调试告警规则调整

Author: acd19ml

client/src/api/index.ts

@@ -2,6 +2,7 @@ import axios from 'axios'
 
 // 创建 axios 实例
 const api = axios.create({
+  // baseURL: (import.meta as any).env?.VITE_API_BASE_URL || 'http://10.210.10.33:8080',
   timeout: 10000,
   headers: {
     'Content-Type': 'application/json'

client/src/mock/services.ts

@@ -1845,27 +1845,9 @@ loadServiceAlertStatus()
  */
 export const serviceVersionAlertStatusMap: Record<string, Record<string, ServiceAlertStatus>> = {}
 
-const saveServiceVersionAlertStatus = () => {
-  try {
-    localStorage.setItem('serviceVersionAlertStatusMap', JSON.stringify(serviceVersionAlertStatusMap))
-    console.log('服务版本告警状态已保存到 localStorage')
-  } catch (error) {
-    console.error('保存服务版本告警状态失败:', error)
-  }
-}
+const saveServiceVersionAlertStatus = () => {}
 
-const loadServiceVersionAlertStatus = () => {
-  try {
-    const data = localStorage.getItem('serviceVersionAlertStatusMap')
-    if (data) {
-      const parsed = JSON.parse(data)
-      Object.assign(serviceVersionAlertStatusMap, parsed)
-      console.log('已从 localStorage 加载服务版本告警状态')
-    }
-  } catch (error) {
-    console.error('从 localStorage 加载服务版本告警状态失败:', error)
-  }
-}
+const loadServiceVersionAlertStatus = () => {}
 
 /**
  * 根据告警状态更新服务版本状态
@@ -1919,8 +1901,7 @@ export const clearServiceVersionAlertStatus = (serviceName: string, version?: st
   console.log(`已清除服务 ${serviceName} ${version ? '版本 ' + version : '所有版本'} 的告警状态`)
 }
 
-// 页面加载时恢复服务版本告警状态
-loadServiceVersionAlertStatus()
+// 页面加载时不再从 localStorage 恢复服务版本告警状态（禁用持久化）
 
 // ==================== 发布任务状态管理 ====================
 // 管理服务的发布任务状态，用于显示发布指示器

client/vite.config.ts

@@ -18,7 +18,7 @@ export default defineConfig({
   server: {
     proxy: {
       '/v1': {
-        target: 'http://127.0.0.1:8080',
+        target: 'http://10.210.10.33:8080',
         changeOrigin: true,
         secure: false,
       }

cmd/zeroops/Dockerfile

@@ -1,11 +1,11 @@
-FROM golang:1.24-alpine AS builder
+FROM docker.m.daocloud.io/library/golang:1.24-alpine AS builder
 WORKDIR /src
 COPY go.mod go.sum ./
 RUN go mod download
 COPY . .
 RUN CGO_ENABLED=0 GOOS=linux GOARCH=amd64 go build -o /out/zeroops ./cmd/zeroops
 
-FROM gcr.io/distroless/base-debian12
+FROM gcr.m.daocloud.io/distroless/base-debian12:nonroot
 WORKDIR /app
 COPY --from=builder /out/zeroops /app/zeroops
 # 复制配置文件目录

cmd/zeroops/main.go

@@ -6,7 +6,7 @@ import (
 	"strconv"
 	"time"
 
-	"github.com/fox-gonic/fox"
+	"github.com/gin-gonic/gin"
 	alertapi "github.com/qiniu/zeroops/internal/alerting/api"
 	adb "github.com/qiniu/zeroops/internal/alerting/database"
 	"github.com/qiniu/zeroops/internal/alerting/service/healthcheck"
@@ -96,22 +96,24 @@ func main() {
 	go rem.Start(ctx, alertCh)
 
