qacafe - ip test solutions
Knowledge Base > What is the state flow of a typical IPv6 startup procedure?

What is the state flow of a typical IPv6 startup procedure?

IPv6

6to4 and 6rd WAN Mode Example Startup Procedure

With the addition of IPv6 functionality to CDRouter, additional sequences are performed during the startup procedures. Using the model of a DUT performing as a 6to4 capable router, both IPv4 and IPv6 stacks must be initialized during the startup procedure in order to accomplish testing.

The first task is for CDRouter to establish an IPv4 WAN connection with the DUT. This can be accomplished through many protocols, such as static, DHCP, PPPoE, et cetera.

After the IPv4 WAN connection has been established, CDRouter will attempt to allocate one or more LAN clients using DHCPv4 served by the DUT.

Once the IPv4 LAN configuration is complete, CDRouter establishes a relay server on the WAN side of the DUT. Neighbor Discovery is activated on the WAN interface. This interface often runs as an un-numbered IPv6 node.

Finally, the LAN side is configured with an IPv6 address. This is possible through autoconf or DHCPv6. In either case, the IPv6 address is deterministic and derived from the WAN IPv4 address (plus other values learned for a 6rd domain, if applicable). At this point, the IPv6 Remote Host is probed with Path MTU Discovery, to learn the appropriate MTU values to use during testing.

Static, DHCP, and Autoconf WAN Mode Example Startup Procedures

CDRouter IPv6 always assumes a dual-stack model. Even with a static IPv6 configuration on the WAN, CDRouter will still initialize the IPv4 portion of the WAN stack first. In these modes, the DUT often has a separate IPv6 address for both the WAN and the LAN interfaces. The WAN address is brought up and discovers the ISP IPv6 address using Neighbor Discovery.

After this has occurred, the LAN client is initialized. It will use either autoconf from Router Advertisements, or DHCPv6, to learn an IPv6 address.

Related articles that may also be helpful: