IPv6 Fundamentals

Why IPv6 Exists

IPv4's 32-bit address space provides about 4.3 billion addresses — long since exhausted for direct global allocation. IPv6 uses 128-bit addresses, providing approximately 340 undecillion addresses, enough that NAT's original purpose (conserving scarce public addresses) simply isn't needed.

Address Notation

An IPv6 address is written as eight groups of four hex digits, separated by colons:

2001:0db8:0000:0000:0000:ff00:0042:8329

Two abbreviation rules make this readable:

• Leading zeros in each group can be dropped: 2001:db8:0:0:0:ff00:42:8329
• One run of consecutive all-zero groups can be replaced with :: (only once per address, since using it twice would be ambiguous): 2001:db8::ff00:42:8329

IPv6 Address Types

Every IPv6-enabled interface gets a link-local address automatically, even with no other configuration — this is fundamentally different from IPv4, where an interface has no usable address until DHCP or manual configuration.

NDP and SLAAC: IPv6's Replacement for ARP and DHCP

• Neighbor Discovery Protocol (NDP) replaces ARP's job — resolving a neighbor's link-layer (MAC) address, plus router discovery and duplicate address detection
• SLAAC (StateLess Address AutoConfiguration) lets a host generate its own global address from a router advertisement's prefix plus its own interface identifier, without needing a DHCP server at all (though DHCPv6 still exists for stateful configuration when needed)

Dual Stack: The Real-World Transition Path

Almost no network flips a switch from IPv4 to IPv6 overnight. Dual stack — running both protocols simultaneously on the same interfaces — is the dominant real-world deployment model, letting traffic use whichever protocol the destination supports, while both stacks need to be independently secured (a common oversight: firewalling IPv4 thoroughly while leaving IPv6 wide open on the same hosts).

Why IPv6 Eliminates the Need for NAT (as an address-conservation tool)

NAT's rise in IPv4 was a direct response to address scarcity — one public IP, many private hosts behind it. With every device able to have its own globally routable IPv6 address, NAT is no longer required to conserve address space (though NAT66 exists for its other side-effect, obscuring internal topology, if an organization still wants that).

Common Pitfalls

• Deploying dual stack and firewalling only IPv4, leaving IPv6 traffic wide open on the same hosts
• Assuming NAT provides meaningful security — it never did; it was an address-conservation side effect, and a stateful firewall (needed regardless of IPv4/IPv6) is the actual control
• Forgetting that every interface has a link-local address by default, reachable on the local segment even if global addressing was never configured