 	// start Prometheus anomaly detection scheduler
-	promInterval := parseDuration(cfg.Alerting.Prometheus.SchedulerInterval, 6*time.Hour)
+	promInterval := parseDuration(cfg.Alerting.Prometheus.SchedulerInterval, 5*time.Minute)
 	promStep := parseDuration(cfg.Alerting.Prometheus.QueryStep, time.Minute)
 	promRange := parseDuration(cfg.Alerting.Prometheus.QueryRange, 6*time.Hour)
 	promCfg := healthcheck.NewPrometheusConfigFromApp(&cfg.Alerting.Prometheus)
-	promClient := healthcheck.NewPrometheusClient(promCfg)
+	anomalyDetectClient := healthcheck.NewAnomalyDetectClient(promCfg)
 	go healthcheck.StartPrometheusScheduler(ctx, healthcheck.PrometheusDeps{
-		DB:               alertDB,
-		PrometheusClient: promClient,
-		Interval:         promInterval,
-		QueryStep:        promStep,
-		QueryRange:       promRange,
-		RulesetBase:      cfg.Alerting.Ruleset.APIBase,
-		RulesetTimeout:   parseDuration(cfg.Alerting.Ruleset.APITimeout, 10*time.Second),
+		DB:                  alertDB,
+		AnomalyDetectClient: anomalyDetectClient,
+		Interval:            promInterval,
+		QueryStep:           promStep,
+		QueryRange:          promRange,
+		RulesetBase:         cfg.Alerting.Ruleset.APIBase,
+		RulesetTimeout:      parseDuration(cfg.Alerting.Ruleset.APITimeout, 10*time.Second),
 	})
 
-	router := fox.New()
+	router := gin.New()
+	router.Use(gin.Logger())
+	router.Use(gin.Recovery())
 	router.Use(middleware.Authentication)
 	alertapi.NewApiWithConfig(router, cfg)
 	if err := serviceManagerSrv.UseApi(router); err != nil {

configs/alerting/rules.json

@@ -6,7 +6,7 @@
       "expr": "histogram_quantile(0.98, sum(rate(http_latency_seconds_bucket{}[2m])) by (service, service_version, le))",
       "op": ">",
       "severity": "P0",
-      "watch_time": "5 minutes",
+      "watch_time": "5m",
       "metas": [
         { "labels": { "service": "storage-service", "service_version": "1.0.0" }, "threshold": 1000 },
         { "labels": { "service": "queue-service", "service_version": "1.0.0" }, "threshold": 1000 },
@@ -19,7 +19,7 @@
       "expr": "histogram_quantile(0.98, sum(rate(http_latency_seconds_bucket{}[2m])) by (service, service_version, le))",
       "op": ">",
       "severity": "P1",
-      "watch_time": "4 minutes",
+      "watch_time": "4m",
       "metas": [
         { "labels": { "service": "storage-service", "service_version": "1.0.0" }, "threshold": 500 },
         { "labels": { "service": "queue-service", "service_version": "1.0.0" }, "threshold": 500 },
@@ -31,7 +31,7 @@
         "description":"HTTP error rate by service P0",
         "op":">",
         "severity":"P0",
-        "watch_time":"5 minutes",
+        "watch_time":"5m",
         "expr":"sum(rate(http_latency_seconds_count{\"http.status_code\"=~\"4..|5..\", \"http.route\"!=\"/metrics\"}[2m])) by (service, service_version) / sum(rate(http_latency_seconds_count{\"http.route\"!=\"/metrics\"}[2m])) by (service, service_version)",
         "metas":[
             {"labels":{"service":"storage-service","service_version":"1.0.0"},"threshold":5},
@@ -44,7 +44,7 @@
         "description":"HTTP error rate by service P1",
         "op":">",
         "severity":"P1",
-        "watch_time":"5 minutes",
+        "watch_time":"5m",
         "expr":"sum(rate(http_latency_seconds_count{\"http.status_code\"=~\"4..|5..\", \"http.route\"!=\"/metrics\"}[2m])) by (service, service_version) / sum(rate(http_latency_seconds_count{\"http.route\"!=\"/metrics\"}[2m])) by (service, service_version)",
         "metas":[
             {"labels":{"service":"storage-service","service_version":"1.0.0"},"threshold":3},

docs/alerting/database-design.md

@@ -13,7 +13,7 @@
 
 ## 数据表设计
 
-### 1) talert_issues（告警问题表）
+### 1) alert_issues（告警问题表）
 
 存储告警问题的主要信息。
 
@@ -23,7 +23,7 @@
 | state | enum(Closed, Open) | 问题状态 |
 | level | varchar(32) | 告警等级：如 P0/P1/Px |
 | alert_state | enum(Pending, Restored, AutoRestored, InProcessing) | 处理状态 |
-| title | varchar(255) | 告警标题 |
+| title | varchar(255) | 告警标题 
 | labels | json | 标签，格式：[{key, value}] |
 | alert_since | TIMESTAMP(6) | 告警发生时间 |
 | resolved_at | TIMESTAMP(6) | 告警结束时间 |
@@ -64,8 +64,6 @@
 | labels | text | labels 的 JSON 字符串表示（规范化后） |
 | old_threshold | numeric | 旧阈值（可空） |
 | new_threshold | numeric | 新阈值（可空） |
-| old_watch | interval | 旧观察窗口（可空） |
-| new_watch | interval | 新观察窗口（可空） |
 
