🔖 dashboard v0.10.0

2025-01-22 12:48:14 -05:00 · 2021-09-27 21:18:09 +08:00 · 2021-09-27 21:18:09 +08:00 · 4249fa82d7
commit 4249fa82d7
parent ac3cfa162c
11 changed files with 81 additions and 69 deletions
--- a/README.md
+++ b/README.md
@ -4,7 +4,7 @@
  <br>
  <small><i>LOGO designed by <a href="https://xio.ng" target="_blank">熊大</a> .</i></small>
  <br><br>
-<img src="https://img.shields.io/github/workflow/status/naiba/nezha/Dashboard%20image?label=Dash%20v0.9.34&logo=github&style=for-the-badge">&nbsp;<img src="https://img.shields.io/github/v/release/naiba/nezha?color=brightgreen&label=Agent&style=for-the-badge&logo=github">&nbsp;<img src="https://img.shields.io/github/workflow/status/naiba/nezha/Agent%20release?label=Agent%20CI&logo=github&style=for-the-badge">&nbsp;<img src="https://img.shields.io/badge/Installer-v0.7.0-brightgreen?style=for-the-badge&logo=linux">
+<img src="https://img.shields.io/github/workflow/status/naiba/nezha/Dashboard%20image?label=Dash%20v0.10.0&logo=github&style=for-the-badge">&nbsp;<img src="https://img.shields.io/github/v/release/naiba/nezha?color=brightgreen&label=Agent&style=for-the-badge&logo=github">&nbsp;<img src="https://img.shields.io/github/workflow/status/naiba/nezha/Agent%20release?label=Agent%20CI&logo=github&style=for-the-badge">&nbsp;<img src="https://img.shields.io/badge/Installer-v0.7.0-brightgreen?style=for-the-badge&logo=linux">
  <br>
  <br>
  <p>:trollface: <b>哪吒监控</b> 一站式轻监控轻运维系统。支持系统状态、HTTP(SSL 证书变更、即将到期、到期)、TCP、Ping 监控报警，命令批量执行和计划任务。</p>
@ -20,7 +20,7 @@

 ## 安装脚本

-**推荐配置：** 安装前准备 _两个域名_，一个可以 **接入 CDN** 作为 _公开访问_，比如 (status.nai.ba)；另外一个解析到面板服务器作为 Agent 连接 Dashboard 使用，**不能接入 CDN** 直接暴露面板主机 IP，比如（randomdashboard.nai.ba）。
+**推荐配置：** 安装前准备 _两个域名_，一个可以 **接入 CDN** 作为 _公开访问_，比如 (status.nai.ba)；另外一个解析到面板服务器作为 Agent 连接 Dashboard 使用，**不能接入 CDN** 直接暴露面板主机 IP，比如（ip-to-dashboard.nai.ba）。

 ```shell
 curl -L https://raw.githubusercontent.com/naiba/nezha/master/script/install.sh  -o nezha.sh && chmod +x nezha.sh
@ -36,6 +36,13 @@ CN=true sudo ./nezha.sh

 _\* 使用 WatchTower 可以自动更新面板，Windows 终端可以使用 nssm 配置自启动（见尾部教程）_

+### 特殊技能
+
+编辑 `/etc/systemd/system/nezha-agent.service`，在 `ExecStart=` 这一行的末尾加上
+
+- `--skip-conn` 不监控连接数，机场/连接密集型机器推荐设置，不然比较占CPU([shirou/gopsutil/issues#220](https://github.com/shirou/gopsutil/issues/220))
+- `--disable-auto-update` 禁止 Agent 自动更新
+
 ## 功能说明

 <details>
@ -63,7 +70,7 @@ URL 里面也可放置占位符，请求时会进行简单的字符串替换。
   - server 酱示例

     - 名称：server 酱
-     - URL：https://sc.ftqq.com/SCUrandomkeys.send?text=#NEZHA#
+     - URL：<https://sc.ftqq.com/SCUrandomkeys.send?text=#NEZHA>#
     - 请求方式: GET
     - 请求类型: 默认
     - Body: 空
