Linux 例程：Shell 脚本——服务器资源监控

目标

写一个实用的 Bash 监控脚本，展示 Linux 管道、重定向、grep/awk/sed 核心技法。

完整脚本

#!/bin/bash
# monitor.sh —— 轻量服务器资源监控

set -euo pipefail  # 严格模式：遇错即停、未定义变量报错、管道失败报错

# ── 配置 ──
LOG_FILE="/tmp/monitor_$(date +%Y%m%d).log"
ALERT_CPU=80        # CPU 告警阈值 %
ALERT_MEM=90        # 内存告警阈值 %

# ── 工具函数 ──
timestamp() {
    date "+%Y-%m-%d %H:%M:%S"
}

log() {
    echo "[$(timestamp)] $*" | tee -a "$LOG_FILE"
}

# ── 监控项 ──
check_cpu() {
    # 从 /proc/stat 计算 CPU 使用率
    local cpu_line
    cpu_line=$(grep '^cpu ' /proc/stat)
    local user system idle
    read -r _ user _ system _ idle _ <;<< "$cpu_line"
    local total=$((user + system + idle))
    local used=$((user + system))

    # 等 1 秒再读一次（计算差值才能得百分比）
    sleep 1
    cpu_line=$(grep '^cpu ' /proc/stat)
    read -r _ user2 _ system2 _ idle2 _ <;<< "$cpu_line"
    local total2=$((user2 + system2 + idle2))
    local used2=$((user2 + system2))

    local cpu_pct=$(( 100 * (used2 - used) / (total2 - total) ))
    echo "$cpu_pct"
}

check_memory() {
    local total used free pct
    read -r total used free <;<< "$(free -m | awk '/^Mem:/ {print $2, $3, $4}')"
    pct=$(( used * 100 / total ))
    log "内存: ${used}MB / ${total}MB (${pct}%)"
    if [ "$pct" -ge "$ALERT_MEM" ]; then
        log "⚠️  内存使用率过高: ${pct}%"
    fi
}

check_disk() {
    df -h / /home 2>;/dev/null | awk 'NR>1 {
        gsub(/%/,"",$5)
        printf "磁盘 %s: %s / %s (%d%%)\n", $6, $3, $2, $5
        if ($5+0 >; 85) print "⚠️  磁盘空间不足: " $6
    }&#39; | while read -r line; do log "$line"; done
}

check_network() {
    local conns
    conns=$(ss -tun state established | tail -n +2 | wc -l)
    log "活跃 TCP 连接数: $conns"
}

check_top_processes() {
    log "=== TOP 5 进程（CPU）==="
    ps aux --sort=-%cpu | head -6 | tail -5 | while read -r line; do
        local user pid cpu mem cmd
        read -r user pid cpu mem _ _ _ _ _ _ _ cmd <;<< "$line"
        log "  PID=${pid} CPU=${cpu}% MEM=${mem}% CMD=${cmd:0:50}"
    done
}

# ── 主循环 ──
main() {
    log "======== 监控开始 ========"

    local cpu
    cpu=$(check_cpu)
    log "CPU 使用率: ${cpu}%"
    if [ "$cpu" -ge "$ALERT_CPU" ]; then
        log "⚠️  CPU 使用率过高: ${cpu}%"
    fi

    check_memory
    check_disk
    check_network
    check_top_processes

    log "======== 监控结束 ========"
}

# 如果传入 --watch 参数，循环执行
if [ "${1:-}" = "--watch" ]; then
    while true; do
        clear
        main
        sleep 5
    done
else
    main
fi

运行步骤

chmod +x monitor.sh
./monitor.sh           # 执行一次
./monitor.sh --watch   # 持续监控

预期输出

[2024-01-15 14:30:00] ======== 监控开始 ========
[2024-01-15 14:30:01] CPU 使用率: 23%
[2024-01-15 14:30:01] 内存: 2048MB / 8192MB (25%)
[2024-01-15 14:30:01] 磁盘 /: 45G / 100G (45%)
[2024-01-15 14:30:01] 活跃 TCP 连接数: 12
[2024-01-15 14:30:01] ======== 监控结束 ========

关键 Shell 技法

技法	示例
管道	`ps aux
重定向	echo log >> file.log 2>&1
命令替换	$(date)
Here String	<<< "$variable"
条件判断	[ "$a" -ge 80 ]
循环读取	while read -r line; do ... done