239 lines
7.0 KiB
Markdown
239 lines
7.0 KiB
Markdown
# 监控系统全面升级实施方案
|
||
|
||
> **For Hermes:** Execute phase by phase, each phase is independently deployable.
|
||
|
||
**Goal:** 构建 4 大监控模块：主机监控、中间件监控、SNMP 设备监控、站点拨测
|
||
|
||
**架构：** Agent 扩展采集器 → Backend 新增路由/处理器 → 前端新建监控面板
|
||
|
||
**Tech Stack:** Go (Agent/Backend), React+TypeScript (Frontend), MySQL+Redis (Storage), gosnmp (SNMP library)
|
||
|
||
---
|
||
|
||
## Phase 1: Agent 指标扩展 + 主机监控仪表盘 🖥️
|
||
|
||
### Task 1.1: Agent 新增采集器 - 网络指标
|
||
**Files:**
|
||
- Create: `agent/collector/network.go`
|
||
- Modify: `agent/model/model.go`
|
||
- Modify: `agent/main.go`
|
||
|
||
**网络指标采集：**
|
||
- 并发连接数 (ESTABLISHED via /proc/net/sockstat)
|
||
- TIME_WAIT 连接数
|
||
- 网卡 rx_bytes/s, tx_bytes/s（两次采集差值 / 间隔）
|
||
- 网卡名列表和速率
|
||
|
||
### Task 1.2: Agent 新增采集器 - 磁盘 IO
|
||
**Files:**
|
||
- Create: `agent/collector/diskio.go`
|
||
- Modify: `agent/model/model.go`
|
||
|
||
**磁盘 IO 指标：**
|
||
- 读取 BPS (bytes/s)
|
||
- 写入 BPS (bytes/s)
|
||
- 读取 IOPS (count/s)
|
||
- 写入 IOPS (count/s)
|
||
|
||
### Task 1.3: Agent 新增采集器 - 进程监控
|
||
**Files:**
|
||
- Create: `agent/collector/process.go`
|
||
- Modify: `agent/model/model.go`
|
||
|
||
**进程监控：**
|
||
- Top N 进程 CPU/内存使用
|
||
- 进程总数
|
||
- 僵尸进程数
|
||
|
||
### Task 1.4: Agent 新增采集器 - 系统信息（启动时一次）
|
||
**Files:**
|
||
- Create: `agent/collector/system.go`
|
||
- Modify: `agent/model/model.go`
|
||
|
||
**系统信息：**
|
||
- OS 版本 (cat /etc/os-release)
|
||
- 内核版本 (uname -r)
|
||
- 运行时间 (uptime)
|
||
- ulimit -n (nofile 上限)
|
||
- ulimit -u (nproc 上限)
|
||
|
||
### Task 1.5: Backend - 扩展 AgentReport 模型
|
||
**Files:**
|
||
- Modify: `backend/internal/models/agent.go`
|
||
- Modify: `backend/internal/api/handlers/agent.go` (ReceiveReport)
|
||
|
||
**扩展字段：**
|
||
- NetworkInfo JSON (net_connections, net_rx_bps, net_tx_bps, nic_name)
|
||
- DiskIOInfo JSON (read_bps, write_bps, read_iops, write_iops)
|
||
- ProcessInfo JSON (top_processes, total_proc, zombie_count)
|
||
- SystemInfo JSON (os_version, kernel, uptime, ulimit_nofile, ulimit_nproc)
|
||
|
||
### Task 1.6: Backend - 主机监控 API
|
||
**Files:**
|
||
- Create: `backend/internal/api/handlers/host_monitor.go`
|
||
- Modify: `backend/internal/api/routes/routes.go`
|
||
|
||
**API 端点：**
|
||
- `GET /api/v1/monitor/host/:hostname/overview` - 主机概览
|
||
- `GET /api/v1/monitor/host/:hostname/metrics?metric=cpu&from=&to=` - 历史指标
|
||
- `GET /api/v1/monitor/host/:hostname/processes` - 进程列表
|
||
- `GET /api/v1/monitor/host/:hostname/network` - 网络指标
|
||
|
||
### Task 1.7: Frontend - 主机监控面板
|
||
**Files:**
|
||
- Create: `frontend/src/pages/HostMonitor.tsx`
|
||
- Create: `frontend/src/components/monitor/HostCard.tsx`
|
||
- Create: `frontend/src/components/monitor/MetricChart.tsx`
|
||
- Modify: `frontend/src/App.tsx` (add route /monitor/host)
|
||
|
||
**展示内容：**
|
||
- 主机卡片：CPU/内存/磁盘使用率仪表盘
|
||
- 趋势图：CPU/内存/磁盘 5m/1h/6h/24h
|
||
- 网络：并发连接数、网卡带宽趋势
|
||
- 进程 TOP 10 表格
|
||
|
||
---
|
||
|
||
## Phase 2: 中间件监控 🔌
|
||
|
||
### Task 2.1: Agent 新增采集器 - Redis
|
||
**Files:**
|
||
- Create: `agent/collector/redis.go`
|
||
- Modify: `agent/model/model.go`
|
||
|
||
**Redis 指标（需要配置 host:port:password）：**
|
||
- 连接数 (connected_clients)
|
||
- 内存使用 (used_memory)
|
||
- 命中率 (keyspace_hits / (hits + misses))
|
||
- Key 数量
|
||
- 慢查询数
|
||
- QPS (instantaneous_ops_per_sec)
|
||
|
||
### Task 2.2: Agent 新增采集器 - MySQL
|
||