@ -71,7 +78,7 @@ URL 里面也可放置占位符，请求时会进行简单的字符串替换。
   - wxpusher 示例，需要关注你的应用

     - 名称: wxpusher
-     - URL：http://wxpusher.zjiecode.com/api/send/message
+     - URL：<http://wxpusher.zjiecode.com/api/send/message>
     - 请求方式: POST
     - 请求类型: JSON
     - Body: `{"appToken":"你的appToken","topicIds":[],"content":"#NEZHA#","contentType":"1","uids":["你的uid"]}`
@ -79,7 +86,7 @@ URL 里面也可放置占位符，请求时会进行简单的字符串替换。
   - telegram 示例 [@haitau](https://github.com/haitau) 贡献

     - 名称：telegram 机器人消息通知
-     - URL：https://api.telegram.org/botXXXXXX/sendMessage?chat_id=YYYYYY&text=#NEZHA#
+     - URL：<https://api.telegram.org/botXXXXXX/sendMessage?chat_id=YYYYYY&text=#NEZHA>#
     - 请求方式: GET
     - 请求类型: 默认
     - Body: 空
@ -130,6 +137,7 @@ URL 里面也可放置占位符，请求时会进行简单的字符串替换。
 - cycle_interval 小时（可以设为 1 月，30\*24）
 - min/max、cover、ignore 参考基本规则配置
 - 示例: 3 号机器的每月 15 号计费的出站月流量 1T 报警 `[{"type":"transfer_out_cycle","max":1000000000000,"cycle_start":"2021-07-15T08:00:00Z","cycle_interval":720,"cover":1,"ignore":{"3":true}}]`
+
 </details>

 <details>
@ -221,19 +229,13 @@ URL 里面也可放置占位符，请求时会进行简单的字符串替换。

 </details>

-<details>
-    <summary>如何禁用网络连接数监控？</summary>
-
-编辑 `/etc/systemd/system/nezha-agent.service`，在 `ExecStart=` 这一行的末尾加上 `--skip-conn` 就是不监控连接数
-
-</details>
-
 <details>
    <summary>Agent 不断重启/无法启动 ？</summary>

 1. 直接执行 `/opt/nezha/agent/nezha-agent -s 面板IP或非CDN域名:面板RPC端口 -p Agent密钥 -d` 查看日志是否是 DNS 问题。
-2. `nc -v 域名/IP 面板RPC端口` 或者 `telnet 域名/IP 面板RPC端口` 检验是否是网络问题，检查本机与面板服务器出入站防火墙，如果单机无法判断可借助 https://port.ping.pe/ 提供的端口检查工具进行检测。
+2. `nc -v 域名/IP 面板RPC端口` 或者 `telnet 域名/IP 面板RPC端口` 检验是否是网络问题，检查本机与面板服务器出入站防火墙，如果单机无法判断可借助 <https://port.ping.pe/> 提供的端口检查工具进行检测。
 3. 如果上面步骤检测正常，Agent 正常上线，尝试关闭 SELinux，[如何关闭 SELinux？](https://www.google.com/search?q=%E5%85%B3%E9%97%ADSELINUX)
+
 </details>

 <details>
--- a/cmd/agent/main.go
+++ b/cmd/agent/main.go
@ -6,7 +6,6 @@ import (
 	"errors"
 	"fmt"
 	"io"
-	"log"
 	"net"
 	"net/http"
 	"os"
@ -29,15 +28,22 @@ import (
 	"github.com/naiba/nezha/service/rpc"
 )

+type AgentConfig struct {
+	SkipConnectionCount bool
+	DisableAutoUpdate   bool
+	Debug               bool
+	Server              string
+	ClientSecret        string
+}
+
 func init() {
 	http.DefaultClient.Timeout = time.Second * 5
 	flag.CommandLine.ParseErrorsWhitelist.UnknownFlags = true
 }

 var (
-	server, clientSecret, version string
-	debug                         bool
-	stateConf                     monitor.GetStateConfig
+	version   string
+	agentConf AgentConfig
 )

 var (
@ -61,13 +67,14 @@ func main() {
 	// 来自于 GoReleaser 的版本号
 	monitor.Version = version

-	flag.BoolVarP(&debug, "debug", "d", false, "开启调试信息")
