.Sso.DingTalk = dingtalk.New(config)
 		}
 		rt.Sso.DingTalk.Reload(config)
+	case feishu.SsoTypeName:
+		var config feishu.Config
+		err := json.Unmarshal([]byte(f.Content), &config)
+		ginx.Dangerous(err)
+		if rt.Sso.FeiShu == nil {
+			rt.Sso.FeiShu = feishu.New(config)
+		}
+		rt.Sso.FeiShu.Reload(config)
 	}
 
 	ginx.NewRender(c).Message(nil)

center/router/router_query.go

@@ -1,13 +1,16 @@
 package router
 
 import (
+	"context"
 	"fmt"
 	"sort"
 	"sync"
 
 	"github.com/ccfos/nightingale/v6/alert/eval"
 	"github.com/ccfos/nightingale/v6/dscache"
+	"github.com/ccfos/nightingale/v6/dskit/doris"
 	"github.com/ccfos/nightingale/v6/models"
+
 	"github.com/gin-gonic/gin"
 	"github.com/toolkits/pkg/ginx"
 	"github.com/toolkits/pkg/logger"
@@ -117,10 +120,13 @@ func (rt *Router) QueryLogBatch(c *gin.Context) {
 }
 
 func QueryDataConcurrently(anonymousAccess bool, ctx *gin.Context, f models.QueryParam) ([]models.DataResp, error) {
-	var resp []models.DataResp
-	var mu sync.Mutex
-	var wg sync.WaitGroup
-	var errs []error
+	var (
+		resp []models.DataResp
+		mu   sync.Mutex
+		wg   sync.WaitGroup
+		errs []error
+		rCtx = ctx.Request.Context()
+	)
 
 	for _, q := range f.Queries {
 		if !anonymousAccess && !CheckDsPerm(ctx, f.DatasourceId, f.Cate, q) {
@@ -132,12 +138,17 @@ func QueryDataConcurrently(anonymousAccess bool, ctx *gin.Context, f models.Quer
 			logger.Warningf("cluster:%d not exists", f.DatasourceId)
 			return nil, fmt.Errorf("cluster not exists")
 		}
+		
+		vCtx := rCtx
+		if f.Cate == models.DORIS {
+			vCtx = context.WithValue(vCtx, doris.NoNeedCheckMaxRow, true)
+		}
 
 		wg.Add(1)
 		go func(query interface{}) {
 			defer wg.Done()
 
-			data, err := plug.QueryData(ctx.Request.Context(), query)
+			data, err := plug.QueryData(vCtx, query)
 			if err != nil {
 				logger.Warningf("query data error: req:%+v err:%v", query, err)
 				mu.Lock()

center/router/router_recording_rule.go

@@ -112,6 +112,7 @@ func (rt *Router) recordingRulePutByFE(c *gin.Context) {
 	}
 
 	rt.bgrwCheck(c, ar.GroupId)
+	rt.bgroCheck(c, f.GroupId)
 
 	f.UpdateBy = c.MustGet("username").(string)
 	ginx.NewRender(c).Message(ar.Update(rt.Ctx, f))

center/router/router_target.go

@@ -38,6 +38,16 @@ func (rt *Router) targetGetsByHostFilter(c *gin.Context) {
 	total, err := models.TargetCountByFilter(rt.Ctx, query)
 	ginx.Dangerous(err)
 
+	models.FillTargetsBeatTime(rt.Redis, hosts)
+	now := time.Now().Unix()
+	for i := 0; i < len(hosts); i++ {
+		if now-hosts[i].BeatTime < 60 {
+			hosts[i].TargetUp = 2
+		} else if now-hosts[i].BeatTime < 180 {
+			hosts[i].TargetUp = 1
+		}
+	}
+
 	ginx.NewRender(c).Data(gin.H{
 		"list":  hosts,
 		"total": total,
@@ -81,9 +91,24 @@ func (rt *Router) targetGets(c *gin.Context) {
 		models.BuildTargetWhereWithBgids(bgids),
 		models.BuildTargetWhereWithDsIds(dsIds),
 		models.BuildTargetWhereWithQuery(query),
-		models.BuildTargetWhereWithDowntime(downtime),
 		models.BuildTargetWhereWithHosts(hosts),
 	}
+
+	// downtime 筛选：从缓存获取心跳时间，选择较小的集合用 IN 或 NOT IN 过滤
+	if downtime != 0 {
+		downtimeOpt, hasMatch := rt.downtimeFilter(downtime)
+		if !hasMatch {
+			ginx.NewRender(c).Data(gin.H{
+				"list":  []*models.Target{},
+				"total": 0,
+			}, nil)
+			return
+		}
+		if downtimeOpt != nil {
+			options = append(options, downtimeOpt)
+		}
+	}
+
 	total, err := models.TargetTotal(rt.Ctx, options...)
 	ginx.Dangerous(err)
 
@@ -102,14 +127,17 @@ func (rt *Router) targetGets(c *gin.Context) {
 		now := time.Now()
 		cache := make(map[int64]*models.BusiGroup)
 
+		// 从 Redis 补全 BeatTime
+		models.FillTargetsBeatTime(rt.Redis, list)
+
 		var keys []string
 		for i := 0; i < len(list); i++ {
 			ginx.Dangerous(list[i].FillGroup(rt.Ctx, cache))
 			keys = append(keys, models.WrapIdent(list[i].Ident))
 
-			if now.Unix()-list[i].UpdateAt < 60 {
+			if now.Unix()-list[i].BeatTime < 60 {
 				list[i].TargetUp = 2
-			} else if now.Unix()-list[i].UpdateAt < 180 {
+			} else if now.Unix()-list[i].BeatTime < 180 {
 				list[i].TargetUp = 1
 			}
 		}
@@ -148,6 +176,43 @@ func (rt *Router) targetGets(c *gin.Context) {
 	}, nil)
 }
 
+// downtimeFilter 从缓存获取心跳时间，生成 downtime 筛选条件
+// 选择匹配集和非匹配集中较小的一方，用 IN 或 NOT IN 来减少 SQL 参数量
+// 返回值：
+//   - option: 筛选条件，nil 表示所有 target 都符合条件（无需过滤）
+//   - hasMatch: 是否有符合条件的 target，false 表示无匹配应返回空结果
+func (rt *Router) downtimeFilter(downtime int64) (option models.BuildTargetWhereOption, hasMatch bool) {
+	now := time.Now().Unix()
+	targets := rt.TargetCache.GetAll()
+	var matchIdents, nonMatchIdents []string
+	for _, target := range targets {
+		matched := false
+		if downtime > 0 {
+			matched = target.BeatTime < now-downtime
+		} else if downtime < 0 {
+			matched = target.BeatTime > now+downtime
+		}
+		if matched {
+			matchIdents = append(matchIdents, target.Ident)
+		} else {
+			nonMatchIdents = append(nonMatchIdents, target.Ident)
+		}
+	}
+
+	if len(matchIdents) == 0 {
+		return nil, false
+	}
+
+	if len(nonMatchIdents) == 0 {
+		return nil, true
+	}
+
+	if len(matchIdents) <= len(nonMatchIdents) {
+		return models.BuildTargetWhereWithIdents(matchIdents), true
+	}
+	return models.BuildTargetWhereExcludeIdents(nonMatchIdents), true
+}
+
 func (rt *Router) targetExtendInfoByIdent(c *gin.Context) {
 	ident := ginx.QueryStr(c, "ident", "")
 	key := models.WrapExtendIdent(ident)

center/sso/init.go

@@ -12,6 +12,7 @@ import (
 	"github.com/ccfos/nightingale/v6/pkg/cas"
 	"github.com/ccfos/nightingale/v6/pkg/ctx"
 	"github.com/ccfos/nightingale/v6/pkg/dingtalk"
+	"github.com/ccfos/nightingale/v6/pkg/feishu"
 	"github.com/ccfos/nightingale/v6/pkg/ldapx"
 	"github.com/ccfos/nightingale/v6/pkg/oauth2x"
 	"github.com/ccfos/nightingale/v6/pkg/oidcx"
@@ -27,6 +28,7 @@ type SsoClient struct {
 	CAS                  *cas.SsoClient
 	OAuth2               *oauth2x.SsoClient
 	DingTalk             *dingtalk.SsoClient
+	FeiShu               *feishu.SsoClient
 	LastUpdateTime       int64
 	configCache          *memsto.ConfigCache
 	configLastUpdateTime int64
@@ -203,6 +205,13 @@ func Init(center cconf.Center, ctx *ctx.Context, configCache *memsto.ConfigCache
 				log.Fatalf("init %s failed: %s", dingtalk.SsoTypeName, err)
 			}
 			ssoClient.DingTalk = dingtalk.New(config)
+		case feishu.SsoTypeName:
+			var config feishu.Config
+			err := json.Unmarshal([]byte(cfg.Content), &config)
+			if err != nil {
+				log.Fatalf("init %s failed: %s", feishu.SsoTypeName, err)
+			}
+			ssoClient.FeiShu = feishu.New(config)
 		}
 	}
 
@@ -291,6 +300,22 @@ func (s *SsoClient) reload(ctx *ctx.Context) error {
 		s.DingTalk = nil
 	}
 
+	if feiShuConfig, ok := ssoConfigMap[feishu.SsoTypeName]; ok {
+		var config feishu.Config
+		err := json.Unmarshal([]byte(feiShuConfig.Content), &config)
+		if err != nil {
+			logger.Warningf("reload %s failed: %s", feishu.SsoTypeName, err)
+		} else {
+			if s.FeiShu != nil {
+				s.FeiShu.Reload(config)
+			} else {
+				s.FeiShu = feishu.New(config)
+			}
+		}
+	} else {
+		s.FeiShu = nil
+	}
+
 	s.LastUpdateTime = lastUpdateTime
 	s.configLastUpdateTime = lastCacheUpdateTime
 	return nil

cli/upgrade/upgrade.sql

@@ -71,7 +71,10 @@ CREATE TABLE `datasource`
     `updated_at` bigint not null default 0,
     `updated_by` varchar(64) not null default '',
     PRIMARY KEY (`id`)
-) ENGINE = InnoDB DEFAULT CHARSET = utf8mb4; 
+) ENGINE = InnoDB DEFAULT CHARSET = utf8mb4;
+
+-- datasource add weight field
+alter table `datasource` add `weight` int not null default 0;
 
 CREATE TABLE `builtin_cate` (
     `id` bigint unsigned not null auto_increment,

datasource/commons/eslike/eslike.go

@@ -565,6 +565,14 @@ func QueryData(ctx context.Context, queryParam interface{}, cliTimeout int64, ve
 		return nil, err
 	}
 
+	// 检查是否有 shard failures，有部分数据时仅记录警告继续处理
+	if shardErr := checkShardFailures(result.Shards, "query_data", searchSourceString); shardErr != nil {
+		if len(result.Aggregations["ts"]) == 0 {
+			return nil, shardErr
+		}
+		// 有部分数据，checkShardFailures 已记录警告，继续处理
+	}
+
 	logger.Debugf("query_data searchSource:%s resp:%s", string(jsonSearchSource), string(result.Aggregations["ts"]))
 
 	js, err := simplejson.NewJson(result.Aggregations["ts"])
@@ -602,6 +610,40 @@ func QueryData(ctx context.Context, queryParam interface{}, cliTimeout int64, ve
 	return items, nil
 }
 
+// checkShardFailures 检查 ES 查询结果中的 shard failures，返回格式化的错误信息
+func checkShardFailures(shards *elastic.ShardsInfo, logPrefix string, queryContext interface{}) error {
+	if shards == nil || shards.Failed == 0 || len(shards.Failures) == 0 {
+		return nil
+	}
+
+	var failureReasons []string
+	for _, failure := range shards.Failures {
+		reason := ""
+		if failure.Reason != nil {
+			if reasonType, ok := failure.Reason["type"].(string); ok {
+				reason = reasonType
+			}
+			if reasonMsg, ok := failure.Reason["reason"].(string); ok {
+				if reason != "" {
+					reason += ": " + reasonMsg
+				} else {
+					reason = reasonMsg
+				}
+			}
+		}
+		if reason != "" {
+			failureReasons = append(failureReasons, fmt.Sprintf("index=%s shard=%d: %s", failure.Index, failure.Shard, reason))
+		}
+	}
+
+	if len(failureReasons) > 0 {
+		errMsg := fmt.Sprintf("elasticsearch shard failures (%d/%d failed): %s", shards.Failed, shards.Total, strings.Join(failureReasons, "; "))
+		logger.Warningf("%s query:%v %s", logPrefix, queryContext, errMsg)
+		return fmt.Errorf("%s", errMsg)
+	}
+	return nil
+}
+
 func HitFilter(typ string) bool {
 	switch typ {
 	case "keyword", "date", "long", "integer", "short", "byte", "double", "float", "half_float", "scaled_float", "unsigned_long":
@@ -678,21 +720,27 @@ func QueryLog(ctx context.Context, queryParam interface{}, timeout int64, versio
 	} else {
 		source = source.From(param.P).Sort(param.DateField, param.Ascending)
 	}
+	sourceBytes, _ := json.Marshal(source)
 	result, err := search(ctx, indexArr, source, param.Timeout, param.MaxShard)
 	if err != nil {
-		logger.Warningf("query data error:%v", err)
+		logger.Warningf("query_log source:%s error:%v", string(sourceBytes), err)
 		return nil, 0, err
 	}
 
+	// 检查是否有 shard failures，有部分数据时仅记录警告继续处理
+	if shardErr := checkShardFailures(result.Shards, "query_log", string(sourceBytes)); shardErr != nil {
+		if len(result.Hits.Hits) == 0 {
+			return nil, 0, shardErr
+		}
+		// 有部分数据，checkShardFailures 已记录警告，继续处理
+	}
+
 	total := result.TotalHits()
-
 	var ret []interface{}
-
-	b, _ := json.Marshal(source)
-	logger.Debugf("query data result query source:%s len:%d total:%d", string(b), len(result.Hits.Hits), total)
+	logger.Debugf("query_log source:%s len:%d total:%d", string(sourceBytes), len(result.Hits.Hits), total)
 
 	resultBytes, _ := json.Marshal(result)
-	logger.Debugf("query data result query source:%s result:%s", string(b), string(resultBytes))
+	logger.Debugf("query_log source:%s result:%s", string(sourceBytes), string(resultBytes))
 
 	if strings.HasPrefix(version, "6") {
 		for i := 0; i < len(result.Hits.Hits); i++ {

datasource/datasource.go

@@ -133,4 +133,5 @@ type DatasourceInfo struct {
 	CreatedAt      int64                  `json:"created_at"`
 	UpdatedAt      int64                  `json:"updated_at"`
 	IsDefault      bool                   `json:"is_default"`
+	Weight         int                    `json:"weight"`
 }

datasource/doris/doris.go

@@ -79,52 +79,19 @@ func (d *Doris) Equal(p datasource.Datasource) bool {
 		return false
 	}
 
-	// only compare first shard
-	if d.Addr != newest.Addr {
-		return false
-	}
-
-	if d.User != newest.User {
-		return false
-	}
-
-	if d.Password != newest.Password {
-		return false
-	}
-
-	if d.EnableWrite != newest.EnableWrite {
-		return false
-	}
-
-	if d.FeAddr != newest.FeAddr {
-		return false
-	}
-
-	if d.MaxQueryRows != newest.MaxQueryRows {
-		return false
-	}
-
-	if d.Timeout != newest.Timeout {
-		return false
-	}
-
-	if d.MaxIdleConns != newest.MaxIdleConns {
-		return false
-	}
-
-	if d.MaxOpenConns != newest.MaxOpenConns {
-		return false
-	}
-
-	if d.ConnMaxLifetime != newest.ConnMaxLifetime {
-		return false
-	}
-
-	if d.ClusterName != newest.ClusterName {
-		return false
-	}
-
-	return true
+	return d.Addr == newest.Addr &&
+		d.FeAddr == newest.FeAddr &&
+		d.User == newest.User &&
+		d.Password == newest.Password &&
+		d.EnableWrite == newest.EnableWrite &&
+		d.UserWrite == newest.UserWrite &&
+		d.PasswordWrite == newest.PasswordWrite &&
+		d.MaxQueryRows == newest.MaxQueryRows &&
+		d.Timeout == newest.Timeout &&
+		d.MaxIdleConns == newest.MaxIdleConns &&
+		d.MaxOpenConns == newest.MaxOpenConns &&
+		d.ConnMaxLifetime == newest.ConnMaxLifetime &&
+		d.ClusterName == newest.ClusterName
 }
 
 func (d *Doris) MakeLogQuery(ctx context.Context, query interface{}, eventTags []string, start, end int64) (interface{}, error) {
@@ -181,7 +148,7 @@ func (d *Doris) QueryData(ctx context.Context, query interface{}) ([]models.Data
 		}
 	}
 
-	items, err := d.QueryTimeseries(context.TODO(), &doris.QueryParam{
+	items, err := d.QueryTimeseries(ctx, &doris.QueryParam{
 		Database: dorisQueryParam.Database,
 		Sql:      dorisQueryParam.SQL,
 		Keys: types.Keys{

docker/compose-postgres/initsql_for_postgres/a-n9e-for-Postgres.sql

@@ -738,6 +738,7 @@ CREATE TABLE datasource
     http varchar(4096) not null default '',
     auth varchar(8192) not null default '',
     is_default boolean not null default false,
+    weight int not null default 0,
     created_at bigint not null default 0,
     created_by varchar(64) not null default '',
     updated_at bigint not null default 0,

docker/initsql/a-n9e.sql

@@ -655,6 +655,7 @@ CREATE TABLE `datasource`
     `http` varchar(4096) not null default '',
     `auth` varchar(8192) not null default '',
     `is_default` boolean COMMENT 'is default datasource',
+    `weight` int not null default 0,
     `created_at` bigint not null default 0,
     `created_by` varchar(64) not null default '',
     `updated_at` bigint not null default 0,

docker/migratesql/migrate.sql

@@ -301,7 +301,7 @@ ALTER TABLE `event_pipeline` ADD COLUMN `trigger_mode` varchar(128) NOT NULL DEF
 ALTER TABLE `event_pipeline` ADD COLUMN `disabled` tinyint(1) NOT NULL DEFAULT 0 COMMENT 'disabled flag';
 ALTER TABLE `event_pipeline` ADD COLUMN `nodes` text COMMENT 'workflow nodes (JSON)';
 ALTER TABLE `event_pipeline` ADD COLUMN `connections` text COMMENT 'node connections (JSON)';
-ALTER TABLE `event_pipeline` ADD COLUMN `env_variables` text COMMENT 'environment variables (JSON)';
+ALTER TABLE `event_pipeline` ADD COLUMN `input_variables` text COMMENT 'input variables (JSON)';
 ALTER TABLE `event_pipeline` ADD COLUMN `label_filters` text COMMENT 'label filters (JSON)';
 
