计算给定网络中可用子网的数量

Question

我在计算网络中可用子网（或主机）的数量时遇到以下问题。例如，主子网是 /24 (255.255.255.0)，其中包含两个 /25 (255.255.255.128) 或四个 /26 (255.255.255.192)。

因此，如果一个节点已被使用，则将只有一个 /25 或只有 3 次 /26。那么我应该如何计算可用子网的数量。换句话说，如何获得余数的数量和类型。 我想在 PHP 中执行这个算法。

Answer 1

我使用 python 来解决这个问题（所有代码都附上了，最后是整个脚本），但我确信 PHP 存在类似的解决方案库。由于这是旧的并且没有发布任何解决方案，我猜任何解决方案（在这种情况下是 Python 解决方案）都比没有解决方案要好。

该脚本包含两个函数，一个使用BFS 计算“主”子网中的可用子网，另一个使用graphviz 创建一个pdf 绘图。

我使用的算法是一个非常简单的 BFS。它基本上从顶部子网（“main”，/24）开始，并找出是否有任何使用的（“taken”）子网与其重叠/匹配。如果其中任何一个出现，算法会将其“子代”(/25) 排入队列，以便在同一进程中进行检查。如果算法命中没有“已采取”子节点的子网，则将其标记为“可用”。如果它命中与“已采用”完全匹配的子网，则将其标记为“已采用”。在任何其他情况下，迭代都会继续。

可视化的绘制过程是相同的。

这里是脚本描述（argparse-generated）：

usage: script.py [-h] -m M -t TAKEN [TAKEN ...] -o OUTPUT

optional arguments:
  -h, --help            show this help message and exit
  -m M, --m M           main subnet to check
  -t TAKEN [TAKEN ...], --taken TAKEN [TAKEN ...]
                        taken subnets
  -o OUTPUT, --output OUTPUT
                        graphviz output file name (.pdf)

使用您的参数运行脚本：

subnet_script.py -m 255.255.255.0/24 -t 255.255.255.192/26 -o test

给出这些结果：

做一些更有趣的事情，比如：

script.py -m 255.255.255.0/24 -t 255.255.255.192/26 255.255.255.128/30 -o test

给出这些结果：

我使用了ipaddress、queue 和graphviz（用于python）模块，以及argparse 使脚本更易于使用。

第一个函数：

def get_available_subnets_set(main, taken):
    # we assume no subnets are available intially
    available = []
    q = queue.Queue()
    # add first node for expansion in the BFS process
    q.put(main)

    while q.qsize() > 0:
        subnet = q.get()
        for taken_subnet in taken:
            if taken_subnet.compare_networks(subnet) == 0:
                # found matching subnet in taken, stop expanding
                print("similar: %s and %s" % (subnet, taken_subnet))
                break
            if taken_subnet.overlaps(subnet):
                # still has overlaps somewhere in children, keep expanding
                print("overlaps: %s and %s" % (subnet, taken_subnet))
                for sub_subnet in subnet.subnets():
                    q.put(sub_subnet)
                break
        else:
            # no overlaps with taken - this subnet is entirely available
            available.append(subnet)

    return set(available)

第二个函数使用graphviz以类似的方式绘制结果：

def make_subnet_graph(main, taken_subnets, available_subnets, filename):

    g = graphviz.Graph()
    q = queue.Queue()
    q.put(main)
    g.node(str(main))

    while q.qsize() > 0:
        subnet = q.get()
        for sub_subnet in subnet.subnets():
            if sub_subnet in available_subnets:
                # draw as available (green)
                g.node(str(sub_subnet), _attributes={"color": "green"})
                g.edge(str(sub_subnet), str(subnet))
                continue
            if sub_subnet in taken_subnets:
                # draw as taken (red)
                g.node(str(sub_subnet), _attributes={"color": "red"})
                g.edge(str(sub_subnet), str(subnet))
                continue

            # has mixed type subnets (taken / available) - go deeper
            g.node(str(sub_subnet))
            g.edge(str(sub_subnet), str(subnet))
            q.put(sub_subnet)

    # write file
    g.render(filename)

整个事情连同 argparse:

#!/usr/bin/env python3.4

import ipaddress
import argparse
import queue
import graphviz


def get_available_subnets(main, taken):
    # we assume no subnets are available intially
    available = []
    q = queue.Queue()
    # add first node for expansion in the BFS process
    q.put(main)

    while q.qsize() > 0:
        subnet = q.get()
        for taken_subnet in taken:
            if taken_subnet.compare_networks(subnet) == 0:
                # found matching subnet in taken, stop expanding
                print("similar: %s and %s" % (subnet, taken_subnet))
                break
            if taken_subnet.overlaps(subnet):
                # still has overlaps somewhere in children, keep expanding
                print("overlaps: %s and %s" % (subnet, taken_subnet))
                for sub_subnet in subnet.subnets():
                    q.put(sub_subnet)
                break
        else:
            # no overlaps with taken - this subnet is entirely available
            available.append(subnet)

    return available

def make_subnet_graph(main, taken_subnets, available_subnets, filename):

    g = graphviz.Graph()
    q = queue.Queue()
    q.put(main)
    g.node(str(main))

    while q.qsize() > 0:
        subnet = q.get()
        for sub_subnet in subnet.subnets():
            if sub_subnet in available_subnets:
                # draw as available (green)
                g.node(str(sub_subnet), _attributes={"color": "green"})
                g.edge(str(sub_subnet), str(subnet))
                continue
            if sub_subnet in taken_subnets:
                # draw as taken (red)
                g.node(str(sub_subnet), _attributes={"color": "red"})
                g.edge(str(sub_subnet), str(subnet))
                continue

            # has mixed type subnets (taken / available) - go deeper
            g.node(str(sub_subnet))
            g.edge(str(sub_subnet), str(subnet))
            q.put(sub_subnet)

    # write file
    g.render(filename)


if "__main__" == __name__:
    parser = argparse.ArgumentParser()
    parser.add_argument('-m', '--m', help='main subnet to check', required=True)
    parser.add_argument('-t', '--taken', nargs='+', help='taken subnets', required=True)
    parser.add_argument('-o', '--output', help='graphviz output file name (.pdf)', required=True)
    args = parser.parse_args()

    taken = [ipaddress.IPv4Network(subnet) for subnet in args.taken]
    main = ipaddress.IPv4Network(args.m)
    available = get_available_subnets_set(main, taken)

    make_subnet_graph(main, taken, available, args.output)