区块链散户一枚

Python中毒爱好者

monit 安装 GPT解释

好的。按“从零到跑通”的顺序，把 Monit（无 HTTP 面板、阈值→Webhook） 全流程整理如下。每一步都有明确命令与目标。

0）可选：清理旧环境（已有就跳过）

systemctl stop monit 2>/dev/null || true
systemctl disable monit 2>/dev/null || true
dnf remove -y monit 2>/dev/null || true
rm -rf /etc/monitrc /etc/monit.d /var/lib/monit /var/log/monit.log /usr/local/bin/monit-webhook.sh

1）安装

dnf install -y monit
monit -V     # 确认版本（你的是 5.33.0）

2）准备 Webhook 发送脚本（统一出口）

将 WEBHOOK_URL 改成你的接口（可带固定 token）。

cat >/usr/local/bin/monit-webhook.sh <<'EOF'
#!/usr/bin/env bash
set -euo pipefail
WEBHOOK_URL="https://your-api.example.com/monitor/hook?token=MY_SECRET_TOKEN"  # TODO: 改成你的
payload="${1:-{}}"
/usr/bin/curl -sS -f --connect-timeout 3 --max-time 5 \
  -H 'Content-Type: application/json' \
  -X POST "$WEBHOOK_URL" -d "$payload" >/dev/null || true
EOF
chmod +x /usr/local/bin/monit-webhook.sh

快速自测脚本是否可用：

/usr/local/bin/monit-webhook.sh '{"alert":"SelfTest","status":"firing","host":"'"$(hostname)"'"}'

3）主配置

/etc/monitrc

（关闭 HTTP；30s/轮）

cat >/etc/monitrc <<'EOF'
set daemon 30
set logfile /var/log/monit.log
set idfile /var/lib/monit/id
set statefile /var/lib/monit/state

include /etc/monit.d/*
EOF
chmod 600 /etc/monitrc
mkdir -p /var/lib/monit

说明：monit status/summary 在 5.33 需要 HTTP 面板；我们此处关闭，所以查询状态请看日志（或见下方“可选开启 HTTP 仅限本机”）。

建议开启http服务
可以添加内容

set httpd port 2812 and
use address 127.0.0.1
allow localhost
allow admin:monit

4）规则

/etc/monit.d/alerts.conf

（含抖动抑制与恢复事件）

mkdir -p /etc/monit.d
cat >/etc/monit.d/alerts.conf <<'EOF'
# 1) CPU：5 分钟平均负载 > 1.2 连续 3 轮
check system cpu_high
  if loadavg (5min) > 1.2 for 3 cycles then exec "/usr/local/bin/monit-webhook.sh '{\"alert\":\"HighCPU\",\"status\":\"firing\",\"host\":\"$(hostname)\",\"detail\":\"loadavg5m>1.2\"}'"
  else if succeeded then exec "/usr/local/bin/monit-webhook.sh '{\"alert\":\"HighCPU\",\"status\":\"resolved\",\"host\":\"$(hostname)\",\"detail\":\"recovered\"}'"

# 2) 内存：> 80% 连续 3 轮
check system mem_high
  if memory usage > 80% for 3 cycles then exec "/usr/local/bin/monit-webhook.sh '{\"alert\":\"LowMemory\",\"status\":\"firing\",\"host\":\"$(hostname)\",\"detail\":\"mem>80%\"}'"
  else if succeeded then exec "/usr/local/bin/monit-webhook.sh '{\"alert\":\"LowMemory\",\"status\":\"resolved\",\"host\":\"$(hostname)\",\"detail\":\"recovered\"}'"

# 3) SWAP：> 20% 连续 5 轮
check system swap_high
  if swap usage > 20% for 5 cycles then exec "/usr/local/bin/monit-webhook.sh '{\"alert\":\"HighSwap\",\"status\":\"firing\",\"host\":\"$(hostname)\",\"detail\":\"swap>20%\"}'"
  else if succeeded then exec "/usr/local/bin/monit-webhook.sh '{\"alert\":\"HighSwap\",\"status\":\"resolved\",\"host\":\"$(hostname)\",\"detail\":\"recovered\"}'"