 CREATE TABLE `event_pipeline_execution` (
@@ -318,7 +318,7 @@ CREATE TABLE `event_pipeline_execution` (
     `finished_at` bigint DEFAULT 0 COMMENT 'finish timestamp',
     `duration_ms` bigint DEFAULT 0 COMMENT 'duration in milliseconds',
     `trigger_by` varchar(64) DEFAULT '' COMMENT 'trigger by',
-    `env_snapshot` text COMMENT 'environment variables snapshot (sanitized)',
+    `inputs_snapshot` text COMMENT 'inputs snapshot',
     PRIMARY KEY (`id`),
     KEY `idx_pipeline_id` (`pipeline_id`),
     KEY `idx_event_id` (`event_id`),
@@ -331,3 +331,33 @@ CREATE TABLE `event_pipeline_execution` (
 ALTER TABLE `builtin_metrics` ADD COLUMN `expression_type` varchar(32) NOT NULL DEFAULT 'promql' COMMENT 'expression type: metric_name or promql';
 ALTER TABLE `builtin_metrics` ADD COLUMN `metric_type` varchar(191) NOT NULL DEFAULT '' COMMENT 'metric type like counter/gauge';
 ALTER TABLE `builtin_metrics` ADD COLUMN `extra_fields` text COMMENT 'custom extra fields';
+
+/* v9 2026-01-16 saved_view */
+CREATE TABLE `saved_view` (
+    `id` bigint NOT NULL AUTO_INCREMENT,
+    `name` varchar(255) NOT NULL COMMENT 'view name',
+    `page` varchar(64) NOT NULL COMMENT 'page identifier',
+    `filter` text COMMENT 'filter config (JSON)',
+    `public_cate` int NOT NULL DEFAULT 0 COMMENT 'public category: 0-self, 1-team, 2-all',
+    `gids` text COMMENT 'team group ids (JSON)',
+    `create_at` bigint NOT NULL DEFAULT 0 COMMENT 'create timestamp',
+    `create_by` varchar(64) NOT NULL DEFAULT '' COMMENT 'creator',
+    `update_at` bigint NOT NULL DEFAULT 0 COMMENT 'update timestamp',
+    `update_by` varchar(64) NOT NULL DEFAULT '' COMMENT 'updater',
+    PRIMARY KEY (`id`),
+    KEY `idx_page` (`page`),
+    KEY `idx_create_by` (`create_by`)
+) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COMMENT='saved views for pages';
+
+CREATE TABLE `user_view_favorite` (
+    `id` bigint NOT NULL AUTO_INCREMENT,
+    `view_id` bigint NOT NULL COMMENT 'saved view id',
+    `user_id` bigint NOT NULL COMMENT 'user id',
+    `create_at` bigint NOT NULL DEFAULT 0 COMMENT 'create timestamp',
+    PRIMARY KEY (`id`),
+    KEY `idx_view_id` (`view_id`),
+    KEY `idx_user_id` (`user_id`)
+) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COMMENT='user favorite views';
+
+/* v9 2026-01-20 datasource weight */
+ALTER TABLE `datasource` ADD COLUMN `weight` int not null default 0 COMMENT 'weight for sorting';

docker/sqlite.sql

@@ -589,6 +589,7 @@ CREATE TABLE `datasource`
     `http` varchar(4096) not null default '',
     `auth` varchar(8192) not null default '',
     `is_default` tinyint not null default 0,
+    `weight` int not null default 0,
     `created_at` bigint not null default 0,
     `created_by` varchar(64) not null default '',
     `updated_at` bigint not null default 0,

dscache/sync.go

@@ -90,7 +90,7 @@ func getDatasourcesFromDBLoop(ctx *ctx.Context, fromAPI bool) {
 					foundDefaultDatasource = true
 				}
 
-				logger.Debugf("get datasource: %+v", item)
+				// logger.Debugf("get datasource: %+v", item)
 				ds := datasource.DatasourceInfo{
 					Id:             item.Id,
 					Name:           item.Name,
@@ -104,6 +104,7 @@ func getDatasourcesFromDBLoop(ctx *ctx.Context, fromAPI bool) {
 					AuthJson:       item.AuthJson,
 					Status:         item.Status,
 					IsDefault:      item.IsDefault,
+					Weight:         item.Weight,
 				}
 
 				if item.PluginType == "elasticsearch" {
@@ -236,5 +237,5 @@ func PutDatasources(items []datasource.DatasourceInfo) {
 		}
 	}
 
-	logger.Debugf("get plugin by type success Ids:%v", ids)
+	// logger.Debugf("get plugin by type success Ids:%v", ids)
 }

dskit/doris/doris.go

@@ -39,6 +39,9 @@ type Doris struct {
 	MaxQueryRows    int    `json:"doris.max_query_rows" mapstructure:"doris.max_query_rows"`
 	ClusterName     string `json:"doris.cluster_name" mapstructure:"doris.cluster_name"`
 	EnableWrite     bool   `json:"doris.enable_write" mapstructure:"doris.enable_write"`
+	// 写用户，用来区分读写用户，减少数据源
+	UserWrite     string `json:"doris.user_write" mapstructure:"doris.user_write"`
+	PasswordWrite string `json:"doris.password_write" mapstructure:"doris.password_write"`
 }
 
 // NewDorisWithSettings initializes a new Doris instance with the given settings
@@ -88,13 +91,13 @@ func (d *Doris) NewConn(ctx context.Context, database string) (*sql.DB, error) {
 
 	var keys []string
 	keys = append(keys, d.Addr)
-	keys = append(keys, d.Password, d.User)
+	keys = append(keys, d.User, d.Password)
 	if len(database) > 0 {
 		keys = append(keys, database)
 	}
-	cachedkey := strings.Join(keys, ":")
+	cachedKey := strings.Join(keys, ":")
 	// cache conn with database
-	conn, ok := pool.PoolClient.Load(cachedkey)
+	conn, ok := pool.PoolClient.Load(cachedKey)
 	if ok {
 		return conn.(*sql.DB), nil
 	}
@@ -102,7 +105,7 @@ func (d *Doris) NewConn(ctx context.Context, database string) (*sql.DB, error) {
 	var err error
 	defer func() {
 		if db != nil && err == nil {
-			pool.PoolClient.Store(cachedkey, db)
+			pool.PoolClient.Store(cachedKey, db)
 		}
 	}()
 
@@ -121,6 +124,79 @@ func (d *Doris) NewConn(ctx context.Context, database string) (*sql.DB, error) {
 	return db, nil
 }
 
+// NewWriteConn establishes a new connection to Doris for write operations
+// When EnableWrite is true and UserWrite is configured, it uses the write user credentials
+// Otherwise, it reuses the read connection from NewConn
+func (d *Doris) NewWriteConn(ctx context.Context, database string) (*sql.DB, error) {
+	// If write user is not configured, reuse the read connection
+	if !d.EnableWrite || len(d.UserWrite) == 0 {
+		return d.NewConn(ctx, database)
+	}
+
+	if len(d.Addr) == 0 {
+		return nil, errors.New("empty fe-node addr")
+	}
+
+	// Set default values similar to postgres implementation
+	if d.Timeout == 0 {
+		d.Timeout = 60000
+	}
+	if d.MaxIdleConns == 0 {
+		d.MaxIdleConns = 10
+	}
+	if d.MaxOpenConns == 0 {
+		d.MaxOpenConns = 100
+	}
+	if d.ConnMaxLifetime == 0 {
+		d.ConnMaxLifetime = 14400
+	}
+	if d.MaxQueryRows == 0 {
+		d.MaxQueryRows = 500
+	}
+
+	// Use write user credentials
+	user := d.UserWrite
+	password := d.PasswordWrite
+
+	var keys []string
+	keys = append(keys, d.Addr)
+	keys = append(keys, user, password)
+	if len(database) > 0 {
+		keys = append(keys, database)
+	}
+	cachedKey := strings.Join(keys, ":")
+	// cache conn with database
+	conn, ok := pool.PoolClient.Load(cachedKey)
+	if ok {
+		return conn.(*sql.DB), nil
+	}
+	var db *sql.DB
+	var err error
+	defer func() {
+		if db != nil && err == nil {
+			pool.PoolClient.Store(cachedKey, db)
+		}
+	}()
+
+	// Simplified connection logic for Doris using MySQL driver
+	dsn := fmt.Sprintf("%s:%s@tcp(%s)/%s?charset=utf8", user, password, d.Addr, database)
+	db, err = sql.Open("mysql", dsn)
+	if err != nil {
+		return nil, err
+	}
+
+	// Set connection pool configuration for write connections
+	// Use more conservative values since write operations are typically less frequent
+	writeMaxIdleConns := max(d.MaxIdleConns/5, 2)
+	writeMaxOpenConns := max(d.MaxOpenConns/10, 5)
+
+	db.SetMaxIdleConns(writeMaxIdleConns)
+	db.SetMaxOpenConns(writeMaxOpenConns)
+	db.SetConnMaxLifetime(time.Duration(d.ConnMaxLifetime) * time.Second)
+
+	return db, nil
+}
+
 // createTimeoutContext creates a context with timeout based on Doris configuration
 func (d *Doris) createTimeoutContext(ctx context.Context) (context.Context, context.CancelFunc) {
 	timeout := d.Timeout
@@ -472,7 +548,7 @@ func (d *Doris) ExecContext(ctx context.Context, database string, sql string) er
 	timeoutCtx, cancel := d.createTimeoutContext(ctx)
 	defer cancel()
 
-	db, err := d.NewConn(timeoutCtx, database)
+	db, err := d.NewWriteConn(timeoutCtx, database)
 	if err != nil {
 		return err
 	}

dskit/doris/logs.go

@@ -10,13 +10,14 @@ const (
 	TimeseriesAggregationTimestamp = "__ts__"
 )
 
+// QueryLogs 查询日志
 // TODO: 待测试, MAP/ARRAY/STRUCT/JSON 等类型能否处理
 func (d *Doris) QueryLogs(ctx context.Context, query *QueryParam) ([]map[string]interface{}, error) {
 	// 等同于 Query()
-	return d.Query(ctx, query)
+	return d.Query(ctx, query, true)
 }
 
-// 本质是查询时序数据, 取第一组, SQL由上层封装, 不再做复杂的解析和截断
+// QueryHistogram 本质是查询时序数据, 取第一组, SQL由上层封装, 不再做复杂的解析和截断
 func (d *Doris) QueryHistogram(ctx context.Context, query *QueryParam) ([][]float64, error) {
 	values, err := d.QueryTimeseries(ctx, query)
 	if err != nil {

dskit/doris/sql_analyzer.go

@@ -0,0 +1,324 @@
+package doris
+
+import (
+	"regexp"
+	"strings"
+
+	"github.com/pingcap/tidb/pkg/parser"
+	"github.com/pingcap/tidb/pkg/parser/ast"
+	_ "github.com/pingcap/tidb/pkg/parser/test_driver" // required for parser
+)
+
+// mapAccessPattern matches Doris map/array access syntax like `col['key']` or col["key"]
+var mapAccessPattern = regexp.MustCompile(`\[['"]\w+['"]\]`)
+
+// castStringPattern matches Doris CAST(... AS STRING) syntax
+var castStringPattern = regexp.MustCompile(`(?i)\bAS\s+STRING\b`)
+
+// macro patterns
+var timeGroupPattern = regexp.MustCompile(`\$__timeGroup\([^)]+\)`)
+var timeFilterPattern = regexp.MustCompile(`\$__timeFilter\([^)]+\)`)
+var intervalPattern = regexp.MustCompile(`\$__interval`)
+
+// SQLAnalyzeResult holds the analysis result of a SQL statement
+type SQLAnalyzeResult struct {
+	IsSelectLike bool   // whether the statement is a SELECT-like query
+	HasTopAgg    bool   // whether the top-level query has aggregate functions
+	LimitConst   *int64 // top-level LIMIT constant value (nil if no LIMIT or non-constant)
+}
+
+// AnalyzeSQL analyzes a SQL statement and extracts top-level features
+func AnalyzeSQL(sql string) (*SQLAnalyzeResult, error) {
+	// Preprocess SQL to remove Doris-specific syntax that TiDB parser doesn't support
+	preprocessedSQL := preprocessDorisSQL(sql)
+
+	p := parser.New()
+	stmtNodes, _, err := p.Parse(preprocessedSQL, "", "")
+	if err != nil {
+		return nil, err
+	}
+	if len(stmtNodes) == 0 {
+		return &SQLAnalyzeResult{}, nil
+	}
+
+	result := &SQLAnalyzeResult{}
+	stmt := stmtNodes[0]
+
+	switch s := stmt.(type) {
+	case *ast.SelectStmt:
+		result.IsSelectLike = true
+		analyzeSelectStmt(s, result)
+	case *ast.SetOprStmt: // UNION / INTERSECT / EXCEPT
+		result.IsSelectLike = true
+		analyzeSetOprStmt(s, result)
+	default:
+		result.IsSelectLike = false
+	}
+
+	return result, nil
+}
+
+// analyzeSelectStmt analyzes a SELECT statement
+func analyzeSelectStmt(sel *ast.SelectStmt, result *SQLAnalyzeResult) {
+	// Check if top-level SELECT has aggregate functions
+	if sel.Fields != nil {
+		for _, field := range sel.Fields.Fields {
+			if field.Expr != nil && hasAggregateFunc(field.Expr) {
+				result.HasTopAgg = true
+				break
+			}
+		}
+	}
+
+	// Check if any CTE has aggregate functions
+	if !result.HasTopAgg && sel.With != nil {
+		for _, cte := range sel.With.CTEs {
+			if selectHasAggregate(cte.Query) {
+				result.HasTopAgg = true
+				break
+			}
+		}
+	}
+
+	// Extract top-level LIMIT
+	if sel.Limit != nil && sel.Limit.Count != nil {
+		if val, ok := extractConstValue(sel.Limit.Count); ok {
+			result.LimitConst = &val
+		}
+	}
+}
+
+// selectHasAggregate checks if a node (SELECT, UNION, or SubqueryExpr) has aggregate functions
+func selectHasAggregate(node ast.Node) bool {
+	switch n := node.(type) {
+	case *ast.SelectStmt:
+		if n.Fields != nil {
+			for _, field := range n.Fields.Fields {
+				if field.Expr != nil && hasAggregateFunc(field.Expr) {
+					return true
+				}
+			}
+		}
+	case *ast.SetOprStmt:
+		// For UNION, check all branches
+		if n.SelectList != nil {
+			for _, sel := range n.SelectList.Selects {
+				if selectHasAggregate(sel) {
+					return true
+				}
+			}
+		}
+	case *ast.SubqueryExpr:
+		// CTE query is wrapped in SubqueryExpr
+		if n.Query != nil {
+			return selectHasAggregate(n.Query)
+		}
+	}
+	return false
+}
+
+// analyzeSetOprStmt analyzes UNION/INTERSECT/EXCEPT statements
+func analyzeSetOprStmt(setOpr *ast.SetOprStmt, result *SQLAnalyzeResult) {
+	// UNION's LIMIT is at the outermost level
+	if setOpr.Limit != nil && setOpr.Limit.Count != nil {
+		if val, ok := extractConstValue(setOpr.Limit.Count); ok {
+			result.LimitConst = &val
+		}
+	}
+
+	// Check if all branches are aggregates (conservative: if any is non-aggregate, don't skip)
+	if setOpr.SelectList == nil || len(setOpr.SelectList.Selects) == 0 {
+		return
+	}
+
+	allAgg := true
+	for _, sel := range setOpr.SelectList.Selects {
+		if selectStmt, ok := sel.(*ast.SelectStmt); ok {
+			if selectStmt.Fields != nil {
+				hasAgg := false
+				for _, field := range selectStmt.Fields.Fields {
+					if field.Expr != nil && hasAggregateFunc(field.Expr) {
+						hasAgg = true
+						break
+					}
+				}
+				if !hasAgg {
+					allAgg = false
+					break
+				}
+			}
+		}
+	}
+	result.HasTopAgg = allAgg
+}
+
+// hasAggregateFunc checks if an expression contains aggregate functions (without entering subqueries)
+func hasAggregateFunc(expr ast.ExprNode) bool {
+	checker := &aggregateChecker{}
+	expr.Accept(checker)
+	return checker.found
+}
+
+// aggregateChecker implements ast.Visitor to find aggregate functions
+type aggregateChecker struct {
+	found bool
+}
+
+func (c *aggregateChecker) Enter(n ast.Node) (ast.Node, bool) {
+	if c.found {
+		return n, true // stop traversal
+	}
+
+	switch node := n.(type) {
+	case *ast.SubqueryExpr:
+		return n, true // don't enter subquery
+	case *ast.AggregateFuncExpr:
+		c.found = true
+		return n, true
+	case *ast.FuncCallExpr:
+		// Check for Doris-specific aggregate/statistic functions
+		funcName := strings.ToUpper(node.FnName.L)
+		if isDorisAggregateFunc(funcName) {
+			c.found = true
+			return n, true
+		}
+	}
+	return n, false // continue traversal
+}
+
+func (c *aggregateChecker) Leave(n ast.Node) (ast.Node, bool) {
+	return n, true
+}
+
+// isDorisAggregateFunc checks if a function is a Doris-specific aggregate/statistic function
+func isDorisAggregateFunc(funcName string) bool {
+	dorisAggFuncs := map[string]bool{
+		// Standard aggregates (in case parser doesn't recognize them)
+		"COUNT":     true,
+		"SUM":       true,
+		"AVG":       true,
+		"MIN":       true,
+		"MAX":       true,
+		"ANY":       true,
+		"ANY_VALUE": true,
+
+		// HLL related
+		"HLL_UNION_AGG":   true,
+		"HLL_RAW_AGG":     true,
+		"HLL_CARDINALITY": true,
+		"HLL_UNION":       true,
+		"HLL_HASH":        true,
+
+		// Bitmap related
+		"BITMAP_UNION":         true,
+		"BITMAP_UNION_COUNT":   true,
+		"BITMAP_INTERSECT":     true,
+		"BITMAP_COUNT":         true,
+		"BITMAP_AND_COUNT":     true,
+		"BITMAP_OR_COUNT":      true,
+		"BITMAP_XOR_COUNT":     true,
+		"BITMAP_AND_NOT_COUNT": true,
+
+		// Other aggregates
+		"PERCENTILE":            true,
+		"PERCENTILE_APPROX":     true,
+		"APPROX_COUNT_DISTINCT": true,
+		"NDV":                   true,
+		"COLLECT_LIST":          true,
+		"COLLECT_SET":           true,
+		"GROUP_CONCAT":          true,
+		"GROUP_BIT_AND":         true,
+		"GROUP_BIT_OR":          true,
+		"GROUP_BIT_XOR":         true,
+		"GROUPING":              true,
+		"GROUPING_ID":           true,
+
+		// Statistical functions
+		"STDDEV":      true,
+		"STDDEV_POP":  true,
+		"STDDEV_SAMP": true,
+		"STD":         true,
+		"VARIANCE":    true,
+		"VAR_POP":     true,
+		"VAR_SAMP":    true,
+		"COVAR_POP":   true,
+		"COVAR_SAMP":  true,
+		"CORR":        true,
+
+		// Window functions that are also aggregates
+		"FIRST_VALUE":  true,
+		"LAST_VALUE":   true,
+		"LAG":          true,
+		"LEAD":         true,
+		"ROW_NUMBER":   true,
+		"RANK":         true,
+		"DENSE_RANK":   true,
+		"NTILE":        true,
+		"CUME_DIST":    true,
+		"PERCENT_RANK": true,
+	}
+	return dorisAggFuncs[funcName]
+}
+
+// extractConstValue extracts constant integer value from an expression
+func extractConstValue(expr ast.ExprNode) (int64, bool) {
+	switch v := expr.(type) {
+	case ast.ValueExpr:
+		switch val := v.GetValue().(type) {
+		case int64:
+			return val, true
+		case uint64:
+			return int64(val), true
+		case float64:
+			return int64(val), true
+		case int:
+			return int64(val), true
+		}
+	}
+	return 0, false
+}
+
+// preprocessDorisSQL removes Doris-specific syntax that TiDB parser doesn't support
+func preprocessDorisSQL(sql string) string {
+	// Remove map/array access syntax like ['key'] or ["key"]
+	// This is used in Doris for accessing map/variant/json fields
+	sql = mapAccessPattern.ReplaceAllString(sql, "")
+
+	// Replace Doris CAST(... AS STRING) with CAST(... AS CHAR)
+	sql = castStringPattern.ReplaceAllString(sql, "AS CHAR")
+
+	// Replace  macros with valid SQL equivalents
+	sql = timeGroupPattern.ReplaceAllString(sql, "ts")
+	sql = timeFilterPattern.ReplaceAllString(sql, "1=1")
+	sql = intervalPattern.ReplaceAllString(sql, "60")
+
+	return sql
+}
+
+// NeedsRowCountCheck determines if a SQL query needs row count checking
+// Returns: needsCheck bool, directReject bool, rejectReason string
+func NeedsRowCountCheck(sql string, maxQueryRows int) (bool, bool, string) {
+	result, err := AnalyzeSQL(sql)
+	if err != nil {
+		// Parse failed, fall back to probe check
+		return true, false, ""
+	}
+
+	if !result.IsSelectLike {
+		// Not a SELECT query, skip check
+		return false, false, ""
+	}
+
+	// Rule 1: Top-level has aggregate functions -> skip check
+	if result.HasTopAgg {
+		return false, false, ""
+	}
+
+	// Rule 2: Top-level LIMIT <= maxRows -> skip check
+	if result.LimitConst != nil && *result.LimitConst <= int64(maxQueryRows) {
+		return false, false, ""
+	}
+
+	// Otherwise, needs probe check (including LIMIT > maxRows, since actual result may be smaller)
+	return true, false, ""
+}