-	flag.StringVarP(&server, "*server", "s", "localhost:5555", "管理面板RPC端口")
-	flag.StringVarP(&clientSecret, "password", "p", "", "Agent连接Secret")
-	flag.BoolVar(&stateConf.SkipConnectionCount, "skip-conn", false, "不监控连接数")
+	flag.BoolVarP(&agentConf.Debug, "debug", "d", false, "开启调试信息")
+	flag.StringVarP(&agentConf.Server, "*server", "s", "localhost:5555", "管理面板RPC端口")
+	flag.StringVarP(&agentConf.ClientSecret, "password", "p", "", "Agent连接Secret")
+	flag.BoolVar(&agentConf.SkipConnectionCount, "skip-conn", false, "不监控连接数")
+	flag.BoolVar(&agentConf.DisableAutoUpdate, "disable-auto-update", false, "禁用自动升级")
 	flag.Parse()

-	if clientSecret == "" {
+	if agentConf.ClientSecret == "" {
 		flag.Usage()
 		return
 	}
@ -77,7 +84,7 @@ func main() {

 func run() {
 	auth := rpc.AuthHandler{
-		ClientSecret: clientSecret,
+		ClientSecret: agentConf.ClientSecret,
 	}

 	go pty.DownloadDependency()
@ -87,7 +94,7 @@ func run() {
 	// 更新IP信息
 	go monitor.UpdateIP()

-	if _, err := semver.Parse(version); err == nil {
+	if _, err := semver.Parse(version); err == nil && !agentConf.DisableAutoUpdate {
 		go func() {
 			for range updateCh {
 				go doSelfUpdate()
@ -111,7 +118,7 @@ func run() {

 	for {
 		timeOutCtx, cancel := context.WithTimeout(context.Background(), networkTimeOut)
-		conn, err = grpc.DialContext(timeOutCtx, server, grpc.WithInsecure(), grpc.WithPerRPCCredentials(&auth))
+		conn, err = grpc.DialContext(timeOutCtx, agentConf.Server, grpc.WithInsecure(), grpc.WithPerRPCCredentials(&auth))
 		if err != nil {
 			println("与面板建立连接失败：", err)
 			cancel()
@ -194,7 +201,7 @@ func reportState() {
 		if client != nil && inited {
 			monitor.TrackNetworkSpeed()
 			timeOutCtx, cancel := context.WithTimeout(context.Background(), networkTimeOut)
-			_, err = client.ReportSystemState(timeOutCtx, monitor.GetState(stateConf).PB())
+			_, err = client.ReportSystemState(timeOutCtx, monitor.GetState(agentConf.SkipConnectionCount).PB())
 			cancel()
 			if err != nil {
 				println("reportState error", err)
@ -334,7 +341,7 @@ func handleTerminalTask(task *pb.Task) {
 		protocol += "s"
 	}
 	header := http.Header{}
-	header.Add("Secret", clientSecret)
+	header.Add("Secret", agentConf.ClientSecret)
 	conn, _, err := websocket.DefaultDialer.Dial(fmt.Sprintf("%s://%s/terminal/%s", protocol, terminal.Host, terminal.Session), header)
 	if err != nil {
 		println("Terminal 连接失败：", err)
@ -404,7 +411,8 @@ func handleTerminalTask(task *pb.Task) {
 }

 func println(v ...interface{}) {
-	if debug {
-		log.Println(v...)
+	if agentConf.Debug {
+		fmt.Printf("NEZHA@%s>> ", time.Now().Format("2006-01-02 15:04:05"))
+		fmt.Println(v...)
