What are Floating IPs and are they useful?

Floating IPs represent a hybrid form that combines dynamic and static IPs, allowing for flexible usage. They are used to enable a specific instance or network to be externally reachable without creating a permanent association between the address and a specific server, gateway, or computer. Floating IPs are publicly visible, can be permanently routed, and are not automatically assigned to a client like dynamic or static IPs. Instead, Floating IPs can "float" or switch between different end devices as needed, for example, to ensure continuous external accessibility in case of hardware failure or overload.

Addressing and DNS Resolution - How do digital networks work?

To uniquely identify a computer or network node, the Internet Protocol (IP) assigns each participant a distinct and unique address used for the delivery and allocation of each request. It consists of a numerical code, which is encoded in the Internet Protocol Version 4 (IPv4) in the decimal system and in Version 6 (IPv6) as a hexadecimal number combination. The transition between the two versions became necessary because IPv4, with its structure of four elements or 32 bits, each with 256 possible values, technically allowed for a maximum of 4,294,967,296 addresses. This seemed more than sufficient in the early days of the internet - however, with the increasing connectivity through the Internet of Things (IoT), it became evident that significantly more addresses were needed. Purely mathematically, it is not even possible to assign a unique address to every other person.

For this reason, a switch to IPv6 took place from 1998 onwards, which theoretically allows for more than 340 sextillion combinations - a number with 38 zeros - through encoding with 128 bits. This also means that there are enough spaces available to assign Floating IPs to all users. In order to transform the addresses, which are cryptic and difficult to understand for humans, into a simple and easy-to-use format, Domain Name Servers (DNS) exist. They break down an internet address like Example.com into its Subdomain Example and the Top Level Domain (TLD) .com, informing a client which unique IPv4 or IPv6 is responsible for processing a request. If you are wondering what a nameserver is and what exactly a DNS entry does, please refer to this article.

What is the difference between dynamic, static, and Floating IPs?

Normally, IPv4 and IPv6 addresses are assigned to a fixed instance (static IPs) or automatically allocated to a pool of clients (dynamic IPs). Static IPs are permanently and continuously assigned to a destination and do not change, for example, if the destination restarts or is out of operation for a certain period. In contrast, dynamic IPs are automatically assigned to a client by a higher authority through protocols like the Dynamic Host Configuration Protocol. The allocation occurs when a client first logs into the network and then receives an IP from a fixed pool of addresses. Upon logging out or after exceeding a specified time period, the IP may remain reserved for a certain period or be immediately released as available.

Floating IPs cannot be clearly classified into one of these categories because they are permanently owned by a customer, but do not point to a specific piece of hardware - such as a gateway or a web server. Like static IPs, they remain identical and are uniquely and permanently assigned to a lower instance. However, Floating IPs can point to different end devices like dynamic IPs and switch between them without any delay. For example, Floating IPs can simultaneously point to an active and a passive dedicated server - if the former fails due to hardware failure or maintenance, the Floating IPs immediately redirect the traffic to the latter.

Which category of devices to use depends not least on their task and the type of hardware. Internet Service Providers (ISPs) typically assign dynamic IPs to their customers, so that the address changes from a router after a restart and at fixed intervals - usually every 24 hours. This means that local networks (LANs) behind a router, such as a home network, cannot be reached directly from the internet permanently. For example, to operate a web server with a domain in your own home, an external service like No-IP.com or DynDNS is required, which assigns the current dynamic IP after each domain change. On the other hand, instances like servers or gateways that need to be permanently reachable under a fixed address use static IPs. Floating IPs, on the other hand, are used in branched networks that require maximum security, performance, or accessibility. They are used to enable redundant structures without downtime, distribute loads across different instances, or address multiple parallel systems through a single address.

What are the advantages and disadvantages of Floating IPs?

Floating IPs can be assigned within a subnet according to individual criteria. This results in both advantages and disadvantages - the network can be publicly accessed and sent packets are subject to permanent routing, but they can be accepted and responded to by different end devices. This concept results in various benefits of Floating IPs that are important for a flexible and high-performance IT infrastructure:

• Redundant systems in case of failure of individual instances
• Permanent assignment of a single IPv4 or IPv6 to a user
• Flexible application of Floating IPs for internal or external networks
• Maximum accessibility even in extreme situations
• Distribution of computing or network load according to individual criteria
• Permanent routing of Floating IPs to multiple instances
• Transfer of Floating IPs between instances without any downtime
• Floating IPs can be combined with dynamic or static IPs

Floating IPs also have certain disadvantages, although these are usually not severe. The most important one is certainly that Floating IPs incur additional costs. They typically consist of a one-time fee for the setup and a monthly fee ranging from single to double-digit euros. Furthermore, they only make sense if the Floating IPs actually point to multiple instances - such as two or more dedicated servers or geographically separated virtual servers. It is also worth noting that not all companies offer Floating IPs - even some large corporations do not provide this service. Lastly, Floating IPs can only fulfill their purpose if they are correctly implemented into a system.

Photo: Michael Schwarzenberger on Pixabay