dskit/doris/sql_analyzer_test.go

@@ -0,0 +1,784 @@
+package doris
+
+import (
+	"testing"
+)
+
+func TestAnalyzeSQL_AggregateQueries(t *testing.T) {
+	tests := []struct {
+		name         string
+		sql          string
+		wantHasAgg   bool
+		wantIsSelect bool
+	}{
+		// Standard aggregate functions - should skip check
+		{
+			name:         "COUNT(*)",
+			sql:          "SELECT COUNT(*) AS `cnt`, FLOOR(UNIX_TIMESTAMP(event_date) DIV 10) * 10 AS `time`, CAST(`labels`['event'] AS STRING) AS `labels.event` FROM `db_insight_doris`.`ewall_event` WHERE `event_date` BETWEEN FROM_UNIXTIME(1768965669) AND FROM_UNIXTIME(1768965969) GROUP BY `time`, `labels.event` ORDER BY `time` ASC",
+			wantHasAgg:   true,
+			wantIsSelect: true,
+		},
+		{
+			name:         "COUNT with column",
+			sql:          "SELECT COUNT(id) FROM users",
+			wantHasAgg:   true,
+			wantIsSelect: true,
+		},
+		{
+			name:         "SUM function",
+			sql:          "SELECT SUM(amount) FROM orders",
+			wantHasAgg:   true,
+			wantIsSelect: true,
+		},
+		{
+			name:         "AVG function",
+			sql:          "SELECT AVG(price) FROM products",
+			wantHasAgg:   true,
+			wantIsSelect: true,
+		},
+		{
+			name:         "MIN function",
+			sql:          "SELECT MIN(created_at) FROM logs",
+			wantHasAgg:   true,
+			wantIsSelect: true,
+		},
+		{
+			name:         "MAX function",
+			sql:          "SELECT MAX(score) FROM results",
+			wantHasAgg:   true,
+			wantIsSelect: true,
+		},
+		{
+			name:         "Multiple aggregates",
+			sql:          "SELECT COUNT(*), SUM(amount), AVG(price) FROM orders",
+			wantHasAgg:   true,
+			wantIsSelect: true,
+		},
+		{
+			name:         "Aggregate with GROUP BY",
+			sql:          "SELECT user_id, COUNT(*) FROM orders GROUP BY user_id",
+			wantHasAgg:   true,
+			wantIsSelect: true,
+		},
+		{
+			name:         "Aggregate with WHERE and GROUP BY",
+			sql:          "SELECT category, SUM(sales) FROM products WHERE status = 'active' GROUP BY category",
+			wantHasAgg:   true,
+			wantIsSelect: true,
+		},
+		{
+			name:         "Aggregate with HAVING",
+			sql:          "SELECT user_id, COUNT(*) as cnt FROM orders GROUP BY user_id HAVING cnt > 10",
+			wantHasAgg:   true,
+			wantIsSelect: true,
+		},
+		// macro queries with aggregates
+		{
+			name:         "COUNT with timeGroup",
+			sql:          "SELECT COUNT(*) AS `cnt`, $__timeGroup(timestamp,$__interval) AS `time` FROM `apm`.`traces_span` WHERE (`service_name` = 'demo-logic-server') AND $__timeFilter(`timestamp`) GROUP BY `time` ORDER BY `time` ASC",
+			wantHasAgg:   true,
+			wantIsSelect: true,
+		},
+		{
+			name:         "CTE with ratio calculation",
+			sql:          "WITH `time_totals` AS (SELECT $__timeGroup(timestamp,$__interval) AS `time`, COUNT(*) AS `total_count` FROM `apm`.`traces_span` WHERE $__timeFilter(`timestamp`) GROUP BY `time`), `time_counts` AS (SELECT ANY_VALUE(`service_name`) AS `service_name`, $__timeGroup(timestamp,$__interval) AS `time`, COUNT(*) AS `count` FROM `apm`.`traces_span` WHERE (`service_name` = 'demo-logic-server') AND $__timeFilter(`timestamp`) GROUP BY `time`) SELECT tc.`service_name`, tc.`time`, ROUND(tc.`count` * 100.0 / tt.`total_count`, 2) AS `ratio` FROM `time_counts` tc JOIN `time_totals` tt ON tc.`time` = tt.`time` ORDER BY tc.`time` ASC",
+			wantHasAgg:   true, // CTE has aggregate functions
+			wantIsSelect: true,
+		},
+		{
+			name:         "CTE with top values and ratio",
+			sql:          "WITH `top_values` AS (SELECT `service_name` FROM `apm`.`traces_span` WHERE $__timeFilter(`timestamp`) GROUP BY `service_name` ORDER BY COUNT(*) DESC LIMIT 5), `time_totals` AS (SELECT $__timeGroup(timestamp,$__interval) AS `time`, COUNT(*) AS `total_count` FROM `apm`.`traces_span` WHERE $__timeFilter(`timestamp`) GROUP BY `time`), `time_counts` AS (SELECT `service_name`, $__timeGroup(timestamp,$__interval) AS `time`, COUNT(*) AS `count` FROM `apm`.`traces_span` WHERE $__timeFilter(`timestamp`) AND `service_name` IN (SELECT `service_name` FROM `top_values`) GROUP BY `service_name`, `time`) SELECT tc.`service_name`, tc.`time`, ROUND(tc.`count` * 100.0 / tt.`total_count`, 2) AS `ratio` FROM `time_counts` tc JOIN `time_totals` tt ON tc.`time` = tt.`time` ORDER BY tc.`time` ASC",
+			wantHasAgg:   true, // CTE has aggregate functions
+			wantIsSelect: true,
+		},
+		{
+			name:         "PERCENTILE_APPROX with timeGroup",
+			sql:          "SELECT PERCENTILE_APPROX(`duration`, 0.95) AS `p95`, $__timeGroup(timestamp,$__interval) AS `time` FROM `apm`.`traces_span` WHERE $__timeFilter(`timestamp`) GROUP BY `time` ORDER BY `time` ASC",
+			wantHasAgg:   true,
+			wantIsSelect: true,
+		},
+		{
+			name:         "COUNT DISTINCT with timeGroup",
+			sql:          "SELECT COUNT(DISTINCT `duration`) AS `unique_count`, $__timeGroup(timestamp,$__interval) AS `time` FROM `apm`.`traces_span` WHERE $__timeFilter(`timestamp`) GROUP BY `time` ORDER BY `time` ASC",
+			wantHasAgg:   true,
+			wantIsSelect: true,
+		},
+		{
+			name:         "CASE WHEN with COUNT and ROUND",
+			sql:          "SELECT ROUND(COUNT(CASE WHEN `duration` IS NOT NULL THEN 1 END) * 100.0 / COUNT(*), 2) AS `exist_ratio`, $__timeGroup(timestamp,$__interval) AS `time` FROM `apm`.`traces_span` WHERE $__timeFilter(`timestamp`) GROUP BY `time` ORDER BY `time` ASC",
+			wantHasAgg:   true,
+			wantIsSelect: true,
+		},
+		{
+			name:         "AVG with timeGroup",
+			sql:          "SELECT AVG(`duration`) AS `avg`, $__timeGroup(timestamp,$__interval) AS `time` FROM `apm`.`traces_span` WHERE $__timeFilter(`timestamp`) GROUP BY `time` ORDER BY `time` ASC",
+			wantHasAgg:   true,
+			wantIsSelect: true,
+		},
+		{
+			name:         "Simple COUNT with timeFilter",
+			sql:          "SELECT COUNT(*) AS `cnt` FROM `apm`.`traces_span` WHERE (`span_name` = 'GET /backend/detail') AND $__timeFilter(`timestamp`)",
+			wantHasAgg:   true,
+			wantIsSelect: true,
+		},
+		{
+			name:         "CTE with CROSS JOIN ratio",
+			sql:          "WITH `total` AS (SELECT COUNT(*) AS `total_count` FROM `apm`.`traces_span` WHERE $__timeFilter(`timestamp`)), `value_counts` AS (SELECT ANY_VALUE(`span_kind`) AS `span_kind`, COUNT(*) AS `count` FROM `apm`.`traces_span` WHERE (`span_kind` = 'SPAN_KIND_SERVER') AND $__timeFilter(`timestamp`)) SELECT vc.`span_kind`, vc.`count` AS `count`, ROUND(vc.`count` * 100.0 / t.`total_count`, 2) AS `ratio` FROM `value_counts` vc CROSS JOIN `total` t ORDER BY vc.`count` DESC;",
+			wantHasAgg:   true, // CTE has aggregate functions
+			wantIsSelect: true,
+		},
+		// Non-aggregate queries - should not skip check
+		{
+			name:         "Simple SELECT *",
+			sql:          "SELECT * FROM users",
+			wantHasAgg:   false,
+			wantIsSelect: true,
+		},
+		{
+			name:         "SELECT with columns",
+			sql:          "SELECT id, name, email FROM users",
+			wantHasAgg:   false,
+			wantIsSelect: true,
+		},
+		{
+			name:         "SELECT with WHERE",
+			sql:          "SELECT * FROM users WHERE status = 'active'",
+			wantHasAgg:   false,
+			wantIsSelect: true,
+		},
+		{
+			name:         "SELECT with JOIN",
+			sql:          "SELECT u.name, o.amount FROM users u JOIN orders o ON u.id = o.user_id",
+			wantHasAgg:   false,
+			wantIsSelect: true,
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			result, err := AnalyzeSQL(tt.sql)
+			if err != nil {
+				t.Fatalf("AnalyzeSQL() error = %v", err)
+			}
+			if result.HasTopAgg != tt.wantHasAgg {
+				t.Errorf("name: %s, HasTopAgg = %v, want %v", tt.name, result.HasTopAgg, tt.wantHasAgg)
+			}
+			if result.IsSelectLike != tt.wantIsSelect {
+				t.Errorf("IsSelectLike = %v, want %v", result.IsSelectLike, tt.wantIsSelect)
+			}
+		})
+	}
+}
+
+func TestAnalyzeSQL_SubqueryWithAggregate(t *testing.T) {
+	// Aggregate in subquery should NOT skip check for main query
+	tests := []struct {
+		name       string
+		sql        string
+		wantHasAgg bool
+	}{
+		{
+			name:       "Aggregate in subquery only",
+			sql:        "SELECT * FROM (SELECT user_id, COUNT(*) as cnt FROM orders GROUP BY user_id) t",
+			wantHasAgg: false, // top-level has no aggregate
+		},
+		{
+			name:       "Aggregate in WHERE subquery",
+			sql:        "SELECT * FROM users WHERE id IN (SELECT user_id FROM orders GROUP BY user_id HAVING COUNT(*) > 5)",
+			wantHasAgg: false, // top-level has no aggregate
+		},
+		{
+			name:       "Both top-level and subquery aggregates",
+			sql:        "SELECT COUNT(*) FROM (SELECT user_id FROM orders GROUP BY user_id) t",
+			wantHasAgg: true, // top-level has aggregate
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			result, err := AnalyzeSQL(tt.sql)
+			if err != nil {
+				t.Fatalf("AnalyzeSQL() error = %v", err)
+			}
+			if result.HasTopAgg != tt.wantHasAgg {
+				t.Errorf("HasTopAgg = %v, want %v", result.HasTopAgg, tt.wantHasAgg)
+			}
+		})
+	}
+}
+
+func TestAnalyzeSQL_LimitQueries(t *testing.T) {
+	tests := []struct {
+		name         string
+		sql          string
+		wantLimit    *int64
+		wantIsSelect bool
+	}{
+		{
+			name:         "LIMIT 10",
+			sql:          "SELECT * FROM users LIMIT 10",
+			wantLimit:    ptr(int64(10)),
+			wantIsSelect: true,
+		},
+		{
+			name:         "LIMIT 100",
+			sql:          "SELECT * FROM users LIMIT 100",
+			wantLimit:    ptr(int64(100)),
+			wantIsSelect: true,
+		},
+		{
+			name:         "LIMIT 1000",
+			sql:          "SELECT * FROM users LIMIT 1000",
+			wantLimit:    ptr(int64(1000)),
+			wantIsSelect: true,
+		},
+		{
+			name:         "LIMIT with OFFSET",
+			sql:          "SELECT * FROM users LIMIT 50 OFFSET 100",
+			wantLimit:    ptr(int64(50)),
+			wantIsSelect: true,
+		},
+		{
+			name:         "No LIMIT",
+			sql:          "SELECT * FROM users",
+			wantLimit:    nil,
+			wantIsSelect: true,
+		},
+		{
+			name:         "LIMIT 0",
+			sql:          "SELECT * FROM users LIMIT 0",
+			wantLimit:    ptr(int64(0)),
+			wantIsSelect: true,
+		},
+		{
+			name:         "LIMIT 1",
+			sql:          "SELECT * FROM users LIMIT 1",
+			wantLimit:    ptr(int64(1)),
+			wantIsSelect: true,
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			result, err := AnalyzeSQL(tt.sql)
+			if err != nil {
+				t.Fatalf("AnalyzeSQL() error = %v", err)
+			}
+			if result.IsSelectLike != tt.wantIsSelect {
+				t.Errorf("IsSelectLike = %v, want %v", result.IsSelectLike, tt.wantIsSelect)
+			}
+			if tt.wantLimit == nil {
+				if result.LimitConst != nil {
+					t.Errorf("LimitConst = %v, want nil", *result.LimitConst)
+				}
+			} else {
+				if result.LimitConst == nil {
+					t.Errorf("LimitConst = nil, want %v", *tt.wantLimit)
+				} else if *result.LimitConst != *tt.wantLimit {
+					t.Errorf("LimitConst = %v, want %v", *result.LimitConst, *tt.wantLimit)
+				}
+			}
+		})
+	}
+}
+
+func TestAnalyzeSQL_UnionQueries(t *testing.T) {
+	tests := []struct {
+		name       string
+		sql        string
+		wantHasAgg bool
+		wantLimit  *int64
+	}{
+		{
+			name:       "UNION without aggregate",
+			sql:        "SELECT id, name FROM users UNION SELECT id, name FROM admins",
+			wantHasAgg: false,
+			wantLimit:  nil,
+		},
+		{
+			name:       "UNION ALL without aggregate",
+			sql:        "SELECT * FROM users UNION ALL SELECT * FROM admins",
+			wantHasAgg: false,
+			wantLimit:  nil,
+		},
+		{
+			name:       "UNION with aggregate in all branches",
+			sql:        "SELECT COUNT(*) FROM users UNION SELECT COUNT(*) FROM admins",
+			wantHasAgg: true,
+			wantLimit:  nil,
+		},
+		{
+			name:       "UNION with aggregate in one branch only",
+			sql:        "SELECT COUNT(*) FROM users UNION SELECT id FROM admins",
+			wantHasAgg: false, // not all branches have aggregate
+			wantLimit:  nil,
+		},
+		{
+			name:       "UNION with outer LIMIT",
+			sql:        "SELECT * FROM users UNION SELECT * FROM admins LIMIT 100",
+			wantHasAgg: false,
+			wantLimit:  ptr(int64(100)),
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			result, err := AnalyzeSQL(tt.sql)
+			if err != nil {
+				t.Fatalf("AnalyzeSQL() error = %v", err)
+			}
+			if result.HasTopAgg != tt.wantHasAgg {
+				t.Errorf("HasTopAgg = %v, want %v", result.HasTopAgg, tt.wantHasAgg)
+			}
+			if tt.wantLimit == nil {
+				if result.LimitConst != nil {
+					t.Errorf("LimitConst = %v, want nil", *result.LimitConst)
+				}
+			} else {
+				if result.LimitConst == nil {
+					t.Errorf("LimitConst = nil, want %v", *tt.wantLimit)
+				} else if *result.LimitConst != *tt.wantLimit {
+					t.Errorf("LimitConst = %v, want %v", *result.LimitConst, *tt.wantLimit)
+				}
+			}
+		})
+	}
+}
+
+func TestAnalyzeSQL_NonSelectStatements(t *testing.T) {
+	tests := []struct {
+		name         string
+		sql          string
+		wantIsSelect bool
+	}{
+		{
+			name:         "SHOW DATABASES",
+			sql:          "SHOW DATABASES",
+			wantIsSelect: false,
+		},
+		{
+			name:         "SHOW TABLES",
+			sql:          "SHOW TABLES",
+			wantIsSelect: false,
+		},
+		{
+			name:         "DESCRIBE table",
+			sql:          "DESCRIBE users",
+			wantIsSelect: false,
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			result, err := AnalyzeSQL(tt.sql)
+			if err != nil {
+				// Some statements may not be parseable, which is fine
+				return
+			}
+			if result.IsSelectLike != tt.wantIsSelect {
+				t.Errorf("IsSelectLike = %v, want %v", result.IsSelectLike, tt.wantIsSelect)
+			}
+		})
+	}
+}
+
+func TestNeedsRowCountCheck(t *testing.T) {
+	maxRows := 500
+
+	tests := []struct {
+		name          string
+		sql           string
+		wantNeedCheck bool
+		wantReject    bool
+	}{
+		// Should skip check (needsCheck = false)
+		{
+			name:          "Aggregate COUNT(*)",
+			sql:           "SELECT COUNT(*) FROM users",
+			wantNeedCheck: false,
+			wantReject:    false,
+		},
+		{
+			name:          "Aggregate SUM",
+			sql:           "SELECT SUM(amount) FROM orders",
+			wantNeedCheck: false,
+			wantReject:    false,
+		},
+		{
+			name:          "Aggregate with GROUP BY",
+			sql:           "SELECT user_id, COUNT(*) FROM orders GROUP BY user_id",
+			wantNeedCheck: false,
+			wantReject:    false,
+		},
+		{
+			name:          "LIMIT equal to max",
+			sql:           "SELECT * FROM users LIMIT 500",
+			wantNeedCheck: false,
+			wantReject:    false,
+		},
+		{
+			name:          "LIMIT less than max",
+			sql:           "SELECT * FROM users LIMIT 100",
+			wantNeedCheck: false,
+			wantReject:    false,
+		},
+		{
+			name:          "LIMIT 1",
+			sql:           "SELECT * FROM users LIMIT 1",
+			wantNeedCheck: false,
+			wantReject:    false,
+		},
+
+		// LIMIT > maxRows still needs probe check (actual result might be smaller)
+		{
+			name:          "LIMIT exceeds max",
+			sql:           "SELECT * FROM users LIMIT 1000",
+			wantNeedCheck: true,
+			wantReject:    false,
+		},
+		{
+			name:          "LIMIT much larger than max",
+			sql:           "SELECT * FROM users LIMIT 10000",
+			wantNeedCheck: true,
+			wantReject:    false,
+		},
+
+		// Should execute probe check (needsCheck = true)
+		{
+			name:          "No LIMIT no aggregate",
+			sql:           "SELECT * FROM users",
+			wantNeedCheck: true,
+			wantReject:    false,
+		},
+		{
+			name:          "SELECT with WHERE no LIMIT",
+			sql:           "SELECT * FROM users WHERE status = 'active'",
+			wantNeedCheck: true,
+			wantReject:    false,
+		},
+		{
+			name:          "SELECT with JOIN no LIMIT",
+			sql:           "SELECT u.*, o.* FROM users u JOIN orders o ON u.id = o.user_id",
+			wantNeedCheck: true,
+			wantReject:    false,
+		},
+		{
+			name:          "Aggregate in subquery only",
+			sql:           "SELECT * FROM (SELECT user_id, COUNT(*) as cnt FROM orders GROUP BY user_id) t",
+			wantNeedCheck: true,
+			wantReject:    false,
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			needsCheck, directReject, _ := NeedsRowCountCheck(tt.sql, maxRows)
+			if needsCheck != tt.wantNeedCheck {
+				t.Errorf("needsCheck = %v, want %v", needsCheck, tt.wantNeedCheck)
+			}
+			if directReject != tt.wantReject {
+				t.Errorf("directReject = %v, want %v", directReject, tt.wantReject)
+			}
+		})
+	}
+}
+
+func TestNeedsRowCountCheck_DorisSpecificFunctions(t *testing.T) {
+	maxRows := 500
+
+	tests := []struct {
+		name          string
+		sql           string
+		wantNeedCheck bool
+	}{
+		// Doris HLL functions
+		{
+			name:          "HLL_UNION_AGG",
+			sql:           "SELECT HLL_UNION_AGG(hll_col) FROM user_stats",
+			wantNeedCheck: false,
+		},
+		{
+			name:          "HLL_CARDINALITY",
+			sql:           "SELECT HLL_CARDINALITY(hll_col) FROM user_stats",
+			wantNeedCheck: false,
+		},
+		// Doris Bitmap functions
+		{
+			name:          "BITMAP_UNION_COUNT",
+			sql:           "SELECT BITMAP_UNION_COUNT(bitmap_col) FROM user_tags",
+			wantNeedCheck: false,
+		},
+		{
+			name:          "BITMAP_UNION",
+			sql:           "SELECT BITMAP_UNION(bitmap_col) FROM user_tags GROUP BY category",
+			wantNeedCheck: false,
+		},
+		// Other Doris aggregate functions
+		{
+			name:          "APPROX_COUNT_DISTINCT",
+			sql:           "SELECT APPROX_COUNT_DISTINCT(user_id) FROM events",
+			wantNeedCheck: false,
+		},
+		{
+			name:          "GROUP_CONCAT",
+			sql:           "SELECT GROUP_CONCAT(name) FROM users GROUP BY department",
+			wantNeedCheck: false,
+		},
+		{
+			name:          "PERCENTILE_APPROX",
+			sql:           "SELECT PERCENTILE_APPROX(latency, 0.99) FROM requests",
+			wantNeedCheck: false,
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			needsCheck, _, _ := NeedsRowCountCheck(tt.sql, maxRows)
+			if needsCheck != tt.wantNeedCheck {
+				t.Errorf("needsCheck = %v, want %v (should skip check for Doris aggregate functions)", needsCheck, tt.wantNeedCheck)
+			}
+		})
+	}
+}
+
+func TestNeedsRowCountCheck_ComplexQueries(t *testing.T) {
+	maxRows := 500
+
+	tests := []struct {
+		name          string
+		sql           string
+		wantNeedCheck bool
+		wantReject    bool
+	}{
+		{
+			name:          "CTE with aggregate",
+			sql:           "WITH user_counts AS (SELECT user_id, COUNT(*) as cnt FROM orders GROUP BY user_id) SELECT * FROM user_counts",
+			wantNeedCheck: false, // CTE has aggregate, skip check
+			wantReject:    false,
+		},
+		{
+			name:          "Complex JOIN with aggregate",
+			sql:           "SELECT u.department, COUNT(*) FROM users u JOIN orders o ON u.id = o.user_id GROUP BY u.department",
+			wantNeedCheck: false, // has aggregate
+			wantReject:    false,
+		},
+		{
+			name:          "Nested subquery",
+			sql:           "SELECT * FROM users WHERE id IN (SELECT user_id FROM orders WHERE amount > 100)",
+			wantNeedCheck: true,
+			wantReject:    false,
+		},
+		{
+			name:          "DISTINCT query",
+			sql:           "SELECT DISTINCT category FROM products",
+			wantNeedCheck: true, // DISTINCT is not aggregate
+			wantReject:    false,
+		},
+		{
+			name:          "ORDER BY with LIMIT",
+			sql:           "SELECT * FROM users ORDER BY created_at DESC LIMIT 100",
+			wantNeedCheck: false, // has valid LIMIT
+			wantReject:    false,
+		},
+		{
+			name:          "Multiple aggregates in single query",
+			sql:           "SELECT COUNT(*), SUM(amount), AVG(amount), MIN(amount), MAX(amount) FROM orders",
+			wantNeedCheck: false,
+			wantReject:    false,
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			needsCheck, directReject, _ := NeedsRowCountCheck(tt.sql, maxRows)
+			if needsCheck != tt.wantNeedCheck {
+				t.Errorf("needsCheck = %v, want %v", needsCheck, tt.wantNeedCheck)
+			}
+			if directReject != tt.wantReject {
+				t.Errorf("directReject = %v, want %v", directReject, tt.wantReject)
+			}
+		})
+	}
+}
+
+func TestNeedsRowCountCheck_EdgeCases(t *testing.T) {
+	maxRows := 500
+
+	tests := []struct {
+		name          string
+		sql           string
+		wantNeedCheck bool
+		wantReject    bool
+	}{
+		{
+			name:          "Empty-ish LIMIT 0",
+			sql:           "SELECT * FROM users LIMIT 0",
+			wantNeedCheck: false,
+			wantReject:    false,
+		},
+		{
+			name:          "LIMIT at boundary",
+			sql:           "SELECT * FROM users LIMIT 501",
+			wantNeedCheck: true, // 501 > 500, needs probe check
+			wantReject:    false,
+		},
+		{
+			name:          "SELECT with trailing semicolon",
+			sql:           "SELECT * FROM users;",
+			wantNeedCheck: true,
+			wantReject:    false,
+		},
+		{
+			name:          "SELECT with extra whitespace",
+			sql:           "  SELECT * FROM users  ",
+			wantNeedCheck: true,
+			wantReject:    false,
+		},
+		{
+			name:          "Lowercase keywords",
+			sql:           "select count(*) from users",
+			wantNeedCheck: false,
+			wantReject:    false,
+		},
+		{
+			name:          "Mixed case keywords",
+			sql:           "Select Count(*) From users",
+			wantNeedCheck: false,
+			wantReject:    false,
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			needsCheck, directReject, _ := NeedsRowCountCheck(tt.sql, maxRows)
+			if needsCheck != tt.wantNeedCheck {
+				t.Errorf("needsCheck = %v, want %v", needsCheck, tt.wantNeedCheck)
+			}
+			if directReject != tt.wantReject {
+				t.Errorf("directReject = %v, want %v", directReject, tt.wantReject)
+			}
+		})
+	}
+}
+
+func TestNeedsRowCountCheck_DifferentMaxRows(t *testing.T) {
+	tests := []struct {
+		name          string
+		sql           string
+		maxRows       int
+		wantNeedCheck bool
+		wantReject    bool
+	}{
+		{
+			name:          "LIMIT 100 with maxRows 50",
+			sql:           "SELECT * FROM users LIMIT 100",
+			maxRows:       50,
+			wantNeedCheck: true, // LIMIT > maxRows, needs probe check
+			wantReject:    false,
+		},
+		{
+			name:          "LIMIT 100 with maxRows 100",
+			sql:           "SELECT * FROM users LIMIT 100",
+			maxRows:       100,
+			wantNeedCheck: false,
+			wantReject:    false,
+		},
+		{
+			name:          "LIMIT 100 with maxRows 200",
+			sql:           "SELECT * FROM users LIMIT 100",
+			maxRows:       200,
+			wantNeedCheck: false,
+			wantReject:    false,
+		},
+		{
+			name:          "No LIMIT with maxRows 1000",
+			sql:           "SELECT * FROM users",
+			maxRows:       1000,
+			wantNeedCheck: true,
+			wantReject:    false,
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			needsCheck, directReject, _ := NeedsRowCountCheck(tt.sql, tt.maxRows)
+			if needsCheck != tt.wantNeedCheck {
+				t.Errorf("needsCheck = %v, want %v", needsCheck, tt.wantNeedCheck)
+			}
+			if directReject != tt.wantReject {
+				t.Errorf("directReject = %v, want %v", directReject, tt.wantReject)
+			}
+		})
+	}
+}
+
+// TestSummary_SkipProbeCheck prints a summary of which SQL patterns skip the probe check
+func TestSummary_SkipProbeCheck(t *testing.T) {
+	maxRows := 500
+
+	skipCheckCases := []struct {
+		category string
+		sql      string
+	}{
+		// Aggregate functions
+		{"Aggregate - COUNT(*)", "SELECT COUNT(*) FROM users"},
+		{"Aggregate - COUNT(col)", "SELECT COUNT(id) FROM users"},
+		{"Aggregate - SUM", "SELECT SUM(amount) FROM orders"},
+		{"Aggregate - AVG", "SELECT AVG(price) FROM products"},
+		{"Aggregate - MIN", "SELECT MIN(created_at) FROM logs"},
+		{"Aggregate - MAX", "SELECT MAX(score) FROM results"},
+		{"Aggregate - GROUP BY", "SELECT user_id, COUNT(*) FROM orders GROUP BY user_id"},
+		{"Aggregate - HAVING", "SELECT user_id, SUM(amount) FROM orders GROUP BY user_id HAVING SUM(amount) > 1000"},
+
+		// Doris specific aggregates
+		{"Doris - HLL_UNION_AGG", "SELECT HLL_UNION_AGG(hll_col) FROM stats"},
+		{"Doris - BITMAP_UNION_COUNT", "SELECT BITMAP_UNION_COUNT(bitmap_col) FROM tags"},
+		{"Doris - APPROX_COUNT_DISTINCT", "SELECT APPROX_COUNT_DISTINCT(user_id) FROM events"},
+		{"Doris - GROUP_CONCAT", "SELECT GROUP_CONCAT(name) FROM users GROUP BY dept"},
+
+		// LIMIT <= maxRows
+		{"LIMIT - Equal to max", "SELECT * FROM users LIMIT 500"},
+		{"LIMIT - Less than max", "SELECT * FROM users LIMIT 100"},
+		{"LIMIT - With OFFSET", "SELECT * FROM users LIMIT 100 OFFSET 50"},
+		{"LIMIT - Value 1", "SELECT * FROM users LIMIT 1"},
+		{"LIMIT - Value 0", "SELECT * FROM users LIMIT 0"},
+	}
+
+	t.Log("=== SQL patterns that SKIP probe check (no extra query needed) ===")
+	for _, tc := range skipCheckCases {
+		needsCheck, _, _ := NeedsRowCountCheck(tc.sql, maxRows)
+		status := "✓ SKIP"
+		if needsCheck {
+			status = "✗ NEEDS CHECK (unexpected)"
+		}
+		t.Logf("  %s: %s\n    SQL: %s", status, tc.category, tc.sql)
+	}
+
+	needsCheckCases := []struct {
+		category string
+		sql      string
+	}{
+		{"No LIMIT - Simple SELECT", "SELECT * FROM users"},
+		{"No LIMIT - With WHERE", "SELECT * FROM users WHERE status = 'active'"},
+		{"No LIMIT - With JOIN", "SELECT u.*, o.* FROM users u JOIN orders o ON u.id = o.user_id"},
+		{"No LIMIT - Subquery with agg", "SELECT * FROM (SELECT user_id, COUNT(*) FROM orders GROUP BY user_id) t"},
+		{"No LIMIT - DISTINCT", "SELECT DISTINCT category FROM products"},
+		{"LIMIT > max (actual may be smaller)", "SELECT * FROM users LIMIT 1000"},
+		{"LIMIT >> max", "SELECT * FROM users LIMIT 10000"},
+	}
+
+	t.Log("\n=== SQL patterns that NEED probe check ===")
+	for _, tc := range needsCheckCases {
+		needsCheck, _, _ := NeedsRowCountCheck(tc.sql, maxRows)
+		status := "✓ NEEDS CHECK"
+		if !needsCheck {
+			status = "✗ SKIP (unexpected)"
+		}
+		t.Logf("  %s: %s\n    SQL: %s", status, tc.category, tc.sql)
+	}
+}
+
+// ptr is a helper function to create a pointer to int64
+func ptr(v int64) *int64 {
+	return &v
+}

