7.9 IP Planning

Organizing Addresses Before You Start

IP planning is the process of deciding in advance how you will use IP address space in a network. Instead of handing out addresses randomly, you design a clear structure so every device, subnet, and site has a predictable place.

Good IP planning makes networks easier to grow, troubleshoot, secure, and document. Poor planning leads to wasted addresses, confusion, and painful redesigns later.

In this chapter, you will see how to think about IPv4 address space at a high level and how to turn requirements into a logical addressing plan. The detailed mechanics of subnetting, CIDR, VLSM, and supernetting are covered in their own chapters, so here we focus on how to use those tools as part of a plan, not on how they work internally.

Starting From Requirements

IP planning always starts from requirements, not from a random IP block.

You usually gather information such as:

1. How many sites or locations exist. For example, head office, two branch offices, a data center, and a cloud VPC.
2. How many subnets each site needs. For example, a user VLAN, a server VLAN, a guest Wi Fi network, a management network, and so on.
3. The size of each subnet. For example, up to 50 users in one office, up to 500 users in another, or only a few devices for network management.
4. Special networks. VPN address pools, DMZ networks, point to point links, lab environments, and test ranges.
5. Growth expectations. Will a small office grow into a large office. Will you add more Wi Fi networks or more servers.

You turn these requirements into a list such as:

This requirement table is the starting point for your IP design. You then map these needs onto actual address blocks with enough space for growth.

Choosing Address Space

In most private networks you use private IPv4 address ranges. The choice of which private block to use is part of IP planning.

The commonly used private ranges are:

You can use more than one private block, but you must avoid overlapping networks inside your own environment. If you connect with partners, MPLS providers, or cloud, it is best to coordinate and avoid using the same ranges they use, to reduce translation work later.

Typical planning approaches:

1. Small environment, such as a single office or small company. Often uses 192.168.0.0/16, divided into multiple /24 subnets.
2. Medium environment with multiple sites. Often uses 172.16.0.0/12. Different sites receive different /16 or /20 blocks.
3. Large environment or service provider. Often uses 10.0.0.0/8 and reserves large chunks for different regions or divisions.

For public IPs, the planning is more constrained and usually managed by the ISP or by network teams that deal with Internet routing. Here, careful conservation and security are important, but the principles of hierarchy and documentation stay the same.

Hierarchical Addressing

Hierarchical addressing means you structure your address space in layers. You reserve large blocks for big areas, and inside those you carve smaller subnets for specific purposes.

This hierarchy usually reflects:

1. Geography, such as countries, regions, sites.
2. Function, such as user networks, server networks, management networks.
3. Security zones, such as inside, DMZ, outside.

You want addresses to be meaningful. By looking at an IP address, you should have a rough idea of where and what it is.

For example, you could plan:

Inside each site block, you then group subnets by function in an orderly way.

This hierarchical design enables route summarization, reduces routing table size, and makes troubleshooting easier because IPs are predictable.

Pairing Sites With Address Blocks

A very common pattern in IP planning is to assign each site a fixed size block, such as a /24 or /22 or /20, depending on size and future growth.

You then use the individual subnets inside that block for specific VLANs or segments. The exact prefix sizes come from your knowledge of CIDR and subnetting, which are treated in the related chapters. Here the focus is the idea of reserving space per site.

Example pattern for a medium sized environment using 172.16.0.0/12:

Inside HQ 172.16.0.0/20, you might later define:

You do not need to use all space on day one. Leaving room for growth is essential.

Functional Grouping of Subnets

Inside each site, it is common to group subnets by their function and assign ranges in a regular pattern. This makes it much easier to navigate the network.

For example, you might decide:

1. User VLANs use the lower part of the site block.
2. Server VLANs use the middle part.
3. Management and infrastructure networks use the upper part.
4. DMZ and externally facing networks use a clearly separated section.

Consider a site block 10.1.0.0/20. You can define a simple pattern:

Inside each region, you carve subnets of appropriate size. The important point is consistency. Once you define that user networks live in the first part of a site block, keep using that pattern for all sites. After a while, you can guess where a subnet belongs just by its address.

Planning for Different Sizes With VLSM

Different networks have different size needs. A voice network with IP phones at a small branch may only need 30 addresses. A user network in HQ may require 600 addresses. Giving every subnet the same size wastes a lot of addresses.

Variable Length Subnet Masking, treated in its own section, lets you create subnets of different sizes from a single larger block. In IP planning, you use VLSM as the technique that lets you match subnet size to real usage.

From a planning perspective, the key ideas are:

1. Assign smaller subnets to networks that will remain small. For example, /27 or /28 for small management or point to point networks.
2. Assign larger subnets to user networks that need many addresses, such as /23 or /22 for large LANs.
3. Keep the subnets inside each site grouped to preserve hierarchy, even when they have different sizes.

A planning workflow might be:

1. Take the site block, for example 192.168.10.0/24 for a small branch.
2. Decide the largest subnet first, such as the user LAN.
3. Place the next largest, such as guest Wi Fi.
4. Use remaining space for small networks like management or printers.

