定期监测 Proxmox VE 系统状态并推送到企业微信

前言

Shell 脚本可以帮助你自动化任务，让你专注于更重要的工作。在本文中，我们将介绍如何编写一个 Shell 脚本，用于定期获取 Proxmox VE 系统的运行状态，结合之前介绍的搭建企业微信推送，并将这些信息推送到企业微信。

脚本功能

• 获取 CPU 使用情况
• 获取内存使用情况
• 获取 IO 使用情况
• 获取网络流量信息
• 获取存储使用情况
• 获取虚拟机信息
• 获取容器信息

脚本实现

我们将使用 Bash 脚本编写此自动化脚本。以下是脚本的实现：

#!/bin/bash

# 获取 CPU 使用情况
print_cpu_info() {
    # Get CPU load
    cpu_load=$(uptime | awk -F'load average:' '{ gsub(/^ +| +$/, "", $2); print $2 }')

    # Print CPU section
    echo "****** CPU ******"
    top -bn1 | grep "Cpu(s)" | \
    awk '{print "👽用户   "$2"%\n🛠️系统   "$4"%\n😴空闲   "$8"%"}'
    printf "🏋️‍♂️负载   %s$cpu_load\n"
    echo
}

# 获取内存使用情况
print_memory_info() {
    # Get Memory usage
    memory_info=$(free -h | grep Mem | awk '{print $3 "\t" $4 "\t" $5 "\t" $6 "\t" $7}')

    # Print Memory section
    echo "****** 内存 ******"
    printf "🗂️已用   %s\n" "$(echo ${memory_info} | awk '{print $1 }')"
    printf "🆓剩余   %s\n" "$(echo ${memory_info} | awk '{print $2 }')"
    printf "♻️共享   %s\n" "$(echo ${memory_info} | awk '{print $3 }')"
    printf "📦缓存   %s\n" "$(echo ${memory_info} | awk '{print $4 }')"
    printf "📤可用   %s\n" "$(echo ${memory_info} | awk '{print $5 }')"
    echo
}

# 获取 IO 使用情况
print_io_info() {
    echo "****** IO ******"
    physical_disks_list=$(lsblk -d -o NAME,TYPE | awk '$2 == "disk" {print $1}')
    for disk in $physical_disks_list; do
        printf "💾磁盘   $disk\n"
        iostat -dx 1 1 | awk '{print $disk  $4 $5 $10}' | grep $disk | \
        awk '{printf "📑读取   %.2fr/s\n📝写入   %.2fw/s\n🕔响应   %.2fms\n\n", $4, $5, $10}'
    done
    echo
}

# 获取网络流量信息
print_traffic_info() {
    # Get Traffic information
    # 设置桥接口名称
    # TODO
    # 根据实际情况修改桥接口名称
    
    bridges=("vmbr0" "vmbr1")    
    echo "****** 流量 ******"
    
    for bridge in "${bridges[@]}"; do
        printf "🌐接口   $bridge\n"
        printf "⬇️接收   %.2fG\n" "$(cat /sys/class/net/$bridge/statistics/rx_bytes | awk '{sum += $1} END {print sum / (1024 * 1024)}')"
        printf "⬆️发送   %.2fG\n" "$(cat /sys/class/net/$bridge/statistics/tx_bytes | awk '{sum += $1} END {print sum / (1024 * 1024)}')"
        printf "\n"
    done
    echo
}

# 获取存储使用情况
print_storage_info() {
    # Get Storage information
    # Run the pvesm status command and store the output
    status=$(/usr/sbin/pvesm status)
    
    # Print the header
    echo "****** 存储 ******"
    
    # Process each line of the output
    echo "$status" | awk 'NR>1 {
        # Convert sizes to GB
        size_gb = $4 / (1024*1024)
        used_gb = $5 / (1024*1024)
        avail_gb = $6 / (1024*1024)
    
        # Print the formatted line
        printf "📁逻辑卷   %s\n", $1
        printf "🔎总大小   %.2fG\n🗂️已使用   %.2fG\n🆓未使用   %.2fG\n📊使用率   %.2f%%\n\n", size_gb, used_gb, avail_gb, $7
    }'
}

# 获取虚拟机信息
print_vm_info() {
    # Get VM information
    # Print VM section
    echo "****** 虚拟机 ******"
    # Run the command to get the list of VMs and process the output
    pvesh get /cluster/resources --output-format json | jq -r '.[] | select(.type == "qemu" and .status == "running") | [.type, .name, .vmid, .cpu, .mem, .maxmem, .disk, .maxdisk, .diskread, .diskwrite, .netin, .netout, .uptime] | @tsv' | awk -F'\t' '{
        # Convert diskread, diskwrite, netin, and netout to GB
        diskread_gb = $9 / (1024 * 1024 * 1024)
        diskwrite_gb = $10 / (1024 * 1024 * 1024)
        netin_gb = $11 / (1024 * 1024)
        netout_gb = $12 / (1024 * 1024)