dskit/doris/timeseries.go

@@ -15,6 +15,10 @@ const (
 	TimeFieldFormatDateTime    = "datetime"
 )
 
+type noNeedCheckMaxRowKey struct{}
+
+var NoNeedCheckMaxRow = noNeedCheckMaxRowKey{}
+
 // 不再拼接SQL, 完全信赖用户的输入
 type QueryParam struct {
 	Database string     `json:"database"`
@@ -39,7 +43,7 @@ var (
 )
 
 // Query executes a given SQL query in Doris and returns the results with MaxQueryRows check
-func (d *Doris) Query(ctx context.Context, query *QueryParam) ([]map[string]interface{}, error) {
+func (d *Doris) Query(ctx context.Context, query *QueryParam, checkMaxRow bool) ([]map[string]interface{}, error) {
 	// 校验SQL的合法性, 过滤掉 write请求
 	sqlItem := strings.Split(strings.ToUpper(query.Sql), " ")
 	for _, item := range sqlItem {
@@ -48,10 +52,12 @@ func (d *Doris) Query(ctx context.Context, query *QueryParam) ([]map[string]inte
 		}
 	}
 
-	// 检查查询结果行数
-	err := d.CheckMaxQueryRows(ctx, query.Database, query.Sql)
-	if err != nil {
-		return nil, err
+	if checkMaxRow {
+		// 检查查询结果行数
+		err := d.CheckMaxQueryRows(ctx, query.Database, query.Sql)
+		if err != nil {
+			return nil, err
+		}
 	}
 
 	rows, err := d.ExecQuery(ctx, query.Database, query.Sql)
@@ -63,8 +69,12 @@ func (d *Doris) Query(ctx context.Context, query *QueryParam) ([]map[string]inte
 
 // QueryTimeseries executes a time series data query using the given parameters with MaxQueryRows check
 func (d *Doris) QueryTimeseries(ctx context.Context, query *QueryParam) ([]types.MetricValues, error) {
-	// 使用 Query 方法执行查询，Query方法内部已包含MaxQueryRows检查
-	rows, err := d.Query(ctx, query)
+	// 默认需要检查，除非调用方声明不需要检查
+	checkMaxRow := true
+	if noCheck, ok := ctx.Value(NoNeedCheckMaxRow).(bool); ok && noCheck {
+		checkMaxRow = false
+	}
+	rows, err := d.Query(ctx, query, checkMaxRow)
 	if err != nil {
 		return nil, err
 	}
@@ -73,35 +83,44 @@ func (d *Doris) QueryTimeseries(ctx context.Context, query *QueryParam) ([]types
 }
 
 // CheckMaxQueryRows checks if the query result exceeds the maximum allowed rows
+// It uses SQL analysis to skip unnecessary checks for aggregate queries or queries with LIMIT <= maxRows
+// For queries that need checking, it uses probe approach (LIMIT maxRows+1) instead of COUNT(*) for better performance
 func (d *Doris) CheckMaxQueryRows(ctx context.Context, database, sql string) error {
+	maxQueryRows := d.MaxQueryRows
+	if maxQueryRows == 0 {
+		maxQueryRows = 500
+	}
+
+	cleanedSQL := strings.TrimSpace(strings.TrimSuffix(strings.TrimSpace(sql), ";"))
+
+	// Step 1: Analyze SQL to determine if check is needed
+	needsCheck, _, _ := NeedsRowCountCheck(cleanedSQL, maxQueryRows)
+	if !needsCheck {
+		return nil
+	}
+
+	// Step 2: Execute probe query (more efficient than COUNT(*))
+	return d.probeRowCount(ctx, database, cleanedSQL, maxQueryRows)
+}
+
+// probeRowCount uses threshold probing to check row count
+// It reads at most maxRows+1 rows, which is O(maxRows) instead of O(totalRows) for COUNT(*)
+// Doris optimizes LIMIT queries by stopping scan early once limit is reached
+func (d *Doris) probeRowCount(ctx context.Context, database, sql string, maxRows int) error {
 	timeoutCtx, cancel := d.createTimeoutContext(ctx)
 	defer cancel()
 
-	cleanedSQL := strings.ReplaceAll(sql, ";", "")
-	checkQuery := fmt.Sprintf("SELECT COUNT(*) as count FROM (%s) AS subquery;", cleanedSQL)
+	// Probe SQL: only need to check if exceeds threshold, not actual data
+	probeSQL := fmt.Sprintf("SELECT 1 FROM (%s) AS __probe_chk LIMIT %d", sql, maxRows+1)
 
-	// 执行计数查询
-	results, err := d.ExecQuery(timeoutCtx, database, checkQuery)
+	results, err := d.ExecQuery(timeoutCtx, database, probeSQL)
 	if err != nil {
 		return err
 	}
 