# 4) 根分区：> 80% 连续 10 轮
check filesystem rootfs with path /
  if space usage > 80% for 10 cycles then exec "/usr/local/bin/monit-webhook.sh '{\"alert\":\"DiskUsage\",\"status\":\"firing\",\"host\":\"$(hostname)\",\"path\":\"/\",\"detail\":\">80%\"}'"
  else if succeeded then exec "/usr/local/bin/monit-webhook.sh '{\"alert\":\"DiskUsage\",\"status\":\"resolved\",\"host\":\"$(hostname)\",\"path\":\"/\",\"detail\":\"recovered\"}'"

# 5) 数据盘：/mnt/blockstorage > 80% 连续 10 轮
check filesystem blockstorage with path /mnt/blockstorage
  if space usage > 80% for 10 cycles then exec "/usr/local/bin/monit-webhook.sh '{\"alert\":\"DiskUsage\",\"status\":\"firing\",\"host\":\"$(hostname)\",\"path\":\"/mnt/blockstorage\",\"detail\":\">80%\"}'"
  else if succeeded then exec "/usr/local/bin/monit-webhook.sh '{\"alert\":\"DiskUsage\",\"status\":\"resolved\",\"host\":\"$(hostname)\",\"path\":\"/mnt/blockstorage\",\"detail\":\"recovered\"}'"

# 6) MySQL 端口：127.0.0.1:3306 连续 2 轮失败
check host mysql_tcp with address 127.0.0.1
  if failed port 3306 for 2 cycles then exec "/usr/local/bin/monit-webhook.sh '{\"alert\":\"MySQLPortDown\",\"status\":\"firing\",\"host\":\"$(hostname)\",\"port\":3306}'"
  else if succeeded then exec "/usr/local/bin/monit-webhook.sh '{\"alert\":\"MySQLPortDown\",\"status\":\"resolved\",\"host\":\"$(hostname)\",\"port\":3306}'"
EOF

5）语法检查 → 启动

monit -t                      # 看到“Control file syntax OK”即可
systemctl enable --now monit
monit reload

6）验证运行

关闭 HTTP 面板时，以下方式验证。

ps -ef | grep -E '[m]onit'          # 进程常驻
tail -n 50 /var/log/monit.log       # 启动/重载/告警日志

模拟触发（任选其一，验证 Webhook）：

临时把某条磁盘阈值改成 > 1% → monit reload → 你的接口应收到 firing；改回→resolved。
暂停 MySQL 容器：docker compose down → 约 1 分钟内收到 MySQLPortDown；恢复后收到 resolved。

7）可选：仅本机启用 HTTP（方便

monit status/summary

）

若需要命令行状态查询或本地 Web 面板：

sed -i '/include \/etc\/monit.d\/\*/i set httpd port 2812 and\n    use address 127.0.0.1\n    allow localhost' /etc/monitrc
monit -t && systemctl restart monit

之后可用：

monit summary
monit status
# Web 面板（仅本机）：http://127.0.0.1:2812

常见问题速解

/etc/monitrc:…: syntax error 'env'：5.33 不支持 set env，已用脚本代替。
Program does not exist: '$CURL'：不要用 $CURL 变量；改为调用 /usr/local/bin/monit-webhook.sh。
Error opening idfile：创建 /var/lib/monit 目录即可。
monit status 提示未启用 HTTP：见第 7 步，启用仅本机 HTTP。

需要我把你当前机器的 WEBHOOK_URL 代入，并生成“可直接复制”的两段文件内容（monitrc 与 alerts.conf）吗？

posted @ 2025-09-03 17:36 就是想学习阅读(7) 评论(0) 收藏举报

刷新页面返回顶部