        # Convert uptime to a more readable format
        uptime = $13
        days = int(uptime / 86400)
        uptime -= days * 86400
        hours = int(uptime / 3600)
        uptime -= hours * 3600
        minutes = int(uptime / 60)
        uptime -= minutes * 60
        seconds = int(uptime)

        uptime_str = ""
        if (days > 0) uptime_str = days "d "
        if (hours > 0) uptime_str = uptime_str hours "h "
        if (minutes > 0) uptime_str = uptime_str minutes "m "
        uptime_str = uptime_str seconds "s"

        # Print the formatted line
        printf "🆔编号 %s\n", $3
        printf "💻名称 %s\n", $2
        printf "#️⃣CPU %.2f%%\n", $4
        printf "🧱内存 %.2f%%\n", ($5 / $6) * 100
        #printf "📦硬盘 %.2f%%\n", ($7 / $8) * 100
        printf "📑读取 %.2fMi\n", diskread_gb
        printf "📝写入 %.2fMi\n", diskwrite_gb
        printf "⬇️接收 %.2fMi\n", netin_gb
        printf "⬆️发送 %.2fMi\n", netout_gb
        printf "⏳运行 %s\n", uptime_str
        printf "\n"
    }'
}

# 获取容器信息
print_lxc_info() {
    # Get VM information
    # Print VM section
    echo "****** 容器 ******"

    # Run the command to get the list of VMs and process the output
    pvesh get /cluster/resources --output-format json | jq -r '.[] | select(.type == "lxc" and .status == "running") | [.type, .name, .vmid, .cpu, .mem, .maxmem, .disk, .maxdisk, .diskread, .diskwrite, .netin, .netout, .uptime] | @tsv' | awk -F'\t' '{
        # Convert diskread, diskwrite, netin, and netout to GB
        diskread_gb = $9 / (1024 * 1024 * 1024)
        diskwrite_gb = $10 / (1024 * 1024 * 1024)
        netin_gb = $11 / (1024 * 1024)
        netout_gb = $12 / (1024 * 1024)

        # Convert uptime to a more readable format
        uptime = $13
        days = int(uptime / 86400)
        uptime -= days * 86400
        hours = int(uptime / 3600)
        uptime -= hours * 3600
        minutes = int(uptime / 60)
        uptime -= minutes * 60
        seconds = int(uptime)

        uptime_str = ""
        if (days > 0) uptime_str = days "d "
        if (hours > 0) uptime_str = uptime_str hours "h "
        if (minutes > 0) uptime_str = uptime_str minutes "m "
        uptime_str = uptime_str seconds "s"

        # Print the formatted line
        printf "🆔编号 %s\n", $3
        printf "🐋名称 %s\n", $2
        printf "#️⃣CPU %.2f%%\n", $4
        printf "🧱内存 %.2f%%\n", ($5 / $6) * 100
        printf "📦硬盘 %.2f%%\n", ($7 / $8) * 100
        printf "📑读取 %.2fMi\n", diskread_gb
        printf "📝写入 %.2fMi\n", diskwrite_gb
        printf "⬇️接收 %.2fMi\n", netin_gb
        printf "⬆️发送 %.2fMi\n", netout_gb
        printf "⏳运行 %s\n", uptime_str
        printf "\n"
    }'
}

# SendMessage 函数用于将信息推送到企业微信
send_message() {
    # 使用 curl 发送 HTTP 请求将信息推送到企业微信应用
    # TODO 
    #这里根据实际公网服务器IP和token内容进行修改
    #
    curl "http://localhost:5000/?token=9a537f5d&msg=%F0%9F%8E%97%EF%B8%8FPVE%E7%B3%BB%E7%BB%9F%E7%9B%91%E6%B5%8B%0a%0a$1"  > /dev/null 2>&1
}

# 主脚本
functions=("print_cpu_info" "print_memory_info" "print_io_info" "print_traffic_info" "print_storage_info" "print_vm_info" "print_lxc_info")
for func in "${functions[@]}"; do
    # 调用各个函数获取信息，并将输出发送到企业微信
    combined_output=$(printf "%s\n%s" "$($func)")
    encoded_string=$(python3 -c "import urllib.parse; print(urllib.parse.quote('''$combined_output'''))")
    send_message "$encoded_string"
done

使用说明

1. 将脚本保存为 monitor.sh。
2. 确保你的系统中安装了所需的依赖，如 curl、python3、jq 等。
3. 设置一个定时任务，以便定期执行该脚本。你可以使用 cron 作业来实现这一点。例如，如果你希望每小时执行一次，可以添加以下 cron 作业：

0 * * * * /path/to/monitor.sh

这将使脚本每小时执行一次，并将最新的 Proxmox VE 系统状态推送到企业微信。

由于企业微信单条信息有长度限制，所以使用每个查询接口单独发送，效果如图

通过这个脚本，你可以轻松地监控 Proxmox VE 系统的运行状态，并在必要时采取相应的行动。