-	if len(results) > 0 {
-		if count, exists := results[0]["count"]; exists {
-			v, err := sqlbase.ParseFloat64Value(count)
-			if err != nil {
-				return err
-			}
-
-			maxQueryRows := d.MaxQueryRows
-			if maxQueryRows == 0 {
-				maxQueryRows = 500
-			}
-
-			if v > float64(maxQueryRows) {
-				return fmt.Errorf("query result rows count %d exceeds the maximum limit %d", int(v), maxQueryRows)
-			}
-		}
+	// If returned rows > maxRows, it exceeds the limit
+	if len(results) > maxRows {
+		return fmt.Errorf("query result rows count exceeds the maximum limit %d", maxRows)
 	}
 
 	return nil

go.mod

@@ -33,6 +33,7 @@ require (
 	github.com/jinzhu/copier v0.4.0
 	github.com/json-iterator/go v1.1.12
 	github.com/koding/multiconfig v0.0.0-20171124222453-69c27309b2d7
+	github.com/larksuite/oapi-sdk-go/v3 v3.5.1
 	github.com/lib/pq v1.10.9
 	github.com/mailru/easyjson v0.7.7
 	github.com/mattn/go-isatty v0.0.19
@@ -42,6 +43,7 @@ require (
 	github.com/opensearch-project/opensearch-go/v2 v2.3.0
 	github.com/patrickmn/go-cache v2.1.0+incompatible
 	github.com/pelletier/go-toml/v2 v2.0.8
+	github.com/pingcap/tidb/pkg/parser v0.0.0-20260120034856-e15515e804da
 	github.com/pkg/errors v0.9.1
 	github.com/prometheus/client_golang v1.20.5
 	github.com/prometheus/common v0.60.1
@@ -101,6 +103,9 @@ require (
 	github.com/jcmturner/gofork v1.7.6 // indirect
 	github.com/jcmturner/gokrb5/v8 v8.4.4 // indirect
 	github.com/jcmturner/rpc/v2 v2.0.3 // indirect
+	github.com/pingcap/errors v0.11.5-0.20250523034308-74f78ae071ee // indirect
+	github.com/pingcap/failpoint v0.0.0-20240528011301-b51a646c7c86 // indirect
+	github.com/pingcap/log v1.1.0 // indirect
 	github.com/rcrowley/go-metrics v0.0.0-20201227073835-cf1acfcdf475 // indirect
 	github.com/remyoudompheng/bigfft v0.0.0-20230129092748-24d4a6f8daec // indirect
 	github.com/rogpeppe/go-internal v1.13.1 // indirect
@@ -108,10 +113,13 @@ require (
 	github.com/valyala/fastrand v1.1.0 // indirect
 	github.com/valyala/histogram v1.2.0 // indirect
 	github.com/yuin/gopher-lua v1.1.1 // indirect
+	go.uber.org/multierr v1.11.0 // indirect
+	go.uber.org/zap v1.27.0 // indirect
 	golang.org/x/sync v0.18.0 // indirect
 	gopkg.in/ini.v1 v1.67.0 // indirect
+	gopkg.in/natefinch/lumberjack.v2 v2.2.1 // indirect
 	modernc.org/libc v1.22.5 // indirect
-	modernc.org/mathutil v1.5.0 // indirect
+	modernc.org/mathutil v1.6.0 // indirect
 	modernc.org/memory v1.5.0 // indirect
 	modernc.org/sqlite v1.23.1 // indirect
 )
@@ -135,7 +143,7 @@ require (
 	github.com/go-playground/locales v0.14.1 // indirect
 	github.com/go-playground/universal-translator v0.18.1 // indirect
 	github.com/go-playground/validator/v10 v10.14.0 // indirect
-	github.com/go-sql-driver/mysql v1.6.0
+	github.com/go-sql-driver/mysql v1.7.1
 	github.com/goccy/go-json v0.10.2 // indirect
 	github.com/golang/freetype v0.0.0-20170609003504-e2365dfdc4a0 // indirect
 	github.com/grafana/regexp v0.0.0-20221122212121-6b5c0a4cb7fd // indirect

go.sum

@@ -101,6 +101,7 @@ github.com/aws/aws-sdk-go-v2/service/sso v1.12.10/go.mod h1:ouy2P4z6sJN70fR3ka3w
 github.com/aws/aws-sdk-go-v2/service/ssooidc v1.14.10/go.mod h1:AFvkxc8xfBe8XA+5St5XIHHrQQtkxqrRincx4hmMHOk=
 github.com/aws/aws-sdk-go-v2/service/sts v1.19.0/go.mod h1:BgQOMsg8av8jset59jelyPW7NoZcZXLVpDsXunGDrk8=
 github.com/aws/smithy-go v1.13.5/go.mod h1:Tg+OJXh4MB2R/uN61Ko2f6hTZwB/ZYGOtib8J3gBHzA=
+github.com/benbjohnson/clock v1.1.0/go.mod h1:J11/hYXuz8f4ySSvYwY0FKfm+ezbsZBKZxNJlLklBHA=
 github.com/beorn7/perks v1.0.1 h1:VlbKKnNfV8bJzeqoa4cOKqO6bYr3WgKZxO8Z16+hsOM=
 github.com/beorn7/perks v1.0.1/go.mod h1:G2ZrVWU2WbWT9wwq4/hrbKbnv/1ERSJQ0ibhJ6rlkpw=
 github.com/bitly/go-simplejson v0.5.1 h1:xgwPbetQScXt1gh9BmoJ6j9JMr3TElvuIyjR8pgdoow=
@@ -195,8 +196,9 @@ github.com/go-playground/universal-translator v0.18.1/go.mod h1:xekY+UJKNuX9WP91
 github.com/go-playground/validator/v10 v10.10.0/go.mod h1:74x4gJWsvQexRdW8Pn3dXSGrTK4nAUsbPlLADvpJkos=
 github.com/go-playground/validator/v10 v10.14.0 h1:vgvQWe3XCz3gIeFDm/HnTIbj6UGmg/+t63MyGU2n5js=
 github.com/go-playground/validator/v10 v10.14.0/go.mod h1:9iXMNT7sEkjXb0I+enO7QXmzG6QCsPWY4zveKFVRSyU=
-github.com/go-sql-driver/mysql v1.6.0 h1:BCTh4TKNUYmOmMUcQ3IipzF5prigylS7XXjEkfCHuOE=
 github.com/go-sql-driver/mysql v1.6.0/go.mod h1:DCzpHaOWr8IXmIStZouvnhqoel9Qv2LBy8hT2VhHyBg=
+github.com/go-sql-driver/mysql v1.7.1 h1:lUIinVbN1DY0xBg0eMOzmmtGoHwWBbvnWubQUrtU8EI=
+github.com/go-sql-driver/mysql v1.7.1/go.mod h1:OXbVy3sEdcQ2Doequ6Z5BW6fXNQTmx+9S1MCJN5yJMI=
 github.com/goccy/go-json v0.9.7/go.mod h1:6MelG93GURQebXPDq3khkgXZkazVtN9CRI+MGFi0w8I=
 github.com/goccy/go-json v0.10.2 h1:CrxCmQqYDkv1z7lO7Wbh2HN93uovUHgrECaO5ZrCXAU=
 github.com/goccy/go-json v0.10.2/go.mod h1:6MelG93GURQebXPDq3khkgXZkazVtN9CRI+MGFi0w8I=
@@ -243,6 +245,7 @@ github.com/gopherjs/gopherjs v0.0.0-20181017120253-0766667cb4d1/go.mod h1:wJfORR
 github.com/gopherjs/gopherjs v0.0.0-20200217142428-fce0ec30dd00/go.mod h1:wJfORRmW1u3UXTncJ5qlYoELFm8eSnnEO6hX4iZ3EWY=
 github.com/gorilla/securecookie v1.1.1/go.mod h1:ra0sb63/xPlUeL+yeDciTfxMRAA+MP+HVt/4epWDjd4=
 github.com/gorilla/sessions v1.2.1/go.mod h1:dk2InVEVJ0sfLlnXv9EAgkf6ecYs/i80K/zI+bUmuGM=
+github.com/gorilla/websocket v1.5.0/go.mod h1:YR8l580nyteQvAITg2hZ9XVh4b55+EU/adAjf1fMHhE=
 github.com/grafana/regexp v0.0.0-20221122212121-6b5c0a4cb7fd h1:PpuIBO5P3e9hpqBD0O/HjhShYuM6XE0i/lbE6J94kww=
 github.com/grafana/regexp v0.0.0-20221122212121-6b5c0a4cb7fd/go.mod h1:M5qHK+eWfAv8VR/265dIuEpL3fNfeC21tXXp9itM24A=
 github.com/hashicorp/errwrap v1.0.0/go.mod h1:YH+1FKiLXxHSkmPseP+kNlulaMuP3n2brvKWEqk/Jc4=
@@ -315,6 +318,8 @@ github.com/kr/text v0.2.0 h1:5Nx0Ya0ZqY2ygV366QzturHI13Jq95ApcVaJBhpS+AY=
 github.com/kr/text v0.2.0/go.mod h1:eLer722TekiGuMkidMxC/pM04lWEeraHUUmBw8l2grE=
 github.com/kylelemons/godebug v1.1.0 h1:RPNrshWIDI6G2gRW9EHilWtl7Z6Sb1BR0xunSBf0SNc=
 github.com/kylelemons/godebug v1.1.0/go.mod h1:9/0rRGxNHcop5bhtWyNeEfOS8JIWk580+fNqagV/RAw=
+github.com/larksuite/oapi-sdk-go/v3 v3.5.1 h1:gX4dz92YU70inuIX+ug+PBe64eHToIN9rHB4Vupv5Eg=
+github.com/larksuite/oapi-sdk-go/v3 v3.5.1/go.mod h1:ZEplY+kwuIrj/nqw5uSCINNATcH3KdxSN7y+UxYY5fI=
 github.com/leodido/go-urn v1.2.1/go.mod h1:zt4jvISO2HfUBqxjfIshjdMTYS56ZS/qv49ictyFfxY=
 github.com/leodido/go-urn v1.2.4 h1:XlAE/cm/ms7TE/VMVoduSpNBoyc2dOxHs5MZSwAN63Q=
 github.com/leodido/go-urn v1.2.4/go.mod h1:7ZrI8mTSeBSHl/UaRyKQW1qZeMgak41ANeCNaVckg+4=
@@ -361,9 +366,19 @@ github.com/pelletier/go-toml/v2 v2.0.8 h1:0ctb6s9mE31h0/lhu+J6OPmVeDxJn+kYnJc2jZ
 github.com/pelletier/go-toml/v2 v2.0.8/go.mod h1:vuYfssBdrU2XDZ9bYydBu6t+6a6PYNcZljzZR9VXg+4=
 github.com/pierrec/lz4/v4 v4.1.22 h1:cKFw6uJDK+/gfw5BcDL0JL5aBsAFdsIT18eRtLj7VIU=
 github.com/pierrec/lz4/v4 v4.1.22/go.mod h1:gZWDp/Ze/IJXGXf23ltt2EXimqmTUXEy0GFuRQyBid4=
+github.com/pingcap/errors v0.11.0/go.mod h1:Oi8TUi2kEtXXLMJk9l1cGmz20kV3TaQ0usTwv5KuLY8=
+github.com/pingcap/errors v0.11.5-0.20250523034308-74f78ae071ee h1:/IDPbpzkzA97t1/Z1+C3KlxbevjMeaI6BQYxvivu4u8=
+github.com/pingcap/errors v0.11.5-0.20250523034308-74f78ae071ee/go.mod h1:X2r9ueLEUZgtx2cIogM0v4Zj5uvvzhuuiu7Pn8HzMPg=
+github.com/pingcap/failpoint v0.0.0-20240528011301-b51a646c7c86 h1:tdMsjOqUR7YXHoBitzdebTvOjs/swniBTOLy5XiMtuE=
+github.com/pingcap/failpoint v0.0.0-20240528011301-b51a646c7c86/go.mod h1:exzhVYca3WRtd6gclGNErRWb1qEgff3LYta0LvRmON4=
+github.com/pingcap/log v1.1.0 h1:ELiPxACz7vdo1qAvvaWJg1NrYFoY6gqAh/+Uo6aXdD8=
+github.com/pingcap/log v1.1.0/go.mod h1:DWQW5jICDR7UJh4HtxXSM20Churx4CQL0fwL/SoOSA4=
+github.com/pingcap/tidb/pkg/parser v0.0.0-20260120034856-e15515e804da h1:PhkRZgMWdq9kTsu7vtVbcDs+SBXjHfFj84027WVZCzI=
+github.com/pingcap/tidb/pkg/parser v0.0.0-20260120034856-e15515e804da/go.mod h1:oHE+ub2QaDERd+UNHe4z2BhFV2jZrm7VNOe6atR9AF4=
 github.com/pkg/browser v0.0.0-20210911075715-681adbf594b8 h1:KoWmjvw+nsYOo29YJK9vDA65RGE3NrOnUtO7a+RF9HU=
 github.com/pkg/browser v0.0.0-20210911075715-681adbf594b8/go.mod h1:HKlIX3XHQyzLZPlr7++PzdhaXEj94dEiJgZDTsxEqUI=
 github.com/pkg/diff v0.0.0-20210226163009-20ebb0f2a09e/go.mod h1:pJLUxLENpZxwdsKMEsNbx1VGcRFpLqf3715MtcvvzbA=
+github.com/pkg/errors v0.8.1/go.mod h1:bwawxfHBFNV+L2hUp1rHADufV3IMtnDRdf1r5NINEl0=
 github.com/pkg/errors v0.9.1 h1:FEBLx1zS214owpjy7qsBeixbURkuhQAwrK5UwLGTwt4=
 github.com/pkg/errors v0.9.1/go.mod h1:bwawxfHBFNV+L2hUp1rHADufV3IMtnDRdf1r5NINEl0=
 github.com/pmezard/go-difflib v1.0.0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZNVY4sRDYZ/4=
@@ -392,7 +407,6 @@ github.com/rcrowley/go-metrics v0.0.0-20201227073835-cf1acfcdf475 h1:N/ElC8H3+5X
 github.com/rcrowley/go-metrics v0.0.0-20201227073835-cf1acfcdf475/go.mod h1:bCqnVzQkZxMG4s8nGwiZ5l3QUCyqpo9Y+/ZMZ9VjZe4=
 github.com/redis/go-redis/v9 v9.0.2 h1:BA426Zqe/7r56kCcvxYLWe1mkaz71LKF77GwgFzSxfE=
 github.com/redis/go-redis/v9 v9.0.2/go.mod h1:/xDTe9EF1LM61hek62Poq2nzQSGj0xSrEtEHbBQevps=
-github.com/remyoudompheng/bigfft v0.0.0-20200410134404-eec4a21b6bb0/go.mod h1:qqbHyh8v60DhA7CoWK5oRCqLrMHRGoxYCSS9EjAz6Eo=
 github.com/remyoudompheng/bigfft v0.0.0-20230129092748-24d4a6f8daec h1:W09IVJc94icq4NjY3clb7Lk8O1qJ8BdBEF8z0ibU0rE=
 github.com/remyoudompheng/bigfft v0.0.0-20230129092748-24d4a6f8daec/go.mod h1:qqbHyh8v60DhA7CoWK5oRCqLrMHRGoxYCSS9EjAz6Eo=
 github.com/rivo/uniseg v0.1.0/go.mod h1:J6wj4VEh+S6ZtnVlnTBMWIodfgj8LQOQFoIToxlJtxc=
@@ -467,13 +481,24 @@ go.opentelemetry.io/otel v1.32.0 h1:WnBN+Xjcteh0zdk01SVqV55d/m62NJLJdIyb4y/WO5U=
 go.opentelemetry.io/otel v1.32.0/go.mod h1:00DCVSB0RQcnzlwyTfqtxSm+DRr9hpYrHjNGiBHVQIg=
 go.opentelemetry.io/otel/trace v1.32.0 h1:WIC9mYrXf8TmY/EXuULKc8hR17vE+Hjv2cssQDe03fM=
 go.opentelemetry.io/otel/trace v1.32.0/go.mod h1:+i4rkvCraA+tG6AzwloGaCtkx53Fa+L+V8e9a7YvhT8=
+go.uber.org/atomic v1.6.0/go.mod h1:sABNBOSYdrvTF6hTgEIbc7YasKWGhgEQZyfxyTvoXHQ=
+go.uber.org/atomic v1.7.0/go.mod h1:fEN4uk6kAWBTFdckzkM89CLk9XfWZrxpCo0nPH17wJc=
+go.uber.org/atomic v1.9.0/go.mod h1:fEN4uk6kAWBTFdckzkM89CLk9XfWZrxpCo0nPH17wJc=
 go.uber.org/atomic v1.11.0 h1:ZvwS0R+56ePWxUNi+Atn9dWONBPp/AUETXlHW0DxSjE=
 go.uber.org/atomic v1.11.0/go.mod h1:LUxbIzbOniOlMKjJjyPfpl4v+PKK2cNJn91OQbhoJI0=
 go.uber.org/automaxprocs v1.4.0/go.mod h1:/mTEdr7LvHhs0v7mjdxDreTz1OG5zdZGqgOnhWiR/+Q=
 go.uber.org/automaxprocs v1.5.2 h1:2LxUOGiR3O6tw8ui5sZa2LAaHnsviZdVOUZw4fvbnME=
 go.uber.org/automaxprocs v1.5.2/go.mod h1:eRbA25aqJrxAbsLO0xy5jVwPt7FQnRgjW+efnwa1WM0=
+go.uber.org/goleak v1.1.10/go.mod h1:8a7PlsEVH3e/a/GLqe5IIrQx6GzcnRmZEufDUTk4A7A=
 go.uber.org/goleak v1.3.0 h1:2K3zAYmnTNqV73imy9J1T3WC+gmCePx2hEGkimedGto=
 go.uber.org/goleak v1.3.0/go.mod h1:CoHD4mav9JJNrW/WLlf7HGZPjdw8EucARQHekz1X6bE=
+go.uber.org/multierr v1.6.0/go.mod h1:cdWPpRnG4AhwMwsgIHip0KRBQjJy5kYEpYjJxpXp9iU=
+go.uber.org/multierr v1.7.0/go.mod h1:7EAYxJLBy9rStEaz58O2t4Uvip6FSURkq8/ppBp95ak=
+go.uber.org/multierr v1.11.0 h1:blXXJkSxSSfBVBlC76pxqeO+LN3aDfLQo+309xJstO0=
+go.uber.org/multierr v1.11.0/go.mod h1:20+QtiLqy0Nd6FdQB9TLXag12DsQkrbs3htMFfDN80Y=
+go.uber.org/zap v1.19.0/go.mod h1:xg/QME4nWcxGxrpdeYfq7UvYrLh66cuVKdrbD1XF/NI=
+go.uber.org/zap v1.27.0 h1:aJMhYGrd5QSmlpLMr2MftRKl7t8J8PTZPA732ud/XR8=
+go.uber.org/zap v1.27.0/go.mod h1:GB2qFLM7cTU87MWRP2mPIjqfIDnGu+VIO4V/SdhGo2E=
 golang.org/x/arch v0.0.0-20210923205945-b76863e36670/go.mod h1:5om86z9Hs0C8fWVUuoMHwpExlXzs5Tkyp9hOrfG7pp8=
 golang.org/x/arch v0.3.0 h1:02VY4/ZcO/gBOH6PUaoiptASxtXU10jazRCP865E97k=
 golang.org/x/arch v0.3.0/go.mod h1:5om86z9Hs0C8fWVUuoMHwpExlXzs5Tkyp9hOrfG7pp8=
@@ -507,6 +532,7 @@ golang.org/x/image v0.18.0/go.mod h1:4yyo5vMFQjVjUcVk4jEQcU9MGy/rulF5WvUILseCM2E
 golang.org/x/lint v0.0.0-20181026193005-c67002cb31c3/go.mod h1:UVdnD1Gm6xHRNCYTkRU2/jEulfH38KcIWyp/GAMgvoE=
 golang.org/x/lint v0.0.0-20190227174305-5b3e6a55c961/go.mod h1:wehouNa3lNwaWXcvxsM5YxQ5yQlVC4a0KAMCusXpPoU=
 golang.org/x/lint v0.0.0-20190313153728-d0100b6bd8b3/go.mod h1:6SW0HCj/g11FgYtHlgUYUwCkIfeOF89ocIRzGO/8vkc=
+golang.org/x/lint v0.0.0-20190930215403-16217165b5de/go.mod h1:6SW0HCj/g11FgYtHlgUYUwCkIfeOF89ocIRzGO/8vkc=
 golang.org/x/mod v0.2.0/go.mod h1:s0Qsj1ACt9ePp/hMypM3fl4fZqREWJwdYDEqhRiZZUA=
 golang.org/x/mod v0.3.0/go.mod h1:s0Qsj1ACt9ePp/hMypM3fl4fZqREWJwdYDEqhRiZZUA=
 golang.org/x/mod v0.6.0-dev.0.20220419223038-86c51ed26bb4/go.mod h1:jJ57K6gSWd91VN4djpZkiMVwK6gcyfeH4XE8wZrZaV4=
@@ -623,6 +649,8 @@ golang.org/x/tools v0.0.0-20190226205152-f727befe758c/go.mod h1:9Yl7xja0Znq3iFh3
 golang.org/x/tools v0.0.0-20190311212946-11955173bddd/go.mod h1:LCzVGOaR6xXOjkQ3onu1FJEFr0SW1gC7cKk1uF8kGRs=
 golang.org/x/tools v0.0.0-20190328211700-ab21143f2384/go.mod h1:LCzVGOaR6xXOjkQ3onu1FJEFr0SW1gC7cKk1uF8kGRs=
 golang.org/x/tools v0.0.0-20190524140312-2c0ae7006135/go.mod h1:RgjU9mgBXZiqYHBnxXauZ1Gv1EHHAz9KjViQ78xBX0Q=
+golang.org/x/tools v0.0.0-20191029041327-9cc4af7d6b2c/go.mod h1:b+2E5dAYhXwXZwtnZ6UAqBI28+e2cm9otk0dWdXHAEo=
+golang.org/x/tools v0.0.0-20191108193012-7d206e10da11/go.mod h1:b+2E5dAYhXwXZwtnZ6UAqBI28+e2cm9otk0dWdXHAEo=
 golang.org/x/tools v0.0.0-20191119224855-298f0cb1881e/go.mod h1:b+2E5dAYhXwXZwtnZ6UAqBI28+e2cm9otk0dWdXHAEo=
 golang.org/x/tools v0.0.0-20200509030707-2212a7e161a5/go.mod h1:EkVYQZoAsY45+roYkvgYkIh4xh/qjgUK9TdY2XT94GE=
 golang.org/x/tools v0.0.0-20200619180055-7c47624df98f/go.mod h1:EkVYQZoAsY45+roYkvgYkIh4xh/qjgUK9TdY2XT94GE=
@@ -668,6 +696,9 @@ gopkg.in/gomail.v2 v2.0.0-20160411212932-81ebce5c23df/go.mod h1:LRQQ+SO6ZHR7tOkp
 gopkg.in/ini.v1 v1.56.0/go.mod h1:pNLf8WUiyNEtQjuu5G5vTm06TEv9tsIgeAvK8hOrP4k=
 gopkg.in/ini.v1 v1.67.0 h1:Dgnx+6+nfE+IfzjUEISNeydPJh9AXNNsWbGP9KzCsOA=
 gopkg.in/ini.v1 v1.67.0/go.mod h1:pNLf8WUiyNEtQjuu5G5vTm06TEv9tsIgeAvK8hOrP4k=
+gopkg.in/natefinch/lumberjack.v2 v2.0.0/go.mod h1:l0ndWWf7gzL7RNwBG7wST/UCcT4T24xpD6X8LsfU/+k=
+gopkg.in/natefinch/lumberjack.v2 v2.2.1 h1:bBRl1b0OH9s/DuPhuXpNl+VtCaJXFZ5/uEFST95x9zc=
+gopkg.in/natefinch/lumberjack.v2 v2.2.1/go.mod h1:YD8tP3GAjkrDg1eZH7EGmyESg/lsYskCTPBJVb9jqSc=
 gopkg.in/square/go-jose.v2 v2.6.0 h1:NGk74WTnPKBNUhNzQX7PYcTLUjoq7mzKk2OKbvwk2iI=
 gopkg.in/square/go-jose.v2 v2.6.0/go.mod h1:M9dMgbHiYLoDGQrXy7OpJDJWiKiU//h+vD76mk0e1AI=
 gopkg.in/yaml.v2 v2.2.2/go.mod h1:hI93XBmqTisBFMUTm0b8Fm+jr3Dg1NNxqwp+5A1VGuI=
@@ -693,8 +724,8 @@ honnef.co/go/tools v0.0.0-20190102054323-c2f93a96b099/go.mod h1:rf3lG4BRIbNafJWh
 honnef.co/go/tools v0.0.0-20190523083050-ea95bdfd59fc/go.mod h1:rf3lG4BRIbNafJWhAfAdb/ePZxsR/4RtNHQocxwk9r4=
 modernc.org/libc v1.22.5 h1:91BNch/e5B0uPbJFgqbxXuOnxBQjlS//icfQEGmvyjE=
 modernc.org/libc v1.22.5/go.mod h1:jj+Z7dTNX8fBScMVNRAYZ/jF91K8fdT2hYMThc3YjBY=
-modernc.org/mathutil v1.5.0 h1:rV0Ko/6SfM+8G+yKiyI830l3Wuz1zRutdslNoQ0kfiQ=
-modernc.org/mathutil v1.5.0/go.mod h1:mZW8CKdRPY1v87qxC/wUdX5O1qDzXMP5TH3wjfpga6E=
+modernc.org/mathutil v1.6.0 h1:fRe9+AmYlaej+64JsEEhoWuAYBkOtQiMEU7n/XgfYi4=
+modernc.org/mathutil v1.6.0/go.mod h1:Ui5Q9q1TR2gFm0AQRqQUaBWFLAhQpCwNcuhBOSedWPo=
 modernc.org/memory v1.5.0 h1:N+/8c5rE6EqugZwHii4IFsaJ7MUhoWX07J5tC/iI5Ds=
 modernc.org/memory v1.5.0/go.mod h1:PkUhL0Mugw21sHPeskwZW4D6VscE/GQJOnIpCnW6pSU=
 modernc.org/sqlite v1.23.1 h1:nrSBg4aRQQwq59JpvGEQ15tNxoO5pX/kUjcRNwSAGQM=

memsto/target_cache.go

@@ -27,7 +27,8 @@ type TargetCacheType struct {
 	redis           storage.Redis
 
 	sync.RWMutex
-	targets map[string]*models.Target // key: ident
+	targets      map[string]*models.Target // key: ident
+	targetsIndex map[string][]string       // key: ip, value: ident list
 }
 
 func NewTargetCache(ctx *ctx.Context, stats *Stats, redis storage.Redis) *TargetCacheType {
@@ -38,6 +39,7 @@ func NewTargetCache(ctx *ctx.Context, stats *Stats, redis storage.Redis) *Target
 		stats:           stats,
 		redis:           redis,
 		targets:         make(map[string]*models.Target),
+		targetsIndex:    make(map[string][]string),
 	}
 
 	tc.SyncTargets()
@@ -51,6 +53,7 @@ func (tc *TargetCacheType) Reset() {
 	tc.statTotal = -1
 	tc.statLastUpdated = -1
 	tc.targets = make(map[string]*models.Target)
+	tc.targetsIndex = make(map[string][]string)
 }
 
 func (tc *TargetCacheType) StatChanged(total, lastUpdated int64) bool {
@@ -62,8 +65,17 @@ func (tc *TargetCacheType) StatChanged(total, lastUpdated int64) bool {
 }
 
 func (tc *TargetCacheType) Set(m map[string]*models.Target, total, lastUpdated int64) {
+	idx := make(map[string][]string, len(m))
+	for ident, target := range m {
+		if _, ok := idx[target.HostIp]; !ok {
+			idx[target.HostIp] = []string{}
+		}
+		idx[target.HostIp] = append(idx[target.HostIp], ident)
+	}
+
 	tc.Lock()
 	tc.targets = m
+	tc.targetsIndex = idx
 	tc.Unlock()
 
 	// only one goroutine used, so no need lock
@@ -78,6 +90,75 @@ func (tc *TargetCacheType) Get(ident string) (*models.Target, bool) {
 	return val, has
 }
 
+func (tc *TargetCacheType) GetByIp(ip string) ([]*models.Target, bool) {
+	tc.RLock()
+	defer tc.RUnlock()
+	idents, has := tc.targetsIndex[ip]
+	if !has {
+		return nil, false
+	}
+	targs := make([]*models.Target, 0, len(idents))
+	for _, ident := range idents {
+		if val, has := tc.targets[ident]; has {
+			targs = append(targs, val)
+		}
+	}
+	return targs, len(targs) > 0
+}
+
+func (tc *TargetCacheType) GetAll() []*models.Target {
+	tc.RLock()
+	defer tc.RUnlock()
+	lst := make([]*models.Target, 0, len(tc.targets))
+	for _, target := range tc.targets {
+		lst = append(lst, target)
+	}
+	return lst
+}
+
+// GetAllBeatTime 返回所有 target 的心跳时间 map，key 为 ident，value 为 BeatTime
+func (tc *TargetCacheType) GetAllBeatTime() map[string]int64 {
+	tc.RLock()
+	defer tc.RUnlock()
+	beatTimeMap := make(map[string]int64, len(tc.targets))
+	for ident, target := range tc.targets {
+		beatTimeMap[ident] = target.BeatTime
+	}
+	return beatTimeMap
+}
+
+// refreshBeatTime 从 Redis 刷新缓存中所有 target 的 BeatTime
+func (tc *TargetCacheType) refreshBeatTime() {
+	if tc.redis == nil {
+		return
+	}
+
+	// 快照 ident 列表，避免持锁访问 Redis
+	tc.RLock()
+	idents := make([]string, 0, len(tc.targets))
+	for ident := range tc.targets {
+		idents = append(idents, ident)
+	}
+	tc.RUnlock()
+
+	if len(idents) == 0 {
+		return
+	}
+
+	beatTimes := models.FetchBeatTimesFromRedis(tc.redis, idents)
+	if len(beatTimes) == 0 {
+		return
+	}
+
+	tc.Lock()
+	for ident, ts := range beatTimes {
+		if target, ok := tc.targets[ident]; ok {
+			target.BeatTime = ts
+		}
+	}
+	tc.Unlock()
+}
+
 func (tc *TargetCacheType) Gets(idents []string) []*models.Target {
 	tc.RLock()
 	defer tc.RUnlock()
@@ -105,7 +186,7 @@ func (tc *TargetCacheType) GetOffsetHost(targets []*models.Target, now, offset i
 			continue
 		}
 
-		if now-target.UpdateAt > 120 {
+		if now-target.BeatTime > 120 {
 			// means this target is not a active host, do not check offset
 			continue
 		}
@@ -147,6 +228,7 @@ func (tc *TargetCacheType) syncTargets() error {
 	}
 
 	if !tc.StatChanged(stat.Total, stat.LastUpdated) {
+		tc.refreshBeatTime()
 		tc.stats.GaugeCronDuration.WithLabelValues("sync_targets").Set(0)
 		tc.stats.GaugeSyncNumber.WithLabelValues("sync_targets").Set(0)
 		dumper.PutSyncRecord("targets", start.Unix(), -1, -1, "not changed")
@@ -170,6 +252,9 @@ func (tc *TargetCacheType) syncTargets() error {
 		}
 	}
 
+	// 从 Redis 批量获取心跳时间填充 BeatTime
+	models.FillTargetsBeatTime(tc.redis, lst)
+
 	for i := 0; i < len(lst); i++ {
 		m[lst[i].Ident] = lst[i]
 	}
@@ -186,57 +271,18 @@ func (tc *TargetCacheType) syncTargets() error {
 
 // get host update time
 func (tc *TargetCacheType) GetHostUpdateTime(targets []string) map[string]int64 {
-	metaMap := make(map[string]int64)
 	if tc.redis == nil {
-		return metaMap
+		return make(map[string]int64)
 	}
 
-	num := 0
-	var keys []string
-	for i := 0; i < len(targets); i++ {
-		keys = append(keys, models.WrapIdentUpdateTime(targets[i]))
-		num++
-		if num == 100 {
-			vals := storage.MGet(context.Background(), tc.redis, keys)
-			for _, value := range vals {
-				var hostUpdateTime models.HostUpdateTime
-				if value == nil {
-					continue
-				}
-
-				err := json.Unmarshal(value, &hostUpdateTime)
-				if err != nil {
-					logger.Errorf("failed to unmarshal host meta: %s value:%v", err, value)
-					continue
-				}
-				metaMap[hostUpdateTime.Ident] = hostUpdateTime.UpdateTime
-			}
-			keys = keys[:0]
-			num = 0
-		}
-	}
-
-	vals := storage.MGet(context.Background(), tc.redis, keys)
-	for _, value := range vals {
-		var hostUpdateTime models.HostUpdateTime
-		if value == nil {
-			continue
-		}
-
-		err := json.Unmarshal(value, &hostUpdateTime)
-		if err != nil {
-			logger.Warningf("failed to unmarshal host err:%v value:%s", err, string(value))
-			continue
-		}
-		metaMap[hostUpdateTime.Ident] = hostUpdateTime.UpdateTime
-	}
+	metaMap := models.FetchBeatTimesFromRedis(tc.redis, targets)
 
 	for _, ident := range targets {
 		if _, ok := metaMap[ident]; !ok {
 			// if not exists, get from cache
 			target, exists := tc.Get(ident)
 			if exists {
-				metaMap[ident] = target.UpdateAt
+				metaMap[ident] = target.BeatTime
 			}
 		}
 	}

models/alert_rule.go

@@ -509,10 +509,16 @@ func (ar *AlertRule) Verify() error {
 
 	ar.AppendTags = strings.TrimSpace(ar.AppendTags)
 	arr := strings.Fields(ar.AppendTags)
+	appendTagKeys := make(map[string]struct{})
 	for i := 0; i < len(arr); i++ {
 		if !strings.Contains(arr[i], "=") {
 			return fmt.Errorf("AppendTags(%s) invalid", arr[i])
 		}
+		pair := strings.SplitN(arr[i], "=", 2)
+		if _, exists := appendTagKeys[pair[0]]; exists {
+			return fmt.Errorf("AppendTags has duplicate key: %s", pair[0])
+		}
+		appendTagKeys[pair[0]] = struct{}{}
 	}
 
 	gids := strings.Fields(ar.NotifyGroups)

models/anomaly_point.go

@@ -46,6 +46,12 @@ func NewAnomalyPoint(key string, labels map[string]string, ts int64, value float
 }
 
 func (v *AnomalyPoint) ReadableValue() string {
+	if len(v.ValuesUnit) > 0 {
+		for _, unit := range v.ValuesUnit { // 配置了单位，优先用配置了单位的值
+			return unit.Text
+		}
+	}
+
 	ret := fmt.Sprintf("%.5f", v.Value)
 	ret = strings.TrimRight(ret, "0")
 	return strings.TrimRight(ret, ".")

models/datasource.go

@@ -45,6 +45,7 @@ type Datasource struct {
 	CreatedBy       string                 `json:"created_by"`
 	UpdatedBy       string                 `json:"updated_by"`
 	IsDefault       bool                   `json:"is_default"`
+	Weight          int                    `json:"weight"`
 	Transport       *http.Transport        `json:"-" gorm:"-"`
 	ForceSave       bool                   `json:"force_save" gorm:"-"`
 }
@@ -517,7 +518,8 @@ func (ds *Datasource) Encrypt(openRsa bool, publicKeyData []byte) error {
 // Decrypt 用于 edge 将从中心同步的数据源解密，中心不可调用
 func (ds *Datasource) Decrypt() error {
 	if rsaConfig == nil {
-		return errors.New("rsa config is nil")
+		logger.Debugf("datasource %s rsa config is nil", ds.Name)
+		return nil
 	}
 
 	if !rsaConfig.OpenRSA {

models/event_pipeline.go

@@ -29,8 +29,8 @@ type EventPipeline struct {
 	Nodes []WorkflowNode `json:"nodes,omitempty" gorm:"type:text;serializer:json"`
 	// 节点连接关系
 	Connections Connections `json:"connections,omitempty" gorm:"type:text;serializer:json"`
-	// 环境变量（工作流级别的配置变量）
-	EnvVariables []EnvVariable `json:"env_variables,omitempty" gorm:"type:text;serializer:json"`
+	// 输入参数（工作流级别的配置变量）
+	Inputs []InputVariable `json:"inputs,omitempty" gorm:"type:text;serializer:json"`
 
 	CreateAt int64  `json:"create_at" gorm:"type:bigint"`
 	CreateBy string `json:"create_by" gorm:"type:varchar(64)"`
@@ -73,8 +73,8 @@ func (e *EventPipeline) Verify() error {
 	if e.Connections == nil {
 		e.Connections = make(Connections)
 	}
-	if e.EnvVariables == nil {
-		e.EnvVariables = make([]EnvVariable, 0)
+	if e.Inputs == nil {
+		e.Inputs = make([]InputVariable, 0)
 	}
 
 	return nil
@@ -245,36 +245,10 @@ func (e *EventPipeline) FillWorkflowFields() {
 	}
 }
 
-func (e *EventPipeline) GetEnvMap() map[string]string {
-	envMap := make(map[string]string)
-	for _, v := range e.EnvVariables {
-		envMap[v.Key] = v.Value
+func (e *EventPipeline) GetInputsMap() map[string]string {
+	inputsMap := make(map[string]string)
+	for _, v := range e.Inputs {
+		inputsMap[v.Key] = v.Value
 	}
-	return envMap
-}
-
-func (e *EventPipeline) GetSecretKeys() map[string]bool {
-	secretKeys := make(map[string]bool)
-	for _, v := range e.EnvVariables {
-		if v.Secret {
-			secretKeys[v.Key] = true
-		}
-	}
-	return secretKeys
-}
-
-func (e *EventPipeline) ValidateEnvVariables(overrides map[string]string) error {
-	// 合并默认值和覆盖值
-	merged := e.GetEnvMap()
-	for k, v := range overrides {
-		merged[k] = v
-	}
-
-	// 校验必填项
-	for _, v := range e.EnvVariables {
-		if v.Required && merged[v.Key] == "" {
-			return fmt.Errorf("required env variable %s is missing", v.Key)
-		}
-	}
-	return nil
+	return inputsMap
 }

models/event_pipeline_execution.go

@@ -2,9 +2,13 @@ package models
 
 import (
 	"encoding/json"
+	"errors"
 	"fmt"
 
 	"github.com/ccfos/nightingale/v6/pkg/ctx"
+	"github.com/ccfos/nightingale/v6/pkg/poster"
+
+	"gorm.io/gorm"
 )
 
 // 执行状态常量
@@ -42,8 +46,8 @@ type EventPipelineExecution struct {
 	// 触发者信息
 	TriggerBy string `json:"trigger_by" gorm:"type:varchar(64)"`
 
-	// 环境变量快照（脱敏后存储）
-	EnvSnapshot string `json:"env_snapshot,omitempty" gorm:"type:text"`
+	// 输入参数快照（脱敏后存储）
+	InputsSnapshot string `json:"inputs_snapshot,omitempty" gorm:"type:text"`
 }
 
 func (e *EventPipelineExecution) TableName() string {
@@ -70,28 +74,31 @@ func (e *EventPipelineExecution) GetNodeResults() ([]*NodeExecutionResult, error
 	return results, err
 }
 
