Conventional NPTv6 using ULA
Unique Local Addresses provide the standards-defined method for stable, non-global IPv6 addressing inside an organization.
2.1Architecture
In the conventional model, the organization generates and uses a ULA prefix internally. NPTv6 maps the relevant internal portion to the currently usable ISP-provided GUA prefix at the network edge.
2.2Advantages
Standards-defined local space
ULA exists specifically to provide stable addressing that is not dependent upon global Internet routing.
Administrative independence
The internal plan does not depend on continued control of an externally registered prefix.
Structured IPv6 subnetting
A locally generated /48 supports conventional /64 LANs and a substantial internal hierarchy.
Clean NPTv6 mapping
Corresponding internal and external subnets can be translated without maintaining per-flow state.
2.3Address-selection implications
ULA is legitimate IPv6 address space, but it is intentionally distinguished from globally scoped addressing. Default source- and destination-address selection policy can therefore produce behaviour that surprises operators whose goal is an IPv6-first dual-stack LAN.
Operational consideration
When a hostname has both a public IPv6 address and an IPv4 address, a host whose candidate IPv6 source is ULA may prefer the IPv4 path under default policy. The precise outcome depends on the host, address candidates, and local policy.
This does not make ULA defective or unusable. It means that a deployment seeking specific source-selection behaviour may need to validate host policy and, where appropriate, adjust it consistently across managed endpoints.
2.4When this model fits
ULA is the natural choice when standards-defined local addressing, administrative independence, and clear separation from global space are more important than retaining ordinary GUA semantics on every internal interface.