 	}
 }
--- a/cmd/agent/monitor/monitor.go
+++ b/cmd/agent/monitor/monitor.go
@ -74,11 +74,7 @@ func GetHost() *model.Host {
 	}
 }

-type GetStateConfig struct {
-	SkipConnectionCount bool
-}
-
-func GetState(conf GetStateConfig) *model.HostState {
+func GetState(skipConnectionCount bool) *model.HostState {
 	hi, _ := host.Info()
 	mv, _ := mem.VirtualMemory()

@ -101,7 +97,7 @@ func GetState(conf GetStateConfig) *model.HostState {

 	var tcpConnCount, udpConnCount uint64

-	if !conf.SkipConnectionCount {
+	if !skipConnectionCount {
 		conns, _ := net.Connections("all")
 		for i := 0; i < len(conns); i++ {
 			switch conns[i].Type {
--- a/cmd/dashboard/controller/common_page.go
+++ b/cmd/dashboard/controller/common_page.go
@ -271,8 +271,8 @@ func (cp *commonPage) terminal(c *gin.Context) {
 	}
 	defer conn.Close()

-	log.Println("terminal connected", isAgent, c.Request.URL)
-	defer log.Println("terminal disconnected", isAgent, c.Request.URL)
+	log.Println("NEZHA>> terminal connected", isAgent, c.Request.URL)
+	defer log.Println("NEZHA>> terminal disconnected", isAgent, c.Request.URL)

 	if isAgent {
 		terminal.agentConn = conn
--- a/cmd/dashboard/main.go
+++ b/cmd/dashboard/main.go
@ -198,7 +198,7 @@ func main() {
 			}
 		}
 		if err := dao.DB.Create(txs).Error; err != nil {
-			log.Println("流量统计入库", err)
+			log.Println("NEZHA>> 流量统计入库", err)
 		}
 		srv.Shutdown(c)
 		return nil
--- a/model/monitor.go
+++ b/model/monitor.go
@ -73,3 +73,7 @@ func (m *Monitor) AfterFind(tx *gorm.DB) error {
 	}
 	return nil
 }
+
+func IsServiceSentinelNeeded(t uint64) bool {
+	return t == TaskTypeCommand || t == TaskTypeTerminal
+}
--- a/service/dao/alertsentinel.go
+++ b/service/dao/alertsentinel.go
@ -54,7 +54,7 @@ func AlertSentinelStart() {
 		checkCount++
 		if lastPrint.Before(startedAt.Add(-1 * time.Hour)) {
 			if Conf.Debug {
-				log.Println("报警规则检测每小时", checkCount, "次", startedAt, time.Now())
+				log.Println("NEZHA>> 报警规则检测每小时", checkCount, "次", startedAt, time.Now())
 			}
 			checkCount = 0
 			lastPrint = startedAt
@ -114,11 +114,11 @@ func checkStatus() {
 			max, passed := alert.Check(alertsStore[alert.ID][server.ID])
 			if !passed {
 				alertsPrevState[alert.ID][server.ID] = _RuleCheckFail
-				message := fmt.Sprintf("报警规则：%s，服务器：%s(%s)，逮到咯，快去看看！", alert.Name, server.Name, utils.IPDesensitize(server.Host.IP))
+				message := fmt.Sprintf("[主机故障] %s(%s) 规则：%s，", server.Name, utils.IPDesensitize(server.Host.IP), alert.Name)
 				go SendNotification(message, true)
 			} else {
 				if alertsPrevState[alert.ID][server.ID] == _RuleCheckFail {
-					message := fmt.Sprintf("报警规则：%s，服务器：%s(%s)，已恢复正常", alert.Name, server.Name, utils.IPDesensitize(server.Host.IP))
+					message := fmt.Sprintf("[主机恢复] %s(%s) 规则：%s", server.Name, utils.IPDesensitize(server.Host.IP), alert.Name)
 					go SendNotification(message, true)
 				}
 				alertsPrevState[alert.ID][server.ID] = _RuleCheckPass
--- a/service/dao/dao.go
+++ b/service/dao/dao.go