-// SetEnvSnapshot 设置环境变量快照（脱敏后存储）
-func (e *EventPipelineExecution) SetEnvSnapshot(env map[string]string) error {
-	data, err := json.Marshal(env)
+// SetInputsSnapshot 设置输入参数快照（脱敏后存储）
+func (e *EventPipelineExecution) SetInputsSnapshot(inputs map[string]string) error {
+	data, err := json.Marshal(inputs)
 	if err != nil {
 		return err
 	}
-	e.EnvSnapshot = string(data)
+	e.InputsSnapshot = string(data)
 	return nil
 }
 
-// GetEnvSnapshot 获取环境变量快照
-func (e *EventPipelineExecution) GetEnvSnapshot() (map[string]string, error) {
-	if e.EnvSnapshot == "" {
+// GetInputsSnapshot 获取输入参数快照
+func (e *EventPipelineExecution) GetInputsSnapshot() (map[string]string, error) {
+	if e.InputsSnapshot == "" {
 		return nil, nil
 	}
-	var env map[string]string
-	err := json.Unmarshal([]byte(e.EnvSnapshot), &env)
-	return env, err
+	var inputs map[string]string
+	err := json.Unmarshal([]byte(e.InputsSnapshot), &inputs)
+	return inputs, err
 }
 
 // CreateEventPipelineExecution 创建执行记录
 func CreateEventPipelineExecution(c *ctx.Context, execution *EventPipelineExecution) error {
+	if !c.IsCenter {
+		return poster.PostByUrls(c, "/v1/n9e/event-pipeline-execution", execution)
+	}
 	return DB(c).Create(execution).Error
 }
 
@@ -105,6 +112,9 @@ func GetEventPipelineExecution(c *ctx.Context, id string) (*EventPipelineExecuti
 	var execution EventPipelineExecution
 	err := DB(c).Where("id = ?", id).First(&execution).Error
 	if err != nil {
+		if errors.Is(err, gorm.ErrRecordNotFound) {
+			return nil, nil
+		}
 		return nil, err
 	}
 	return &execution, nil
@@ -145,12 +155,15 @@ func ListEventPipelineExecutionsByEventID(c *ctx.Context, eventID int64) ([]*Eve
 }
 
 // ListAllEventPipelineExecutions 获取所有 Pipeline 的执行记录列表
-func ListAllEventPipelineExecutions(c *ctx.Context, pipelineName, mode, status string, limit, offset int) ([]*EventPipelineExecution, int64, error) {
+func ListAllEventPipelineExecutions(c *ctx.Context, pipelineId int64, pipelineName, mode, status string, limit, offset int) ([]*EventPipelineExecution, int64, error) {
 	var executions []*EventPipelineExecution
 	var total int64
 
 	session := DB(c).Model(&EventPipelineExecution{})
 
+	if pipelineId > 0 {
+		session = session.Where("pipeline_id = ?", pipelineId)
+	}
 	if pipelineName != "" {
 		session = session.Where("pipeline_name LIKE ?", "%"+pipelineName+"%")
 	}
@@ -270,8 +283,8 @@ func GetEventPipelineExecutionStatistics(c *ctx.Context, pipelineID int64) (*Eve
 // EventPipelineExecutionDetail 执行详情（包含解析后的节点结果）
 type EventPipelineExecutionDetail struct {
 	EventPipelineExecution
-	NodeResultsParsed []*NodeExecutionResult `json:"node_results_parsed"`
-	EnvSnapshotParsed map[string]string      `json:"env_snapshot_parsed"`
+	NodeResultsParsed    []*NodeExecutionResult `json:"node_results_parsed"`
+	InputsSnapshotParsed map[string]string      `json:"inputs_snapshot_parsed"`
 }
 
 // GetEventPipelineExecutionDetail 获取执行详情
@@ -281,6 +294,10 @@ func GetEventPipelineExecutionDetail(c *ctx.Context, id string) (*EventPipelineE
 		return nil, err
 	}
 
+	if execution == nil {
+		return &EventPipelineExecutionDetail{}, nil
+	}
+
 	detail := &EventPipelineExecutionDetail{
 		EventPipelineExecution: *execution,
 	}
@@ -292,12 +309,12 @@ func GetEventPipelineExecutionDetail(c *ctx.Context, id string) (*EventPipelineE
 	}
 	detail.NodeResultsParsed = nodeResults
 
-	// 解析环境变量快照
-	envSnapshot, err := execution.GetEnvSnapshot()
+	// 解析输入参数快照
+	inputsSnapshot, err := execution.GetInputsSnapshot()
 	if err != nil {
-		return nil, fmt.Errorf("parse env snapshot error: %w", err)
+		return nil, fmt.Errorf("parse inputs snapshot error: %w", err)
 	}
-	detail.EnvSnapshotParsed = envSnapshot
+	detail.InputsSnapshotParsed = inputsSnapshot
 
 	return detail, nil
 }

models/migrate/migrate.go

@@ -234,6 +234,7 @@ type Target struct {
 type Datasource struct {
 	IsDefault  bool   `gorm:"column:is_default;type:boolean;comment:is default datasource"`
 	Identifier string `gorm:"column:identifier;type:varchar(255);default:'';comment:identifier"`
+	Weight     int    `gorm:"column:weight;type:int;default:0;comment:weight for sorting"`
 }
 
 type Configs struct {

models/recording_rule.go

@@ -44,6 +44,7 @@ type QueryConfig struct {
 	Exp               string  `json:"exp"`
 	WriteDatasourceId int64   `json:"write_datasource_id"`
 	Delay             int     `json:"delay"`
+	WritebackEnabled  bool    `json:"writeback_enabled"` // 是否写入与查询数据源相同的数据源
 }
 
 type Query struct {
@@ -211,7 +212,6 @@ func (re *RecordingRule) Update(ctx *ctx.Context, ref RecordingRule) error {
 
 	ref.FE2DB()
 	ref.Id = re.Id
-	ref.GroupId = re.GroupId
 	ref.CreateAt = re.CreateAt
 	ref.CreateBy = re.CreateBy
 	ref.UpdateAt = time.Now().Unix()

models/target.go

@@ -1,6 +1,8 @@
 package models
 
 import (
+	"context"
+	"encoding/json"
 	"log"
 	"sort"
 	"strings"
@@ -8,6 +10,7 @@ import (
 
 	"github.com/ccfos/nightingale/v6/pkg/ctx"
 	"github.com/ccfos/nightingale/v6/pkg/poster"
+	"github.com/ccfos/nightingale/v6/storage"
 	"golang.org/x/exp/slices"
 
 	"github.com/pkg/errors"
@@ -36,6 +39,7 @@ type Target struct {
 	OS           string            `json:"os" gorm:"column:os"`
 	HostTags     []string          `json:"host_tags" gorm:"serializer:json"`
 
+	BeatTime   int64    `json:"beat_time" gorm:"-"` // 实时心跳时间，从 Redis 获取
 	UnixTime   int64    `json:"unixtime" gorm:"-"`
 	Offset     int64    `json:"offset" gorm:"-"`
 	TargetUp   float64  `json:"target_up" gorm:"-"`
@@ -97,12 +101,6 @@ func (t *Target) MatchGroupId(gid ...int64) bool {
 }
 
 func (t *Target) AfterFind(tx *gorm.DB) (err error) {
-	delta := time.Now().Unix() - t.UpdateAt
-	if delta < 60 {
-		t.TargetUp = 2
-	} else if delta < 180 {
-		t.TargetUp = 1
-	}
 	t.FillTagsMap()
 	return
 }
@@ -182,6 +180,24 @@ func BuildTargetWhereWithHosts(hosts []string) BuildTargetWhereOption {
 	}
 }
 
+func BuildTargetWhereWithIdents(idents []string) BuildTargetWhereOption {
+	return func(session *gorm.DB) *gorm.DB {
+		if len(idents) > 0 {
+			session = session.Where("ident in (?)", idents)
+		}
+		return session
+	}
+}
+
+func BuildTargetWhereExcludeIdents(idents []string) BuildTargetWhereOption {
+	return func(session *gorm.DB) *gorm.DB {
+		if len(idents) > 0 {
+			session = session.Where("ident not in (?)", idents)
+		}
+		return session
+	}
+}
+
 func BuildTargetWhereWithQuery(query string) BuildTargetWhereOption {
 	return func(session *gorm.DB) *gorm.DB {
 		if query != "" {
@@ -203,17 +219,6 @@ func BuildTargetWhereWithQuery(query string) BuildTargetWhereOption {
 	}
 }
 
-func BuildTargetWhereWithDowntime(downtime int64) BuildTargetWhereOption {
-	return func(session *gorm.DB) *gorm.DB {
-		if downtime > 0 {
-			session = session.Where("target.update_at < ?", time.Now().Unix()-downtime)
-		} else if downtime < 0 {
-			session = session.Where("target.update_at > ?", time.Now().Unix()+downtime)
-		}
-		return session
-	}
-}
-
 func buildTargetWhere(ctx *ctx.Context, options ...BuildTargetWhereOption) *gorm.DB {
 	sub := DB(ctx).Model(&Target{}).Distinct("target.ident")
 	for _, opt := range options {
@@ -264,21 +269,6 @@ func TargetCountByFilter(ctx *ctx.Context, query []map[string]interface{}) (int6
 	return Count(session)
 }
 
-func MissTargetGetsByFilter(ctx *ctx.Context, query []map[string]interface{}, ts int64) ([]*Target, error) {
-	var lst []*Target
-	session := TargetFilterQueryBuild(ctx, query, 0, 0)
-	session = session.Where("update_at < ?", ts)
-
-	err := session.Order("ident").Find(&lst).Error
-	return lst, err
-}
-
-func MissTargetCountByFilter(ctx *ctx.Context, query []map[string]interface{}, ts int64) (int64, error) {
-	session := TargetFilterQueryBuild(ctx, query, 0, 0)
-	session = session.Where("update_at < ?", ts)
-	return Count(session)
-}
-
 func TargetFilterQueryBuild(ctx *ctx.Context, query []map[string]interface{}, limit, offset int) *gorm.DB {
 	sub := DB(ctx).Model(&Target{}).Distinct("target.ident").Joins("left join " +
 		"target_busi_group on target.ident = target_busi_group.target_ident")
@@ -619,6 +609,66 @@ func (t *Target) FillMeta(meta *HostMeta) {
 	t.RemoteAddr = meta.RemoteAddr
 }
 
+// FetchBeatTimesFromRedis 从 Redis 批量获取心跳时间，返回 ident -> updateTime 的映射
+func FetchBeatTimesFromRedis(redis storage.Redis, idents []string) map[string]int64 {
+	result := make(map[string]int64, len(idents))
+	if redis == nil || len(idents) == 0 {
+		return result
+	}
+
+	num := 0
+	var keys []string
+	for i := 0; i < len(idents); i++ {
+		keys = append(keys, WrapIdentUpdateTime(idents[i]))
+		num++
+		if num == 100 {
+			fetchBeatTimeBatch(redis, keys, result)
+			keys = keys[:0]
+			num = 0
+		}
+	}
+
+	if len(keys) > 0 {
+		fetchBeatTimeBatch(redis, keys, result)
+	}
+
+	return result
+}
+
+func fetchBeatTimeBatch(redis storage.Redis, keys []string, result map[string]int64) {
+	vals := storage.MGet(context.Background(), redis, keys)
+	for _, value := range vals {
+		if value == nil {
+			continue
+		}
+		var hut HostUpdateTime
+		if err := json.Unmarshal(value, &hut); err != nil {
+			logger.Warningf("failed to unmarshal host update time: %v", err)
+			continue
+		}
+		result[hut.Ident] = hut.UpdateTime
+	}
+}
+
+// FillTargetsBeatTime 从 Redis 批量获取心跳时间填充 target.BeatTime
+func FillTargetsBeatTime(redis storage.Redis, targets []*Target) {
+	if len(targets) == 0 {
+		return
+	}
+
+	idents := make([]string, len(targets))
+	for i, t := range targets {
+		idents[i] = t.Ident
+	}
+
+	beatTimes := FetchBeatTimesFromRedis(redis, idents)
+	for _, t := range targets {
+		if ts, ok := beatTimes[t.Ident]; ok {
+			t.BeatTime = ts
+		}
+	}
+}
+
 func TargetIdents(ctx *ctx.Context, ids []int64) ([]string, error) {
 	var ret []string
 

models/user.go

@@ -315,6 +315,18 @@ func (u *User) UpdatePassword(ctx *ctx.Context, password, updateBy string) error
 	}).Error
 }
 
+func (u *User) UpdateUserGroup(ctx *ctx.Context, userGroupIds []int64) error {
+
+	count := len(userGroupIds)
+	for i := 0; i < count; i++ {
+		err := UserGroupMemberAdd(ctx, userGroupIds[i], u.Id)
+		if err != nil {
+			return err
+		}
+	}
+	return nil
+}
+
 func UpdateUserLastActiveTime(ctx *ctx.Context, userId int64, lastActiveTime int64) error {
 	return DB(ctx).Model(&User{}).Where("id = ?", userId).Updates(map[string]interface{}{
 		"last_active_time": lastActiveTime,

models/workflow.go

@@ -33,13 +33,11 @@ type ConnectionTarget struct {
 	Index int    `json:"index"` // 目标节点的输入端口索引
 }
 
-// EnvVariable 环境变量
-type EnvVariable struct {
+// InputVariable 输入参数
+type InputVariable struct {
 	Key         string `json:"key"`                   // 变量名
 	Value       string `json:"value"`                 // 默认值
 	Description string `json:"description,omitempty"` // 描述
-	Secret      bool   `json:"secret,omitempty"`      // 是否敏感（日志脱敏）
-	Required    bool   `json:"required,omitempty"`    // 是否必填
 }
 
 // NodeOutput 节点执行输出
@@ -104,8 +102,8 @@ type WorkflowTriggerContext struct {
 	// 请求ID（API/Cron 触发使用）
 	RequestID string `json:"request_id"`
 
-	// 环境变量覆盖
-	EnvOverrides map[string]string `json:"env_overrides"`
+	// 输入参数覆盖
+	InputsOverrides map[string]string `json:"inputs_overrides"`
 
 	// 流式输出（API 调用时动态指定）
 	Stream bool `json:"stream"`
@@ -118,7 +116,7 @@ type WorkflowTriggerContext struct {
 
 type WorkflowContext struct {
 	Event    *AlertCurEvent         `json:"event"`            // 当前事件
-	Env      map[string]string      `json:"env"`              // 环境变量/配置（静态，来自 Pipeline 配置）
+	Inputs   map[string]string      `json:"inputs"`           // 前置输入参数（静态，用户配置）
 	Vars     map[string]interface{} `json:"vars"`             // 节点间传递的数据（动态，运行时产生）
 	Metadata map[string]string      `json:"metadata"`         // 执行元数据（request_id、start_time 等）
 	Output   map[string]interface{} `json:"output,omitempty"` // 输出结果（非告警场景使用）
@@ -128,19 +126,6 @@ type WorkflowContext struct {
 	StreamChan chan *StreamChunk `json:"-"` // 流式数据通道（不序列化）
 }
 
-// SanitizedEnv 返回脱敏后的环境变量（用于日志和存储）
-func (ctx *WorkflowContext) SanitizedEnv(secretKeys map[string]bool) map[string]string {
-	sanitized := make(map[string]string)
-	for k, v := range ctx.Env {
-		if secretKeys[k] {
-			sanitized[k] = "******"
-		} else {
-			sanitized[k] = v
-		}
-	}
-	return sanitized
-}
-
 // StreamChunk 类型常量
 const (
 	StreamTypeThinking   = "thinking"    // AI 思考过程（ReAct Thought）

pkg/cfg/scan.go

@@ -1,7 +1,7 @@
 package cfg
 
 import (
-	"io/ioutil"
+	"os"
 )
 
 type scanner struct {
@@ -23,6 +23,6 @@ func (s *scanner) Data() []byte {
 
 func (s *scanner) Read(file string) {
 	if s.err == nil {
-		s.data, s.err = ioutil.ReadFile(file)
+		s.data, s.err = os.ReadFile(file)
 	}
 }