 
 **索引建议：**

go.mod

@@ -3,7 +3,7 @@ module github.com/qiniu/zeroops
 go 1.24
 
 require (
-	github.com/fox-gonic/fox v0.0.6
+	github.com/gin-gonic/gin v1.10.1
 	github.com/google/uuid v1.6.0
 	github.com/jackc/pgx/v5 v5.5.5
 	github.com/lib/pq v1.10.9

internal/alerting/README.md

@@ -170,9 +170,9 @@ docker exec -i zeroops-postgres-1 psql -U postgres -d zeroops -c \
   "CREATE TABLE IF NOT EXISTS alert_issue_comments (issue_id text, create_at timestamp, content text, PRIMARY KEY(issue_id, create_at));"
 ```
 
-### 2) 初始化/重置规则表（alert_rules / alert_rule_metas）
+### 2) 初始化/重置规则表（alert_rules / alert_rule_metas / alert_meta_change_logs）
 
-注意：该脚本会 DROP 并重建 `alert_rules` 与 `alert_rule_metas`，仅用于本地/开发环境。
+注意：该脚本会 DROP 并重建 `alert_rules`、`alert_rule_metas` 和 `alert_meta_change_logs`，仅用于本地/开发环境。
 
 脚本位置：`scripts/sql/alert_rules_bootstrap.sql`
 
@@ -191,12 +191,13 @@ psql -U postgres -d zeroops -f scripts/sql/alert_rules_bootstrap.sql
 ```bash
 docker exec -i zeroops-postgres-1 psql -U postgres -d zeroops -c "SELECT name, severity FROM alert_rules;"
 docker exec -i zeroops-postgres-1 psql -U postgres -d zeroops -c "SELECT alert_name, labels, threshold FROM alert_rule_metas;"
+docker exec -i zeroops-postgres-1 psql -U postgres -d zeroops -c "SELECT alert_name, change_type, change_time FROM alert_meta_change_logs;"
 ```
 
 ### 2) 清空数据库与缓存（可选，保证从空开始）
 
 ```bash
-docker exec -i zeroops-pg psql -U postgres -d zeroops -c "TRUNCATE TABLE alert_issue_comments, service_states, alert_issues;"
+docker exec -i zeroops-pg psql -U postgres -d zeroops -c "TRUNCATE TABLE alert_issue_comments, service_states, alert_issues, alert_meta_change_logs;"
 docker exec -i zeroops-redis redis-cli --raw DEL $(docker exec -i zeroops-redis redis-cli --raw KEYS 'alert:*' | tr '\n' ' ') 2>/dev/null || true
 docker exec -i zeroops-redis redis-cli --raw DEL $(docker exec -i zeroops-redis redis-cli --raw KEYS 'service_state:*' | tr '\n' ' ') 2>/dev/null || true
 ```

internal/alerting/api/api.go

@@ -3,7 +3,7 @@ package api
 import (
 	"fmt"
 
-	"github.com/fox-gonic/fox"
+	"github.com/gin-gonic/gin"
 	adb "github.com/qiniu/zeroops/internal/alerting/database"
 	"github.com/qiniu/zeroops/internal/alerting/service/healthcheck"
 	receiver "github.com/qiniu/zeroops/internal/alerting/service/receiver"
@@ -12,15 +12,15 @@ import (
 
 type Api struct{}
 
-func NewApi(router *fox.Engine) *Api { return NewApiWithConfig(router, nil) }
+func NewApi(router *gin.Engine) *Api { return NewApiWithConfig(router, nil) }
 
-func NewApiWithConfig(router *fox.Engine, cfg *config.Config) *Api {
+func NewApiWithConfig(router *gin.Engine, cfg *config.Config) *Api {
 	api := &Api{}
 	api.setupRouters(router, cfg)
 	return api
 }
 
-func (api *Api) setupRouters(router *fox.Engine, cfg *config.Config) {
+func (api *Api) setupRouters(router *gin.Engine, cfg *config.Config) {
 	var h *receiver.Handler
 	var alertDB *adb.Database
 	if cfg != nil {