@ -13,7 +13,7 @@ import (
 	pb "github.com/naiba/nezha/proto"
 )

-var Version = "v0.9.34" // ！！记得修改 README 中的 badge 版本！！
+var Version = "v0.10.0" // ！！记得修改 README 中的 badge 版本！！

 var (
 	Conf  *model.Config
@ -64,7 +64,7 @@ func ManualTrigger(c *model.Cron) {
 				Type: model.TaskTypeCommand,
 			})
 		} else {
-			SendNotification(fmt.Sprintf("计划任务：%s，服务器：%s 离线，无法执行。", c.Name, ServerList[c.Servers[j]].Name), false)
+			SendNotification(fmt.Sprintf("[任务失败] %s，服务器 %s 离线，无法执行。", c.Name, ServerList[c.Servers[j]].Name), false)
 		}
 	}
 }
@ -91,7 +91,7 @@ func CronTrigger(cr model.Cron) func() {
 					Type: model.TaskTypeCommand,
 				})
 			} else {
-				SendNotification(fmt.Sprintf("计划任务：%s，服务器：%s 离线，无法执行。", cr.Name, s.Name), false)
+				SendNotification(fmt.Sprintf("[任务失败] %s，服务器 %s 离线，无法执行。", cr.Name, s.Name), false)
 			}
 		}
 	}
--- a/service/dao/notification.go
+++ b/service/dao/notification.go
@ -71,7 +71,7 @@ func SendNotification(desc string, muteable bool) {

 		if !flag {
 			if Conf.Debug {
-				log.Println("静音的重复通知：", desc, muteable)
+				log.Println("NEZHA>> 静音的重复通知：", desc, muteable)
 			}
 			return
 		}
@ -81,7 +81,7 @@ func SendNotification(desc string, muteable bool) {
 	defer notificationsLock.RUnlock()
 	for i := 0; i < len(notifications); i++ {
 		if err := notifications[i].Send(desc); err != nil {
-			log.Println("发送通知失败：", err)
+			log.Println("NEZHA>> 发送通知失败：", err)
 		}
 	}
 }
--- a/service/dao/servicesentinel.go
+++ b/service/dao/servicesentinel.go
@ -150,6 +150,7 @@ func (ss *ServiceSentinel) loadMonitorHistory() {
 		monitors[i].CronJobID, err = ss.dispatchCron.AddFunc(task.CronSpec(), func() {
 			ss.dispatchBus <- task
 		})
+		log.Println("NEZHA>> 服务监控任务", monitors[i].ID, monitors[i].Name, monitors[i].CronJobID)
 		if err != nil {
 			panic(err)
 		}
@ -282,7 +283,8 @@ func getStateStr(percent uint64) string {

 func (ss *ServiceSentinel) worker() {
 	for r := range ss.serviceReportChannel {
-		if ss.monitors[r.Data.GetId()].ID == 0 {
+		if ss.monitors[r.Data.GetId()] == nil || ss.monitors[r.Data.GetId()].ID == 0 {
+			log.Printf("NEZAH>> 错误的服务监控上报 %+v", r)
 			continue
 		}
 		mh := model.PB2MonitorHistory(r.Data)
@ -315,7 +317,7 @@ func (ss *ServiceSentinel) worker() {
 			ss.serviceCurrentStatusIndex[mh.MonitorID] = 0
 			dataToSave := ss.serviceCurrentStatusData[mh.MonitorID]
 			if err := DB.Create(&dataToSave).Error; err != nil {
-				log.Println("服务监控数据持久化失败：", err)
+				log.Println("NEZHA>> 服务监控数据持久化失败：", err)
 			}
 		}
 		// 更新当前状态
@ -337,7 +339,7 @@ func (ss *ServiceSentinel) worker() {
 		stateStr := getStateStr(upPercent)
 		if !mh.Successful {
 			ServerLock.RLock()
-			log.Println("服务故障上报：", ss.monitors[mh.MonitorID].Target, stateStr, "上报者：", ServerList[r.Reporter].Name, "请求输出：", mh.Data)
+			log.Println("NEZHA>> 服务故障上报：", ss.monitors[mh.MonitorID].Target, stateStr, "上报者：", ServerList[r.Reporter].Name, "请求输出：", mh.Data)
 			ServerLock.RUnlock()
 		}
