The following excerpt is from Datacate’s Colocation Survival Guide – get the full Guide here.

Every Internet presence project will necessarily involve the use of static public IP address space, except in the rare instances where no public IP connectivity is required. Whether you receive your Internet connectivity – and therefor, your IP address space – from your colocation provider or from a dedicated carrier, you will be asked how many IP addresses you need, and if the quantity that you request is substantial, you may be required to submit justification information.

Both carriers and colocation providers will issue address space allocations as dedicated subnets of a fixed size, which most closely matches the quantity of addresses you requested or were approved for. An IP ‘subnet‘ (short for ‘sub network‘), is just a group of contiguous IP addresses of a predetermined size that has been carved out of a larger group of IPs, in a way that allows it to be treated separately from those that come before and after it. There is a more concise technical explanation to be sure, but it’s really not necessary to go into that. The important point is that, rather than giving you a random scattering of IPs from widely varying ranges, a subnet is a uniform chunk of addresses, all from the same range, assigned solely to you. The size of the subnet can be varied, from as little as a single IP address, on up to thousands or millions… but in all likelihood, your provider will assign you a subnet of perhaps eight, 16, or 32 IP addresses, depending upon how many you can justify.

A more in-depth discussion of the technical aspects of IP address subnets can be the topic of another article, but for now there are two considerations you need to be aware of:

1. IP address subnets are generally not expandable in size once issued, so if you need more addresses later, your only option will be to take an additional subnet from a different range
2. The global supply of IPv4 addresses (the addressing scheme that most people are familiar with) is nearly exhausted, so addresses are in very short supply. Because of this, providers are likely to be rather stingy with IPv4 allocations, and will probably charge you a monthly usage fee for IP addresses. On the other hand, carriers and providers will usually promote the adoption of the newer IPv6 standard, encouraging you to take an IPv6 allocation, typically at no charge.

If you reach a point where you need more IP addresses, the first thing you’ll want to determine is: do you REALLY need more IP addresses? In most cases, multiple applications and services can be run on a single IP address. There are only a few instances in which a separate distinct IP address is truly needed in order to add a service or a function, so do your research if you are unsure.

As mentioned earlier, IP subnets usually cannot be expanded in size once they have been allocated – it never hurts to ask the provider if it can be done, but expect a “no”. Instead, assuming that your provider approves your request for additional IPs (it is possible that you will be denied), you would be issued a second IP subnet of the appropriate size. While there is nothing wrong with this, you need to be aware of it as it introduces some additional technical considerations, which are beyond the scope of this guide to discuss. Furthermore, in the current climate of IPv4 global exhaustion, you can expect to be charged a monthly fee for additional IPv4 allocations.