Example for 192.168.10.0/24:

Details of how these subnet sizes are calculated come from subnetting knowledge, but the planning idea is clear: different needs, different subnet sizes, all fitting inside a planned block.

Summarization Friendly Design

Route summarization, discussed elsewhere, works best when your addressing has clean boundaries. IP planning must respect these boundaries if you want simple summarization.

Key planning hints:

1. Use contiguous blocks per site or region. Do not scatter subnets for a site all over your entire IPv4 space.
2. Use prefixes that can be summarized into a single route. For example, if all HQ networks fall under 10.1.0.0/16, you can advertise one summary route instead of many smaller ones.
3. Align summaries on natural binary boundaries. This means designing blocks that match clean prefix sizes such as /16, /20, /24, instead of odd shaped collections.

An example of summarization aware design:

Even if you do not use every subnet in those /16 blocks, the ability to summarize simplifies routing and often improves performance.

Address Roles and Conventions

IP planning includes decisions about how particular addresses inside each subnet are used. These are conventions, not strict rules of the protocol, but they help maintain consistency.

Typical conventions include:

1. Default gateway placement. Often the first usable IP of a subnet, such as x.x.x.1, or the last usable, such as x.x.x.254. Pick one convention and keep it.
2. Static vs dynamic ranges. IPs reserved for servers, printers, network devices and IPs handed out by DHCP.
3. Infrastructure addresses. A special pattern for network devices, such as always using low numbers in each subnet for routers, switches, and firewalls.
4. Reserved addresses. Specific parts of each network held aside for future changes or special functions.

An example layout for a /24 subnet 192.168.5.0/24:

These numbers are only an example. Your plan might differ, but the key is to define a convention before deployment and document it.

Planning for Special Networks

Beyond regular LAN segments, certain types of networks require specific planning attention.

Common special networks include:

1. WAN and point to point links. These networks usually need only two usable addresses. You can assign very small subnets and group them in a dedicated address block, so all transport links are easy to identify.
2. DMZ networks. These host publicly reachable services. It is helpful to locate them in a recognizable range, separate from internal user and server networks.
3. Management networks. For example, out of band management or dedicated management VLANs. These should be isolated and follow a clear pattern for all devices.
4. VPN address pools. Remote users and site to site VPNs often use their own pools, which must not conflict with internal networks and partner networks.
5. Lab and test ranges. Commonly placed in a predictable but separate portion of the address space to allow experiments without impacting production.

A simple way to treat these is to reserve a dedicated portion of your global private space:

Inside each block, you can further subdivide by site and use VLSM to assign exact sizes.

Growth, Renumbering, and Future Proofing

IP planning must think years ahead. The worst case is designing a tight address plan that looks efficient at first but leaves no room for growth.

Important principles:

1. Overprovision address space, within reason. If a site has 100 users today but may have 500 in three years, assign a block that can grow to that size.
2. Reserve extra space for new services. For example, new Wi Fi SSIDs, new application environments, or new security zones.
3. Avoid hardcoding addresses everywhere. Use DHCP and DNS to keep renumbering easier if you must change a subnet later.
4. Separate addressing for internal and external use. For example, do not expose internal IPs directly when you can use NAT or load balancers to front services.
5. Consider mergers, acquisitions, and partner connections. Overlapping private ranges are very common in such situations. A clear, documented plan helps resolve overlaps or avoid them beforehand.

Renumbering an active network is slow, risky, and tends to break things temporarily. It is far better to leave extra address space upfront than to run out and be forced to renumber.

Documentation and IP Address Management

A good IP plan is useless if it lives only in someone’s head. Once you decide on blocks, subnets, and conventions, you must document them clearly and keep them updated.

At a minimum, documentation should capture:

1. Global blocks and their purpose. For example, 10.0.0.0/8 for internal, 192.168.0.0/16 for labs.
2. Site allocations. For example, which block each site owns and its summary.
3. Individual subnets. Network address, prefix length, default gateway, VLAN ID, purpose, and size.
4. Address roles. Which ranges are static, which are DHCP pools, and which are reserved.
5. Special networks. WAN links, DMZ, management, VPN pools.
6. Contact and ownership. Who is responsible for a site or block.

You can start with a simple spreadsheet that includes columns such as:

As networks grow, organizations often adopt dedicated IP Address Management tools, sometimes called IPAM, to track all this information, integrate with DHCP and DNS, and avoid conflicts.

Putting It All Together

IP planning combines several ideas:

You start from requirements about sites, subnet counts, and sizes. You choose appropriate private address blocks and partition them hierarchically by region, site, and function. You use VLSM to match subnet sizes to actual needs, but always leave room for growth. You group related networks into neat blocks to allow summarization. You define clear conventions for gateways, static ranges, and DHCP pools. You reserve dedicated ranges for special networks such as WAN, DMZ, management, and VPN. Finally, you record everything in a living document or IPAM system.

With a solid IP plan in place, all the technical tools from subnetting, CIDR, VLSM, and routing become easier to apply, and the entire network is more predictable, scalable, and maintainable.