 		if stateStr == "故障" || stateStr != ss.lastStatus[mh.MonitorID] {
@ -345,7 +347,7 @@ func (ss *ServiceSentinel) worker() {
 			isNeedSendNotification := (ss.lastStatus[mh.MonitorID] != "" || stateStr == "故障") && ss.monitors[mh.MonitorID].Notify
 			ss.lastStatus[mh.MonitorID] = stateStr
 			if isNeedSendNotification {
-				go SendNotification(fmt.Sprintf("服务监控：%s 服务状态：%s", ss.monitors[mh.MonitorID].Name, stateStr), true)
+				go SendNotification(fmt.Sprintf("[服务%s] %s", stateStr, ss.monitors[mh.MonitorID].Name), true)
 			}
 			ss.monitorsLock.RUnlock()
 		}
@ -389,7 +391,7 @@ func (ss *ServiceSentinel) worker() {
 		if errMsg != "" {
 			ss.monitorsLock.RLock()
 			if ss.monitors[mh.MonitorID].Notify {
-				go SendNotification(fmt.Sprintf("服务监控：%s %s", ss.monitors[mh.MonitorID].Name, errMsg), true)
+				go SendNotification(fmt.Sprintf("[SSL] %s %s", ss.monitors[mh.MonitorID].Name, errMsg), true)
 			}
 			ss.monitorsLock.RUnlock()
 		}
--- a/service/rpc/nezha.go
+++ b/service/rpc/nezha.go
@ -21,12 +21,7 @@ func (s *NezhaHandler) ReportTask(c context.Context, r *pb.TaskResult) (*pb.Rece
 	if clientID, err = s.Auth.Check(c); err != nil {
 		return nil, err
 	}
-	if r.GetType() != model.TaskTypeCommand {
-		dao.ServiceSentinelShared.Dispatch(dao.ReportData{
-			Data:     r,
-			Reporter: clientID,
-		})
-	} else {
+	if r.GetType() == model.TaskTypeCommand {
 		// 处理上报的计划任务
 		dao.CronLock.RLock()
 		defer dao.CronLock.RUnlock()
@ -35,16 +30,21 @@ func (s *NezhaHandler) ReportTask(c context.Context, r *pb.TaskResult) (*pb.Rece
 			dao.ServerLock.RLock()
 			defer dao.ServerLock.RUnlock()
 			if cr.PushSuccessful && r.GetSuccessful() {
-				dao.SendNotification(fmt.Sprintf("成功计划任务：%s ，服务器：%s，日志：\n%s", cr.Name, dao.ServerList[clientID].Name, r.GetData()), false)
+				dao.SendNotification(fmt.Sprintf("[任务成功] %s ，服务器：%s，日志：\n%s", cr.Name, dao.ServerList[clientID].Name, r.GetData()), false)
 			}
 			if !r.GetSuccessful() {
-				dao.SendNotification(fmt.Sprintf("失败计划任务：%s ，服务器：%s，日志：\n%s", cr.Name, dao.ServerList[clientID].Name, r.GetData()), false)
+				dao.SendNotification(fmt.Sprintf("[任务失败] %s ，服务器：%s，日志：\n%s", cr.Name, dao.ServerList[clientID].Name, r.GetData()), false)
 			}
 			dao.DB.Model(cr).Updates(model.Cron{
 				LastExecutedAt: time.Now().Add(time.Second * -1 * time.Duration(r.GetDelay())),
 				LastResult:     r.GetSuccessful(),
 			})
 		}
+	} else if model.IsServiceSentinelNeeded(r.GetType()) {
+		dao.ServiceSentinelShared.Dispatch(dao.ReportData{
+			Data:     r,
+			Reporter: clientID,
+		})
 	}
 	return &pb.Receipt{Proced: true}, nil
 }
@ -101,7 +101,7 @@ func (s *NezhaHandler) ReportSystemInfo(c context.Context, r *pb.Host) (*pb.Rece
 		host.IP != "" &&
 		dao.ServerList[clientID].Host.IP != host.IP {
 		dao.SendNotification(fmt.Sprintf(
-			"IP变更提醒 服务器：%s ，旧IP：%s，新IP：%s。",
+			"[IP变更] %s ，旧IP：%s，新IP：%s。",
 			dao.ServerList[clientID].Name, utils.IPDesensitize(dao.ServerList[clientID].Host.IP), utils.IPDesensitize(host.IP)), true)
 	}