pkg/feishu/feishu.go

@@ -0,0 +1,348 @@
+package feishu
+
+import (
+	"bytes"
+	"context"
+	"fmt"
+	"net/url"
+	"strings"
+	"sync"
+	"time"
+
+	"github.com/ccfos/nightingale/v6/storage"
+	"github.com/google/uuid"
+	"github.com/pkg/errors"
+	"github.com/toolkits/pkg/logger"
+
+	lark "github.com/larksuite/oapi-sdk-go/v3"
+	larkcore "github.com/larksuite/oapi-sdk-go/v3/core"
+	larkauthen "github.com/larksuite/oapi-sdk-go/v3/service/authen/v1"
+	larkcontact "github.com/larksuite/oapi-sdk-go/v3/service/contact/v3"
+)
+
+const defaultAuthURL = "https://accounts.feishu.cn/open-apis/authen/v1/authorize"
+const SsoTypeName = "feishu"
+
+type SsoClient struct {
+	Enable       bool
+	FeiShuConfig *Config `json:"-"`
+	Ctx          context.Context
+	client       *lark.Client
+	sync.RWMutex
+}
+
+type Config struct {
+	Enable            bool     `json:"enable"`
+	AuthURL           string   `json:"auth_url"`
+	DisplayName       string   `json:"display_name"`
+	AppID             string   `json:"app_id"`
+	AppSecret         string   `json:"app_secret"`
+	RedirectURL       string   `json:"redirect_url"`
+	UsernameField     string   `json:"username_field"`  // name, email, phone
+	FeiShuEndpoint    string   `json:"feishu_endpoint"` // 飞书API端点，默认为 open.feishu.cn
+	Proxy             string   `json:"proxy"`
+	CoverAttributes   bool     `json:"cover_attributes"`
+	DefaultRoles      []string `json:"default_roles"`
+	DefaultUserGroups []int64  `json:"default_user_groups"`
+}
+
+type CallbackOutput struct {
+	Redirect    string `json:"redirect"`
+	Msg         string `json:"msg"`
+	AccessToken string `json:"accessToken"`
+	Username    string `json:"Username"`
+	Nickname    string `json:"Nickname"`
+	Phone       string `yaml:"Phone"`
+	Email       string `yaml:"Email"`
+}
+
+func wrapStateKey(key string) string {
+	return "n9e_feishu_oauth_" + key
+}
+
+// createClient 创建飞书SDK客户端（v3版本）
+func (c *Config) createClient() (*lark.Client, error) {
+	opts := []lark.ClientOptionFunc{
+		lark.WithLogLevel(larkcore.LogLevelInfo),
+		lark.WithEnableTokenCache(true), // 启用token缓存
+	}
+
+	if c.FeiShuEndpoint != "" {
+		lark.FeishuBaseUrl = c.FeiShuEndpoint
+	}
+
+	// 创建客户端（v3版本）
+	client := lark.NewClient(
+		c.AppID,
+		c.AppSecret,
+		opts...,
+	)
+
+	return client, nil
+}
+
+func New(cf Config) *SsoClient {
+	var s = &SsoClient{}
+	if !cf.Enable {
+		return s
+	}
+	s.Reload(cf)
+	return s
+}
+
+func (s *SsoClient) AuthCodeURL(state string) (string, error) {
+	var buf bytes.Buffer
+	feishuAuthURL := defaultAuthURL
+	if s.FeiShuConfig.AuthURL != "" {
+		feishuAuthURL = s.FeiShuConfig.AuthURL
+	}
+	buf.WriteString(feishuAuthURL)
+	v := url.Values{
+		"app_id": {s.FeiShuConfig.AppID},
+		"state":  {state},
+	}
+	v.Set("redirect_uri", s.FeiShuConfig.RedirectURL)
+
+	if s.FeiShuConfig.RedirectURL == "" {
+		return "", errors.New("FeiShu OAuth RedirectURL is empty")
+	}
+
+	if strings.Contains(feishuAuthURL, "?") {
+		buf.WriteByte('&')
+	} else {
+		buf.WriteByte('?')
+	}
+	buf.WriteString(v.Encode())
+
+	return buf.String(), nil
+}
+
+// GetUserToken 通过授权码获取用户access token和user_id（使用SDK v3）
+func (s *SsoClient) GetUserToken(code string) (string, string, error) {
+	if s.client == nil {
+		return "", "", errors.New("feishu client is not initialized")
+	}
+
+	ctx := context.Background()
+
+	// 使用SDK v3的authen服务获取access token
+	req := larkauthen.NewCreateAccessTokenReqBuilder().
+		Body(larkauthen.NewCreateAccessTokenReqBodyBuilder().
+			GrantType("authorization_code").
+			Code(code).
+			Build()).
+		Build()
+
+	resp, err := s.client.Authen.AccessToken.Create(ctx, req)
+	if err != nil {
+		return "", "", fmt.Errorf("feishu get access token error: %w", err)
+	}
+
+	// 检查响应
+	if !resp.Success() {
+		return "", "", fmt.Errorf("feishu api error: code=%d, msg=%s", resp.Code, resp.Msg)
+	}
+
+	if resp.Data == nil {
+		return "", "", errors.New("feishu api returned empty data")
+	}
+
+	userID := ""
+	if resp.Data.UserId != nil {
+		userID = *resp.Data.UserId
+	}
+	if userID == "" {
+		return "", "", errors.New("feishu api returned empty user_id")
+	}
+
+	accessToken := ""
+	if resp.Data.AccessToken != nil {
+		accessToken = *resp.Data.AccessToken
+	}
+	if accessToken == "" {
+		return "", "", errors.New("feishu api returned empty access_token")
+	}
+
+	return accessToken, userID, nil
+}
+
+// GetUserInfo 通过user_id获取用户详细信息（使用SDK v3）
+// 注意：SDK内部会自动管理token，所以不需要传入accessToken
+func (s *SsoClient) GetUserInfo(userID string) (*larkcontact.GetUserRespData, error) {
+	if s.client == nil {
+		return nil, errors.New("feishu client is not initialized")
+	}
+
+	ctx := context.Background()
+
+	// 使用SDK v3的contact服务获取用户详情
+	req := larkcontact.NewGetUserReqBuilder().
+		UserId(userID).
+		UserIdType(larkcontact.UserIdTypeUserId).
+		Build()
+
+	resp, err := s.client.Contact.User.Get(ctx, req)
+	if err != nil {
+		return nil, fmt.Errorf("feishu get user detail error: %w", err)
+	}
+
+	// 检查响应
+	if !resp.Success() {
+		return nil, fmt.Errorf("feishu api error: code=%d, msg=%s", resp.Code, resp.Msg)
+	}
+
+	if resp.Data == nil || resp.Data.User == nil {
+		return nil, errors.New("feishu api returned empty user data")
+	}
+
+	return resp.Data, nil
+}
+
+func (s *SsoClient) Reload(feishuConfig Config) {
+	s.Lock()
+	defer s.Unlock()
+	s.Enable = feishuConfig.Enable
+	s.FeiShuConfig = &feishuConfig
+
+	// 重新创建客户端
+	if feishuConfig.Enable && feishuConfig.AppID != "" && feishuConfig.AppSecret != "" {
+		client, err := feishuConfig.createClient()
+		if err != nil {
+			logger.Errorf("create feishu client error: %v", err)
+		} else {
+			s.client = client
+		}
+	}
+}
+
+func (s *SsoClient) GetDisplayName() string {
+	s.RLock()
+	defer s.RUnlock()
+	if !s.Enable {
+		return ""
+	}
+
+	return s.FeiShuConfig.DisplayName
+}
+
+func (s *SsoClient) Authorize(redis storage.Redis, redirect string) (string, error) {
+	state := uuid.New().String()
+	ctx := context.Background()
+
+	err := redis.Set(ctx, wrapStateKey(state), redirect, time.Duration(300*time.Second)).Err()
+	if err != nil {
+		return "", err
+	}
+
+	s.RLock()
+	defer s.RUnlock()
+
+	return s.AuthCodeURL(state)
+}
+
+func (s *SsoClient) Callback(redis storage.Redis, ctx context.Context, code, state string) (*CallbackOutput, error) {
+	// 通过code获取access token和user_id
+	accessToken, userID, err := s.GetUserToken(code)
+	if err != nil {
+		return nil, fmt.Errorf("feishu GetUserToken error: %s", err)
+	}
+
+	// 获取用户详细信息
+	userData, err := s.GetUserInfo(userID)
+	if err != nil {
+		return nil, fmt.Errorf("feishu GetUserInfo error: %s", err)
+	}
+
+	// 获取redirect URL
+	redirect := ""
+	if redis != nil {
+		redirect, err = fetchRedirect(redis, ctx, state)
+		if err != nil {
+			logger.Errorf("get redirect err:%v code:%s state:%s", err, code, state)
+		}
+	}
+	if redirect == "" {
+		redirect = "/"
+	}
+
+	err = deleteRedirect(redis, ctx, state)
+	if err != nil {
+		logger.Errorf("delete redirect err:%v code:%s state:%s", err, code, state)
+	}
+
+	var callbackOutput CallbackOutput
+	if userData == nil || userData.User == nil {
+		return nil, fmt.Errorf("feishu GetUserInfo failed, user data is nil")
+	}
+
+	user := userData.User
+	logger.Debugf("feishu get user info userID %s result %+v", userID, user)
+
+	// 提取用户信息
+	username := ""
+	if user.UserId != nil {
+		username = *user.UserId
+	}
+	if username == "" {
+		return nil, errors.New("feishu user_id is empty")
+	}
+
+	nickname := ""
+	if user.Name != nil {
+		nickname = *user.Name
+	}
+
+	phone := ""
+	if user.Mobile != nil {
+		phone = *user.Mobile
+	}
+
+	email := ""
+	if user.Email != nil {
+		email = *user.Email
+	}
+
+	if email == "" {
+		if user.EnterpriseEmail != nil {
+			email = *user.EnterpriseEmail
+		}
+	}
+
+	callbackOutput.Redirect = redirect
+	callbackOutput.AccessToken = accessToken
+
+	// 根据UsernameField配置确定username
+	switch s.FeiShuConfig.UsernameField {
+	case "userid":
+		callbackOutput.Username = username
+	case "name":
+		if nickname == "" {
+			return nil, errors.New("feishu user name is empty")
+		}
+		callbackOutput.Username = nickname
+	case "phone":
+		if phone == "" {
+			return nil, errors.New("feishu user phone is empty")
+		}
+		callbackOutput.Username = phone
+	default:
+		if email == "" {
+			return nil, errors.New("feishu user email is empty")
+		}
+		callbackOutput.Username = email
+	}
+
+	callbackOutput.Nickname = nickname
+	callbackOutput.Email = email
+	callbackOutput.Phone = phone
+
+	return &callbackOutput, nil
+}
+
+func fetchRedirect(redis storage.Redis, ctx context.Context, state string) (string, error) {
+	return redis.Get(ctx, wrapStateKey(state)).Result()
+}
+
+func deleteRedirect(redis storage.Redis, ctx context.Context, state string) error {
+	return redis.Del(ctx, wrapStateKey(state)).Err()
+}

pkg/tplx/fns.go

@@ -531,8 +531,13 @@ func Printf(format string, value interface{}) string {
 
 	switch valType {
 	case reflect.String:
+		strValue := value.(string)
+		// Check if it's a value with unit (contains both digits and non-numeric chars like letters or %)
+		if isValueWithUnit(strValue) {
+			return strValue
+		}
 		// Try converting string to float
-		if floatValue, err := strconv.ParseFloat(value.(string), 64); err == nil {
+		if floatValue, err := strconv.ParseFloat(strValue, 64); err == nil {
 			return fmt.Sprintf(format, floatValue)
 		}
 		return fmt.Sprintf(format, value)
@@ -544,6 +549,32 @@ func Printf(format string, value interface{}) string {
 	}
 }
 
+// isValueWithUnit checks if a string is a numeric value with unit
+// e.g., "11.5%", "100MB", "10a" returns true
+// e.g., "11", "11.11", "-3.14" returns false
+func isValueWithUnit(s string) bool {
+	if s == "" {
+		return false
+	}
+
+	hasDigit := false
+	hasUnit := false
+
+	for _, r := range s {
+		if r >= '0' && r <= '9' {
+			hasDigit = true
+		} else if r == '.' || r == '-' || r == '+' {
+			// These are valid numeric characters, not units
+			continue
+		} else {
+			// Any other character (letters, %, etc.) is considered a unit
+			hasUnit = true
+		}
+	}
+
+	return hasDigit && hasUnit
+}
+
 func floatToTime(v float64) (*time.Time, error) {
 	if math.IsNaN(v) || math.IsInf(v, 0) {
 		return nil, errNaNOrInf

pkg/tplx/tplx.go

@@ -7,7 +7,7 @@ import (
 	"regexp"
 	"strings"
 	templateT "text/template"
-
+    "encoding/base64"
 	"github.com/toolkits/pkg/logger"
 )
 
@@ -64,6 +64,16 @@ var TemplateFuncMap = template.FuncMap{
 	"jsonMarshal":                   JsonMarshal,
 	"mapDifference":                 MapDifference,
 	"tagsMapToStr":                  TagsMapToStr,
+    "b64enc": func(s string) string {
+        return base64.StdEncoding.EncodeToString([]byte(s))
+    },
+    "b64dec": func(s string) string {
+        data, err := base64.StdEncoding.DecodeString(s)
+        if err != nil {
+            return s
+        }
+        return string(data)
+    },
 }
 
 // NewTemplateFuncMap copy on write for TemplateFuncMap

pushgw/idents/idents.go

@@ -106,6 +106,7 @@ func (s *Set) UpdateTargets(lst []string, now int64) error {
 		return nil
 	}
 
+	// 心跳时间只写入 Redis，不再写入 MySQL update_at
 	err := s.updateTargetsUpdateTs(lst, now, s.redis)
 	if err != nil {
 		logger.Errorf("update_ts: failed to update targets: %v error: %v", lst, err)
@@ -133,12 +134,7 @@ func (s *Set) UpdateTargets(lst []string, now int64) error {
 		return nil
 	}
 
-	if s.configs.UpdateDBTargetTimestampDisable {
-		// 如果 mysql 压力太大，关闭更新 db 的操作
-		return nil
-	}
-
-	// there are some idents not found in db, so insert them
+	// 新 target 仍需 INSERT 注册到 MySQL
 	var exists []string
 	err = s.ctx.DB.Table("target").Where("ident in ?", lst).Pluck("ident", &exists).Error
 	if err != nil {
@@ -153,38 +149,13 @@ func (s *Set) UpdateTargets(lst []string, now int64) error {
 		}
 	}
 
-	// 从批量更新一批机器的时间戳，改成逐台更新，是为了避免批量更新时，mysql的锁竞争问题
-	start := time.Now()
-	duration := time.Since(start).Seconds()
-	if len(exists) > 0 {
-		sema := semaphore.NewSemaphore(s.configs.UpdateDBTargetConcurrency)
-		wg := sync.WaitGroup{}
-		for i := 0; i < len(exists); i++ {
-			sema.Acquire()
-			wg.Add(1)
-			go func(ident string) {
-				defer sema.Release()
-				defer wg.Done()
-				s.updateDBTargetTs(ident, now)
-			}(exists[i])
-		}
-		wg.Wait()
-	}
-	pstat.DBOperationLatency.WithLabelValues("update_targets_ts").Observe(duration)
-
 	return nil
 }
 
-func (s *Set) updateDBTargetTs(ident string, now int64) {
-	err := s.ctx.DB.Exec("UPDATE target SET update_at = ? WHERE ident = ?", now, ident).Error
-	if err != nil {
-		logger.Error("update_target: failed to update target:", ident, "error:", err)
-	}
-}
-
 func (s *Set) updateTargetsUpdateTs(lst []string, now int64, redis storage.Redis) error {
 	if redis == nil {
-		return fmt.Errorf("redis is nil")
+		logger.Debugf("update_ts: redis is nil")
+		return nil
 	}
 
 	newMap := make(map[string]interface{}, len(lst))
@@ -247,7 +218,7 @@ func (s *Set) writeTargetTsInRedis(ctx context.Context, redis storage.Redis, con
 
 	for i := 0; i < retryCount; i++ {
 		start := time.Now()
-		err := storage.MSet(ctx, redis, content)
+		err := storage.MSet(ctx, redis, content, 24*time.Hour)
 		duration := time.Since(start).Seconds()
 
 		logger.Debugf("update_ts: write target ts in redis, keys: %v, retryCount: %d, retryInterval: %v, error: %v", keys, retryCount, retryInterval, err)

pushgw/pconf/conf.go

@@ -18,8 +18,6 @@ type Pushgw struct {
 	UpdateTargetRetryIntervalMills int64
 	UpdateTargetTimeoutMills       int64
 	UpdateTargetBatchSize          int
-	UpdateDBTargetConcurrency      int
-	UpdateDBTargetTimestampDisable bool
 	PushConcurrency                int
 	UpdateTargetByUrlConcurrency   int
 
@@ -129,10 +127,6 @@ func (p *Pushgw) PreCheck() {
 		p.UpdateTargetBatchSize = 20
 	}
 
-	if p.UpdateDBTargetConcurrency <= 0 {
-		p.UpdateDBTargetConcurrency = 16
-	}
-
 	if p.PushConcurrency <= 0 {
 		p.PushConcurrency = 16
 	}

pushgw/router/fns.go

@@ -109,21 +109,30 @@ func (rt *Router) debugSample(remoteAddr string, v *prompb.TimeSeries) {
 }
 
 func (rt *Router) DropSample(v *prompb.TimeSeries) bool {
-	filters := rt.Pushgw.DropSample
-	if len(filters) == 0 {
+	// 快速路径：检查仅 __name__ 的过滤器 O(1)
+	if len(rt.dropByNameOnly) > 0 {
+		for i := 0; i < len(v.Labels); i++ {
+			if v.Labels[i].Name == "__name__" {
+				if _, ok := rt.dropByNameOnly[v.Labels[i].Value]; ok {
+					return true
+				}
+				break // __name__ 只会出现一次，找到后直接跳出
+			}
+		}
+	}
+
+	// 慢速路径：处理复杂的多条件过滤器
+	if len(rt.dropComplex) == 0 {
 		return false
 	}
 
-	labelMap := make(map[string]string)
+	// 只有复杂过滤器存在时才创建 labelMap
+	labelMap := make(map[string]string, len(v.Labels))
 	for i := 0; i < len(v.Labels); i++ {
 		labelMap[v.Labels[i].Name] = v.Labels[i].Value
 	}
 
-	for _, filter := range filters {
-		if len(filter) == 0 {
-			continue
-		}
-
+	for _, filter := range rt.dropComplex {
 		if matchSample(filter, labelMap) {
 			return true
 		}

pushgw/router/router.go

@@ -6,6 +6,7 @@ import (
 
 	"github.com/gin-gonic/gin"
 	"github.com/prometheus/prometheus/prompb"
+	"github.com/toolkits/pkg/logger"
 
 	"github.com/ccfos/nightingale/v6/alert/aconf"
 	"github.com/ccfos/nightingale/v6/center/metas"
@@ -33,6 +34,10 @@ type Router struct {
 	Ctx            *ctx.Context
 	HandleTS       HandleTSFunc
 	HeartbeatApi   string
+
+	// 预编译的 DropSample 过滤器
+	dropByNameOnly map[string]struct{} // 仅 __name__ 条件的快速匹配
+	dropComplex    []map[string]string // 多条件的复杂匹配
 }
 
 func stat() gin.HandlerFunc {
@@ -51,7 +56,7 @@ func stat() gin.HandlerFunc {
 func New(httpConfig httpx.Config, pushgw pconf.Pushgw, aconf aconf.Alert, tc *memsto.TargetCacheType, bg *memsto.BusiGroupCacheType,
 	idents *idents.Set, metas *metas.Set,
 	writers *writer.WritersType, ctx *ctx.Context) *Router {
-	return &Router{
+	rt := &Router{
 		HTTP:           httpConfig,
 		Pushgw:         pushgw,
 		Aconf:          aconf,
@@ -63,6 +68,38 @@ func New(httpConfig httpx.Config, pushgw pconf.Pushgw, aconf aconf.Alert, tc *me
 		MetaSet:        metas,
 		HandleTS:       func(pt *prompb.TimeSeries) *prompb.TimeSeries { return pt },
 	}
+
+	// 预编译 DropSample 过滤器
+	rt.initDropSampleFilters()
+
+	return rt
+}
+
+// initDropSampleFilters 预编译 DropSample 过滤器，将单条件 __name__ 过滤器
+// 放入 map 实现 O(1) 查找，多条件过滤器保留原有逻辑
+func (rt *Router) initDropSampleFilters() {
+	rt.dropByNameOnly = make(map[string]struct{})
+	rt.dropComplex = make([]map[string]string, 0)
+
+	for _, filter := range rt.Pushgw.DropSample {
+		if len(filter) == 0 {
+			continue
+		}
+
+		// 如果只有一个条件且是 __name__，放入快速匹配 map
+		if len(filter) == 1 {
+			if name, ok := filter["__name__"]; ok {
+				rt.dropByNameOnly[name] = struct{}{}
+				continue
+			}
+		}
+
+		// 其他情况放入复杂匹配列表
+		rt.dropComplex = append(rt.dropComplex, filter)
+	}
+
+	logger.Infof("DropSample filters initialized: %d name-only, %d complex",
+		len(rt.dropByNameOnly), len(rt.dropComplex))
 }
 
 func (rt *Router) Config(r *gin.Engine) {

storage/redis.go

@@ -163,10 +163,10 @@ func MGet(ctx context.Context, r Redis, keys []string) [][]byte {
 	return vals
 }
 
-func MSet(ctx context.Context, r Redis, m map[string]interface{}) error {
+func MSet(ctx context.Context, r Redis, m map[string]interface{}, expiration time.Duration) error {
 	pipe := r.Pipeline()
 	for k, v := range m {
-		pipe.Set(ctx, k, v, 0)
+		pipe.Set(ctx, k, v, expiration)
 	}
 	_, err := pipe.Exec(ctx)
 	return err

storage/redis_test.go

@@ -30,7 +30,7 @@ func TestMiniRedisMGet(t *testing.T) {
 	mp["key2"] = "value2"
 	mp["key3"] = "value3"
 
-	err = MSet(context.Background(), rdb, mp)
+	err = MSet(context.Background(), rdb, mp, 0)
 	if err != nil {
 		t.Fatalf("failed to set miniredis value: %v", err)
 	}