The Exhaustion of IPv4

With only $2^{32} \approx 4.3$ billion addresses, IPv4 has officially reached exhaustion in most regional registries (RIRs). IPv6, providing $2^{128}$ addresses, is the solution, but the migration is complex because the protocols are not naturally backward compatible.

Primary Transition Strategies

1. Dual Stack (The Gold Standard)

Nodes run both IPv4 and IPv6 protocol stacks simultaneously. Applications choose the protocol based on DNS resolution (preferring IPv6 via 'Happy Eyeballs' algorithms).

2. Tunneling (Encapsulation)

Encapsulating IPv6 packets inside IPv4 headers (or vice-versa) to traverse non-compatible infrastructure. Common types include:

• 6in4: IPv6 packets in IPv4 (fixed endpoints).
• 6rd: RAD-based rapid deployment.
• 464XLAT: Double translation for mobile networks.

3. Translation (NAT64/DNS64)

Used when an IPv6-only client needs to talk to an IPv4-only server. A NAT64 gateway maps the IPv6 address to a temporary IPv4 address.

MTU Challenges in Transition

Tunneling introduces header overhead, which reduces the effective Payload MTU. If the path doesn't support larger frames, fragmentation occurs.

For example, standard IPv4 encapsulation adds 20 bytes, reducing a 1500B MTU to 1480B.

Conclusion

While Dual Stack is ideal, the scarcity of remaining IPv4 addresses makes it increasingly expensive. Mechanisms like NAT64 and 464XLAT are becoming the default for modern mobile and ISP infrastructure.