Unicast-Prefix-based IPv6 Multicast Addresses
updated by
Extracted elements (17)
Multicast address lifetime is tied to the unicast prefix's Valid Lifetime in Router Advertisement messages to support the Abstract API for Multicast Address Allocation (RFC 2771), though the RA lifetime value typically remains constant until a renumbering event.
The network prefix portion of the multicast address is limited to at most 64 bits because Interface Identifier requirements in Section 2.5.1 of the IPv6 addressing architecture effectively restrict unicast prefix length to 64 bits.
The proposal reduces the number of protocols needed for dynamic multicast address allocation by eliminating the need for the Multicast Address Allocation Protocol (AAP) and the Multicast Address-Set Claim (MASC) Protocol, replacing them with unicast prefix delegation.
Using unicast prefix-based multicast addresses after the unicast prefix has become invalid may cause operational problems, such as routers performing policy checks comparing the multicast prefix against the unicast prefix assigned to an AS discarding packets.
When P=0, the address format reverts to the standard IPv6 multicast address as defined in the addressing architecture (RFC 2373/ADDRARCH), preserving backward compatibility with existing multicast address usage.
All non-significant bits of the network prefix field SHOULD be zero when P=1.
If P=1, T MUST be set to 1. This ensures that unicast-prefix-based multicast addresses are always marked as dynamically assigned.
The lifetime of a unicast prefix-based multicast address SHOULD NOT exceed the Valid Lifetime field in the Prefix Information option of the Neighbor Discovery Router Advertisement message corresponding to the embedded unicast prefix.
The reserved field MUST be zero.
The scope of the unicast-prefix-based multicast address MUST NOT exceed the scope of the unicast prefix embedded in the multicast address.
To form a Source-Specific Multicast address, a node MUST set P=1, plen=0, and network prefix=0, yielding the SSM range FF3x::/32.
Source-Specific Multicast (SSM) addresses are formed by setting P=1, plen=0, and network prefix=0, creating the address range FF3x::/32 where 'x' is any valid scope value. The IPv6 source address identifies the owner.
The flags field (flgs) is 4 bits with the two high bits reserved (0), P bit indicating unicast-prefix-based allocation, and T bit indicating transient/dynamic address. P=1 signals the address is assigned based on the network prefix.
The plen field (8 bits) indicates the actual number of bits in the network prefix field that identify the subnet when P=1. The network prefix field (64 bits) contains the unicast prefix assigned to the domain owning the multicast address.
The embedded unicast prefix can aid in identifying the allocation domain of a multicast address. However, an allocation domain can avoid traceability by using a prefix not assigned to it or using a smaller scope embedded prefix — no enforcement mechanism is provided.
Using source-specific multicast addresses can sometimes aid in preventing denial-of-service attacks by arbitrary sources, though no guarantee is provided. Full SSM security considerations are in the SSM architecture document.
The unicast-prefix-based IPv6 multicast address format is: 8 bits (11111111) | 4 bits flags | 4 bits scope | 8 bits reserved | 8 bits plen | 64 bits network prefix | 32 bits group ID. This replaces the original 112-bit group ID field with structured prefix information.