Merge remote-tracking branch 'qiniu/develop'

Author: aice030

ENV_SETUP.md

@@ -1,10 +1,27 @@
-# 环境变量设置指南
+# 配置文件与环境变量指南
 
 ## 概述
 
-ZeroOps 项目需要配置 API 密钥来使用 AI 模型。本文档说明如何正确设置环境变量，包括模型配置、搜索后端、评估器设置等。
+ZeroOps 现已支持通过配置文件集中管理参数（推荐）。同时也兼容环境变量方式。
 
-## 方法1：使用 .env 文件（推荐）
+## 使用配置文件（推荐）
+
+1) 复制示例并编辑
+
+```bash
+cp config.example.json config.json
+vi config.json
+```
+
+2) 以配置文件启动服务
+
+```bash
+go run ./cmd/zeroops -f $(pwd)/config.json
+```
+
+配置文件中的字段包括：`server`、`logging`、`database`、`redis`、`alerting.healthcheck/prometheus/remediation/ruleset/receiver`。
+
+## 方法1：使用 .env 文件（可选）
 
 ### 步骤1：创建 .env 文件
 在项目根目录创建 `.env` 文件：

api/openapi/alerting-ml.yaml

@@ -0,0 +1,94 @@
+openapi: 3.0.3
+info:
+  title: alerting-ml service
+  version: 0.1.0
+  description: HTTP API for time series anomaly detection using IsolationForest
+servers:
+  - url: http://localhost:8081
+paths:
+  /api/v1/anomaly/detect:
+    post:
+      summary: Detect anomalies in a single time series
+      operationId: detectAnomalies
+      requestBody:
+        required: true
+        content:
+          application/json:
+            schema:
+              $ref: '#/components/schemas/DetectRequest'
+      responses:
+        '200':
+          description: Detection result
+          content:
+            application/json:
+              schema:
+                $ref: '#/components/schemas/DetectResponse'
+        '400':
+          description: Invalid input
+        '500':
+          description: Internal error
+components:
+  schemas:
+    Metadata:
+      type: object
+      properties:
+        alert_name:
+          type: string
+        severity:
+          type: string
+        labels:
+          type: object
+          additionalProperties:
+            type: string
+    DataPoint:
+      type: object
+      required: [timestamp, value]
+      properties:
+        timestamp:
+          type: string
+          format: date-time
+        value:
+          type: number
+          format: double
+    DetectRequest:
+      type: object
+      required: [data]
+      properties:
+        metadata:
+          $ref: '#/components/schemas/Metadata'
+        data:
+          type: array
+          items:
+            $ref: '#/components/schemas/DataPoint'
+        contamination:
+          type: number
+          default: 0.05
+        random_state:
+          type: integer
+          default: 42
+        ratio_threshold:
+          type: number
+          default: 0.2
+        streak_threshold:
+          type: integer
+          default: 20
+    Anomaly:
+      type: object
+      required: [start, end]
+      properties:
+        start:
+          type: string
+          format: date-time
+        end:
+          type: string
+          format: date-time
+    DetectResponse:
+      type: object
+      properties:
+        metadata:
+          $ref: '#/components/schemas/Metadata'
+        anomalies:
+          type: array
+          items:
+            $ref: '#/components/schemas/Anomaly'
+

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/src/views/ChangeLogView.vue

@@ -84,11 +84,31 @@
 import { ref, computed, onMounted, watch } from 'vue'
 import { useAppStore, type ChangeItem, type AlarmChangeItem } from '@/stores/app'
 import { mockApi } from '@/mock/api'
-import type { DeploymentChangelogResponse, DeploymentChangelogItem, AlertRuleChangelogResponse, AlertRuleChangeItem } from '@/mock/services'
+import { apiService } from '@/api'
+import type { DeploymentChangelogResponse } from '@/mock/services'
 import ChangeCard from '@/components/ChangeCard.vue'
 import AlarmChangeCard from '@/components/AlarmChangeCard.vue'
 import { ArrowLeft, Loading } from '@element-plus/icons-vue'
 
+interface AlertRuleChangeValue {
+  name: string
+  old: string
+  new: string
+}
+
+interface AlertRuleChangeItem {
+  name: string
+  editTime: string
+  scope: string
+  values: AlertRuleChangeValue[]
+  reason: string
+}
+
+interface AlertRuleChangelogResponse {
+  items: AlertRuleChangeItem[]
+  next?: string
+}
+
 const appStore = useAppStore()
 
 const activeTab = ref('service')
@@ -160,7 +180,7 @@ const transformAlertRuleChangelogToAlarmChangeItems = (changelogData: AlertRuleC
     const serviceName = item.scope?.startsWith('service:') ? item.scope.slice('service:'.length) + '服务' : '全局服务'
     
     // 构建变更描述
-    const changeDescription = item.values.map(value => {
+  const changeDescription = item.values.map((value) => {
       return `${value.name}: ${value.old} -> ${value.new}`
     }).join(', ')
     
@@ -200,21 +220,21 @@ const loadDeploymentChangelog = async (start?: string, limit?: number) => {
   }
 }
 
-// 加载告警规则变更记录
+// 加载告警规则变更记录（使用真实 API）
 const loadAlertRuleChangelog = async (start?: string, limit?: number) => {
   if (alertRuleLoading.value) return // 防止重复加载
   
   try {
     alertRuleLoading.value = true
     error.value = null
     
-    const response = await mockApi.getAlertRuleChangelog(start, limit)
-    alertRuleChangelog.value = response
+    const response = await apiService.getAlertRuleChangelog(start, limit ?? 10)
+    alertRuleChangelog.value = response.data
     
