轻量监控：不装 Prometheus 也能知道机器什么时候出事

# /etc/logrotate.d/nginx（默认已有，确认一下）
/var/log/nginx/*.log {
    daily
    rotate 14
    compress
    delaycompress
    missingok
    notifempty
    sharedscripts
    postrotate
        [ -f /run/nginx.pid ] && kill -USR1 $(cat /run/nginx.pid)
    endscript
}

logrotate -d /etc/logrotate.d/nginx   # 调试模式，看会做什么
logrotate -f /etc/logrotate.d/nginx   # 强制执行一次

# journald 也要限制
sudo journalctl --vacuum-size=300M

#!/usr/bin/env bash
# /opt/scripts/watch.sh
THRESHOLD=80
usage=$(df -h / | awk "NR==2 {print $5}" | tr -d "%")

if [ "$usage" -gt "$THRESHOLD" ]; then
  echo "[$(date)] disk usage ${usage}% over ${THRESHOLD}%" >> /var/log/watch.log
  # 这里换成你自己的通知方式
  curl -s -X POST "https://your-notify-hook" \
    -d "disk ${usage}% on $(hostname)" > /dev/null
fi

chmod +x /opt/scripts/watch.sh
crontab -e
*/10 * * * * /opt/scripts/watch.sh

#!/usr/bin/env bash
# /opt/scripts/healthcheck.sh
URL="https://www.bzii.cn"
code=$(curl -o /dev/null -s -w "%{http_code}" --max-time 10 "$URL")
cost=$(curl -o /dev/null -s -w "%{time_total}" --max-time 10 "$URL")

if [ "$code" != "200" ]; then
  echo "[$(date)] site down, code=$code" >> /var/log/health.log
  systemctl restart nginx    # 简单的自愈尝试
fi

# 响应超过 3 秒也记一笔，慢也是故障
awk -v c="$cost" "BEGIN{exit !(c>3)}" && echo "[$(date)] slow: ${cost}s" >> /var/log/health.log

#!/usr/bin/env bash
end=$(echo | openssl s_client -connect www.bzii.cn:443 -servername www.bzii.cn 2>/dev/null \
  | openssl x509 -noout -enddate | cut -d= -f2)
left=$(( ( $(date -d "$end" +%s) - $(date +%s) ) / 86400 ))
echo "证书剩余天数: $left"
[ "$left" -lt 15 ] && echo "证书即将过期！" >> /var/log/health.log

apt install -y btop
btop

echo "负载: $(cat /proc/loadavg | cut -d" " -f1-3) | 内存: $(free -m | awk "NR==2{printf "%.0f%%", $3*100/$2}") | 磁盘: $(df -h / | awk "NR==2{print $5}")"