.png)
How the IPv4 shortage affects cloud providers and ISPs
- LARUS Foundation
- Aug 6, 2025
- 8 min read
Updated: Aug 15, 2025
Table of Contents
Introduction: The IPv4 shortage crisis
The global shortage of IPv4 addresses is not a future issue. It is happening now. IPv4 gives about 4.3 billion unique addresses. This number is not enough for how large the Internet is today. IANA used up its last address blocks in 2011. Regional registries such as APNIC, ARIN, and RIPE NCC also ran out in the years that followed (en.wikipedia.org).
There are no more unused IPv4 addresses. Cloud providers and ISPs have to use old or transferred address blocks. This makes their costs go up. It also makes their networks harder to manage. It becomes more difficult to grow their systems. Many networks still do not fully utilize it. Although more addresses are indeed provided by IPv6, most vendors still choose to run both IPv4 and IPv6 simultaneously.. They also need tools like NAT to keep their services working.
IPv4 depletion timeline: Milestones and consequences
The shortage of IPv4 addresses started step by step. In 2011, IANA gave out its last group of IPv4 addresses. This action caused each regional Internet registry to begin running out of their own supply.These registries then set clear rules. New users could only receive a small block, like a /22, and this was allowed only for networks that were moving to IPv6.
In 2025, this situation changed. No large blocks are left to give. IPv4 addresses are now bought, rented, or reused by those who still need them. There is no way to make more of them. The number stays the same. People and companies now decide how to share or trade these addresses. The supply depends on the market, not on technical growth or new needs.
How ISPs manage IPv4 scarcity
Internet service providers use different ways to deal with the shortage of IPv4 addresses. One common way is Carrier-Grade NAT (CGNAT). In this setup, providers give users private IP addresses. These addresses do not work on the public Internet. When users go online, the network changes private addresses into public ones. This change happens inside the ISP’s system on a large scale.
CGNAT lets many users share the same group of public addresses. It helps providers continue service with limited resources. This also makes the network harder to manage. It adds extra work to system teams. Some online tools do not work well, such as video calls or file-sharing software. These tools often need direct connections. Some websites or games may not run smoothly. The use of CGNAT also makes it harder to track network activity or match users to specific traffic. This is because many users may appear to share one public IP address.
Providers also try to solve the problem by joining the IPv4 address market. In this market, they can get addresses by buying or renting them from others. Many providers now choose this way. The cost keeps going up.
A report from RIPE NCC in 2024 showed that cloud companies and ISPs often say they cannot get new IPv4 addresses from public pools. They now turn to leasing or trading. They say they feel pressure from limited supply. Most no longer wait for free allocations. They now must use what the market offers, based on need and cost.
Cloud providers: IPv4 exhaustion and deployment challenges
Cloud service providers that give users public IP addresses or flexible IP plans now deal with serious problems from the lack of IPv4 supply. These problems change how services work, how they are priced, and how users access them. In many cases, they also affect how well customers can use the cloud.
IP scarcity limits service offerings: When a customer starts a new virtual machine or container, it often needs a public IPv4 address. In the past, this task was simple. Now, the number of free IPv4 addresses is too small. Some providers limit how many public addresses a customer can get. Some charge more for each one. This makes it harder to grow systems. It also makes users change how they plan their network use.
Cost inflation: Providers that do not have enough IPv4 addresses must go to the open market. They buy or rent addresses from brokers or other companies. These prices keep rising. This makes the cost to run the network go up. These higher costs are not only on the provider. Many providers raise fees or add charges.
Transition complexity: Nowadays, there are quite a few providers that support both IPv6 and IPv4. The reason for this is that some users' old systems still rely on IPv4.This setup is called dual-stack. To make this work, providers build tools that change IPv6 to IPv4. These tools include NAT64 and other systems that help traffic move between the two. In virtual networks, these tools make traffic slower. They also use more computer power. They make setup and management harder. This brings more delay and may hurt the performance of the services.
IPv6 transition: Adoption status and barriers
More networks use IPv6 now, and the number of users with IPv6 access gets higher. But most systems still depend on IPv4. Many services run on IPv4 because the older systems were built that way. These systems are still in place, so they cannot change fast. Many providers use both IPv4 and IPv6. They do this to support users who need either one. This setup is called dual-stack, and it helps networks stay connected for all types of users.
Many users still work on IPv4. Most tools and services are also made for IPv4. When a network wants to support both, it needs more tools. These tools let IPv6 users reach IPv4 systems. These tools include translators and gateways. They help with the traffic, but they also make the setup more complex. This adds time to the connection and may cause slower speeds or other issues.
IPv4 addresses are now spread out in small blocks. These blocks were moved or sold, and they are not in the same place anymore. Routers must hold more records to track each small block. This uses more memory and slows down updates. When routers get too full, the network may become harder to manage. Engineers must spend more time keeping the system working the right way. These tasks get harder as more blocks move to more places.