     // 转换数据格式
-    alarmChangeItems.value = transformAlertRuleChangelogToAlarmChangeItems(response.items)
+    alarmChangeItems.value = transformAlertRuleChangelogToAlarmChangeItems(response.data.items)
     
-    console.log('告警规则变更记录加载成功:', response)
+    console.log('告警规则变更记录加载成功:', response.data)
   } catch (err) {
     error.value = '加载告警规则变更记录失败'
     console.error('加载告警规则变更记录失败:', err)

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

@@ -0,0 +1,17 @@
+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.m.daocloud.io/distroless/base-debian12:nonroot
+WORKDIR /app
+COPY --from=builder /out/zeroops /app/zeroops
+# 复制配置文件目录
+COPY configs/ /app/configs/
+EXPOSE 8080
+ENTRYPOINT ["/app/zeroops"]
+CMD ["-f", "/app/config.json"]
+
+

cmd/zeroops/main.go

@@ -3,11 +3,10 @@ package main
 import (
 	"context"
 	"fmt"
-	"os"
 	"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"
@@ -17,16 +16,34 @@ import (
 	servicemanager "github.com/qiniu/zeroops/internal/service_manager"
 
 	// releasesystem "github.com/qiniu/zeroops/internal/release_system/api"
+	"strings"
+
+	"github.com/rs/zerolog"
 	"github.com/rs/zerolog/log"
 )
 
 func main() {
+	// load config first
 	log.Info().Msg("Starting zeroops api server")
 	cfg, err := config.Load()
 	if err != nil {
 		log.Fatal().Err(err).Msg("failed to load config")
 	}
 
+	// configure log level from config
+	switch strings.ToLower(cfg.Logging.Level) {
+	case "trace":
+		zerolog.SetGlobalLevel(zerolog.TraceLevel)
+	case "debug":
+		zerolog.SetGlobalLevel(zerolog.DebugLevel)
+	case "warn", "warning":
+		zerolog.SetGlobalLevel(zerolog.WarnLevel)
+	case "error":
+		zerolog.SetGlobalLevel(zerolog.ErrorLevel)
+	default:
+		zerolog.SetGlobalLevel(zerolog.DebugLevel)
+	}
+
 	serviceManagerSrv, err := servicemanager.NewServiceManagerServer(cfg)
 	if err != nil {
 		log.Fatal().Err(err).Msg("failed to create release system api")
@@ -52,28 +69,51 @@ func main() {
 	// start healthcheck scheduler and remediation consumer
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
-	interval := parseDuration(os.Getenv("HC_SCAN_INTERVAL"), 10*time.Second)
-	batch := parseInt(os.Getenv("HC_SCAN_BATCH"), 200)
-	workers := parseInt(os.Getenv("HC_WORKERS"), 1)
+
+	// bootstrap alert rules from config if provided
+	if err := healthcheck.BootstrapRulesFromConfigWithApp(ctx, alertDB, &cfg.Alerting.Ruleset); err != nil {
+		log.Error().Err(err).Msg("bootstrap rules from config failed")
+	}
+	interval := parseDuration(cfg.Alerting.Healthcheck.Interval, 10*time.Second)
+	batch := cfg.Alerting.Healthcheck.Batch
+	workers := cfg.Alerting.Healthcheck.Workers
 	if workers < 1 {
 		workers = 1
 	}
-	alertChSize := parseInt(os.Getenv("REMEDIATION_ALERT_CHAN_SIZE"), 1024)
+	alertChSize := cfg.Alerting.Healthcheck.AlertChanSize
 	alertCh := make(chan healthcheck.AlertMessage, alertChSize)
 
 	for i := 0; i < workers; i++ {
 		go healthcheck.StartScheduler(ctx, healthcheck.Deps{
 			DB:       alertDB,
-			Redis:    healthcheck.NewRedisClientFromEnv(),
+			Redis:    healthcheck.NewRedisClientFromConfig(&cfg.Redis),
 			AlertCh:  alertCh,
 			Batch:    batch,
 			Interval: interval,
 		})
 	}
-	rem := remediation.NewConsumer(alertDB, healthcheck.NewRedisClientFromEnv())
+	rem := remediation.NewConsumer(alertDB, healthcheck.NewRedisClientFromConfig(&cfg.Redis)).WithConfig(&cfg.Alerting.Remediation)
 	go rem.Start(ctx, alertCh)
 
-	router := fox.New()
+	// start Prometheus anomaly detection scheduler
+	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)
+	anomalyDetectClient := healthcheck.NewAnomalyDetectClient(promCfg)
+	go healthcheck.StartPrometheusScheduler(ctx, healthcheck.PrometheusDeps{
+		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 := 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 {

config.docker.json

@@ -0,0 +1,19 @@
+{
+  "server": { "bindAddr": ":8080" },
+  "logging": { "level": "info" },
+  "database": {
+    "host": "host.docker.internal",
+    "port": 5532,
+    "user": "postgres",
+    "password": "postgres123",
+    "dbname": "zeroops",
+    "sslmode": "disable"
+  },
+  "redis": { "addr": "host.docker.internal:16379", "password": "", "db": 0 },
+  "alerting": {
+    "healthcheck": { "interval": "10s", "batch": 200, "workers": 2, "alertChanSize": 1024 },
+    "prometheus": { "url": "http://10.210.10.33:9090", "queryTimeout": "30s", "anomalyAPIUrl": "http://alerting-ml:8081/api/v1/anomaly/detect", "anomalyAPITimeout": "10s", "schedulerInterval": "6h", "queryStep": "1m", "queryRange": "6h" },
+    "ruleset": { "configFile": "/app/configs/alerting/rules.json", "apiBase": "http://10.210.10.33:9999", "apiTimeout": "10s" },
+    "receiver": { "basicUser": "alert", "basicPass": "REDACTED", "bearer": "" }
+  }
+}