**Files:**
|
||
- Create: `agent/collector/mysql.go`
|
||
|
||
**MySQL 指标：**
|
||
- 当前连接数 / 最大连接数
|
||
- QPS (Questions 差值)
|
||
- 慢查询数
|
||
- Buffer Pool 命中率
|
||
- 复制延迟（如有）
|
||
|
||
### Task 2.3: Backend - 中间件监控 API + 配置管理
|
||
**Files:**
|
||
- Create: `backend/internal/api/handlers/middleware_monitor.go`
|
||
- Create: `backend/internal/models/middleware_config.go` (RedisConfig, MySQLConfig)
|
||
- Modify: `backend/internal/api/routes/routes.go`
|
||
|
||
**API：**
|
||
- `POST /api/v1/monitor/middleware/config` - 配置中间件认证
|
||
- `GET /api/v1/monitor/middleware/redis/:hostname` - Redis 指标
|
||
- `GET /api/v1/monitor/middleware/mysql/:hostname` - MySQL 指标
|
||
|
||
### Task 2.4: Frontend - 中间件监控面板
|
||
**Files:**
|
||
- Create: `frontend/src/pages/MiddlewareMonitor.tsx`
|
||
|
||
---
|
||
|
||
## Phase 3: SNMP 设备监控 📡
|
||
|
||
### Task 3.1: Backend - SNMP 采集引擎 + OID 模板系统
|
||
**Files:**
|
||
- Create: `backend/internal/api/handlers/snmp_monitor.go`
|
||
- Create: `backend/internal/models/snmp_template.go`
|
||
- Create: `backend/pkg/snmp/collector.go` (uses gosnmp)
|
||
- Add dep: `github.com/gosnmp/gosnmp`
|
||
|
||
**SNMP 模板系统：**
|
||
- 模板包含一组 OID → 指标名映射
|
||
- 预置模板：Cisco 交换机、H3C 交换机、华为交换机、服务器 iLO/iDRAC
|
||
- 用户可自定义 OID 模板
|
||
|
||
### Task 3.2: Backend - SNMP 定时采集任务
|
||
**Files:**
|
||
- Create: `backend/internal/scheduler/snmp_scheduler.go`
|
||
|
||
**定时采集：**
|
||
- 每 5 分钟对配置的 SNMP 设备执行采集
|
||
- 存储到 agent_reports 表（type=snmp）
|
||
|
||
### Task 3.3: Frontend - SNMP 设备管理 + 监控面板
|
||
**Files:**
|
||
- Create: `frontend/src/pages/SNMPMonitor.tsx`
|
||
- Create: `frontend/src/components/monitor/SNMPDeviceCard.tsx`
|
||
|
||
---
|
||
|
||
## Phase 4: 站点拨测 🌐
|
||
|
||
### Task 4.1: Backend - 拨测引擎
|
||
**Files:**
|
||
- Create: `backend/internal/api/handlers/site_monitor.go`
|
||
- Create: `backend/internal/models/site_task.go`
|
||
- Create: `backend/pkg/probe/` (HTTP, PING, WebSocket, DNS, TCP, UDP, SMTP, POP3, FTP probes)
|
||
|
||
**8 种拨测类型实现：**
|
||
- HTTP(S): GET/POST，检查状态码 + 响应时间 + 正则匹配 body
|
||
- PING: ICMP echo
|
||
- WebSocket: 连接 + 心跳
|
||
- DNS: 解析时间 + 结果验证
|
||
- TCP: 端口连通性
|
||
- UDP: 端口连通性
|
||
- SMTP: 连接 + STARTTLS
|
||
- POP3: 连接 + 认证
|
||
- FTP: 连接 + 登录
|
||
|
||
### Task 4.2: Backend - 拨测定时任务
|
||
**Files:**
|
||
- Modify: `backend/internal/scheduler/` (add probe scheduler)
|
||
|
||
### Task 4.3: Frontend - 站点拨测管理面板
|
||
**Files:**
|
||
- Create: `frontend/src/pages/SiteMonitor.tsx`
|
||
- Create: `frontend/src/components/monitor/ProbeTaskForm.tsx`
|
||
- Create: `frontend/src/components/monitor/ProbeResultChart.tsx`
|
||
|
||
**展示内容：**
|
||
- 拨测任务列表（类型、URL、频率、状态）
|
||
- 响应时间趋势图
|
||
- 可用率统计
|
||
- 告警配置
|
||
|
||
---
|
||
|
||
## Phase 5: 导航重构 🧭
|
||
|
||
### Task 5.1: 导航结构重组
|
||
**Files:**
|
||
- Modify: `frontend/src/components/layout/MainLayout.tsx` 或 Sidebar
|
||
|
||
**新导航结构：**
|
||
```
|
||
📊 监控告警
|
||
├── 🖥️ 主机监控 /monitor/host
|
||
├── 🔌 中间件监控 /monitor/middleware
|
||
├── 📡 SNMP 监控 /monitor/snmp
|
||
├── 🌐 站点拨测 /monitor/site
|
||
├── 📈 Agent 监控 /agent-monitor (已有)
|
||
├── 🚨 告警中心 /alert-center (已有)
|
||
└── 📜 告警规则 /agent-alerts (已有)
|
||
```
|
||
|
||
---
|
||
|
||
## 执行顺序
|
||
|
||
1. **Phase 1**（最优先，用户当前最需要）
|
||
→ Agent 扩展网络/磁盘IO/进程/系统信息
|
||
→ 前端主机监控面板
|
||
|
||
2. **Phase 2** → 中间件监控
|
||
|
||
3. **Phase 3** → SNMP 设备监控（较复杂，需 gosnmp 依赖）
|
||
|
||
4. **Phase 4** → 站点拨测
|
||
|
||
5. **Phase 5** → 导航重构（收尾）
|