DEV Community

Tanveer
Tanveer

Posted on

IPv4 Exhaustion: How the Internet Survives Despite 5.5 Billion Users

IPv4 addresses are exhausted, yet billions of users still connect daily. Let’s unravel the duct-tape solutions keeping the internet alive—and why IPv6 is the future.

🌐 The IPv4 Crisis: A Numbers Game

IPv4, the internet’s original addressing system, was designed in the 1980s with 4.3 billion unique addresses (like 192.168.1.1). Fast-forward to 2023: there are 5.5 billion active internet users and over 16 billion devices (phones, laptops, IoT).

The problem is obvious: There aren’t enough IPv4 addresses for everyone. Yet, the internet still works. How? Let’s break down the clever (and messy) workarounds.

🔧 Workarounds Keeping IPv4 Alive

1. NAT: The Magic of Shared Addresses

  • Network Address Translation (NAT) lets multiple devices share one public IP.
    • Example: Your home router uses one public IP but assigns private IPs (e.g., 192.168.1.2) to your phone, laptop, and smart fridge.
  • Carrier-Grade NAT (CGNAT): ISPs take this further, sharing a single IP across thousands of users. Mobile networks rely heavily on this.

2. Dynamic IP Allocation

  • Most ISPs assign IPs temporarily. When you disconnect, your IP goes back into a pool for others.

3. Private IP Ranges

  • Reserved ranges like 10.x.x.x and 192.168.x.x are reused globally. Your home network and a café’s Wi-Fi can both use 192.168.1.1 without conflict.

4. IPv4 Address Black Markets

  • With IPv4 exhausted, addresses are traded like real estate, costing $30–$50 each.

💥 What Happens if 5.5 Billion Users Hit IPv4 Simultaneously?

Even with these fixes, IPv4 struggles at scale:

1. Port Exhaustion

  • NAT uses ports (16-bit numbers, ~65k per IP) to track connections.
    • If 1,000 users share an IP via CGNAT, each gets ~65 ports. Result: Apps fail when ports run out.

2. CGNAT Overload

  • NAT tables (which map connections) get overwhelmed, causing timeouts and lag. Gamers and VoIP users suffer most.

3. IPv4’s Performance Tax

  • Multiple NAT layers add latency. Ever wondered why your Zoom call lags on hotel Wi-Fi? Thank CGNAT.

4. The IPv6 Lifeline

  • ~40% of users already connect via IPv6 (Google, 2023). Services like Netflix and Facebook use IPv6 to bypass IPv4 bottlenecks.

🚀 Why IPv6 Is the Only Real Fix

IPv6 uses 128-bit addresses (e.g., 2001:0db8:85a3::8a2e:0370:7334), offering enough unique IPs for every atom on Earth. Benefits include:

  • No more NAT: Every device gets a public IP.
  • Faster routing: Simplified headers and no NAT translation.
  • Built-in security: IPsec encryption is mandatory.

🔮 The Future: A Slow, Uneven Transition

  • Dual-stack networks: Most modern ISPs support IPv4 and IPv6 side-by-side.
  • Legacy systems: Older infrastructure still clings to IPv4, but IPv6 adoption grows ~10% yearly.
  • Developer action: Test your apps for IPv6 compatibility—tools like test-IPv6.com can help.

🛠️ What Can You Do?

  1. Advocate for IPv6: Push your company/ISP to prioritize it.
  2. Build IPv6-ready apps: Use libraries that support IPv6 (most modern frameworks do).
  3. Learn networking basics: Understand NAT, CGNAT, and subnetting.

TL;DR

IPv4 is held together by duct-tape solutions like NAT and CGNAT. While it barely works today, scaling further requires IPv6—a system with enough IPs for every grain of sand on Earth. The transition is slow but inevitable.

Top comments (1)

Collapse
 
zorton profile image
zorton

Your article is incorrect:

  1. IPSec is not mandatory. It was removed as a requirement in IPv6 and changed to a recommendation. There are several reasons for this.

  2. IPv6 most assuredly does not grow 10% yearly. There is absolutely no proof anywhere to support this claim.

Furthermore:

  1. IPSec was backported to IPv4.

  2. Every device getting a public IP is an idea that engineers like (or liked in the 1990s). That paradigm is no longer important. Content delivery, load-balancing and other modern needs are about data, not old phone-style point-to-point infrastructure. Plus: NAT is a feature, not a bug.

  3. NAT is not security but neither is a large address space. What is security is the built-in firewall consumers get with their routers. NAT helps make that process easy. The router is a hardware firewall that protects the entire household or site.

  4. VPNs represent massive business, expected to grow to nearly $500 billion in ten years, with about 1/3 of all Internet users using them. The point? Thanks to current speeds, users are fine with the performance hit. VPNs turn NAT into a profit center and users get security and privacy.

  5. IPv6 has one killer app - the address space. However, the addresses are horrible and IPv4 has CG NAT. Ways to expand IPv4 are likely to continue.

  6. Average users don't need lots of ports. Studies show typical households use ~20–50 concurrent ports max. For games, a few static ports will do and that is easily possible even with CG NAT. I've used CG NAT and it was fine.

  7. Everything that home users and SMB owners/engineers need for their LAN is solved by IPv4.

  8. IPv6's SLAAC created significant privacy risks, leading to Privacy Extensions (RFC 4941/8981). While mostly effective, these introduce challenges with network congestion, configuration complexity, reduced adoption in enterprises due to management difficulties, security trade-offs in tracking and filtering, and connection disruptions for applications requiring stable addresses.

  9. IPv6 is not dominant after 30 years. Cloudflare reported it at 39%. Much of the Internet is IPv4-only. I can buy static IPv4 IPs for $5-10 (with VPS hosts adding them on for free). If people want to fix the address "crisis", they should seriously consider ezIP or another solution compatible with IPv4.

The IPv4 part of the Internet is massive and highly functional. Your article unfairly and incorrectly paints it as something barely alive and on the verge of collapse. Reality is that IPv6 failed to overtake IPv4 and has not evolved. New and better protocols are being tested and are the real future. It's time to stop living in 1995 (or 1998 if you prefer) and work on a new solution.