Cost dynamics: IPv4 markets and leasing
IPv4 addresses are running out, so they are now treated like a limited item. They are not easy to get from registries. Many companies that need them go to private markets. They rent or buy them through brokers,Some brokers are IPXO and Brander Group. These brokers connect companies that own extra addresses with those that need more. Demand stays high, but supply does not grow.
A10 Networks has a tool that shows possible future prices. It says that one address may cost $83 in 2026 if the market stays stable. If many companies try to buy quickly, the cost may rise to $112 or higher. This puts more pressure on providers. They pay more for each address. This can change how they set prices for customers.
To lower the need for new addresses, some providers try ways to share what they already have. One way is address plus port, where many users get the same address but with different ports. Another way is CGNAT, which lets hundreds of users use the same public IP. These steps help them use fewer addresses. But they cause new problems. Tracking each user becomes harder. Some services run more slowly or have more errors. These limits make network planning more difficult.
Technical impacts: performance, complexity, and scalability
NAT layers and network complexity
When networks use more NAT layers, the system becomes harder to manage. These layers may begin at the user’s home router and also appear deeper in the provider’s network. Each layer changes the address of a device, and this adds more places where problems can happen. It also makes port-forwarding harder, because the real address is hidden. Some tools need a direct link between two devices to work well. When NAT is in the path, the link breaks. This causes problems for games, video calls, or file sharing, because they all depend on full connection between both ends.
Routing table bloat
When IPv4 blocks are split and moved, the result is many small address groups. These groups are sent to different networks. Routers must keep a full list of where each group goes. When there are too many small groups, the router’s memory fills up. It also takes longer to update the list when networks change. When the list gets too long, the system runs slower. This slows down how traffic moves, and this also makes it harder to recover when something goes wrong.
Virtualisation latency and overhead
In cloud systems, many servers run virtual machines. These machines often use IPv6, and the network needs to move their traffic across IPv4. To make this work, the system wraps the data by using 6to4, Teredo, or ISATAP. This sends the traffic through IPv4, and it lets older systems keep working. This method adds more steps. The data must be wrapped, passed through the network, and then unwrapped. These steps take time and also reduce the speed. Some tasks need a fast response, like live video or online trading. If there is too much delay or jitter, the traffic cannot keep up, and the task does not run well.
Responses and perspectives: how ISPs and cloud providers are adapting
Cloud providers and ISPs now take different steps to deal with the shortage of IPv4 addresses. Many run both IPv4 and IPv6. This setup is called dual-stack. It supports older systems that still use IPv4 and also helps the network move toward IPv6. Some networks use IPv6 as the main layer. These networks include tools that let IPv6 traffic reach IPv4-only systems. This makes sure that older services can still work.
Some providers do not buy new address blocks. They lease them from other networks or brokers. This gives them more choice. They can get addresses when needed and return them later. This avoids large one-time payments. IPv4 prices are going up, so short-term leases may cost less than buying full blocks. Many providers use this method to control how much they spend.
To reduce how many addresses they use, some networks group users. One method is shared address space. Another is address plus port, or A + P. This lets many users share the same address by giving each one a different port. Some also use CGNAT. This allows even more users to connect through the same IP. These methods help save addresses. They also cause problems. Networks may become slower. It may also be harder to find and fix user issues when trouble happens.
Some networks change how they are built. They add tools that allow both IPv4 and IPv6. These include NAT64 and DNS64, which help traffic move between both versions. Some networks also use tunnels. These put IPv6 data inside IPv4 packets. These steps help with coverage, and they also add delay and extra load. Networks must check how much these tools affect speed and service.
Many people who work in this field have said that the lack of IPv4 is now a serious issue. Alexander Timokhin from InterLIR said the shortage changes cost, growth, and network results. In 2024, RIPE NCC shared a report. It said that most providers now feel pressure to find more IPv4 addresses. They often rent or trade them. Experts also said that splitting address blocks makes routing more complex. The routing table grows larger. This gives networks more tasks and slows down how they respond to change.
FAQs
1. Why are IPv4 addresses so scarce today?
All the original pools from IANA and regional registries have been used, and the rest must come from old blocks that are reused or moved.
2. How do cloud providers acquire IPv4 addresses now?
They rent or buy addresses from other holders, and they also work with brokers who help transfer them between networks.
3. How do networks handle the lack of public IPv4 addresses?
They give users private addresses, and then turn them into public ones at the main gateway because the available public range is too small.
4. How do ISPs share limited IPv4 addresses among users?
They give each user a private address, and then change it into a public one at the network level because public addresses are not enough.
5. How does IPv4 scarcity affect network performance?
More NAT layers slow things down, and routing tables grow larger, and tunnelling over IPv6 can also raise delay and lower quality.
Comments