OSPFv3 Configuration
The process of configuring OSPFv3 involves the following steps:
Step 1. Initialize the routing process. As a prerequisite, ipv6 unicast-routing must be enabled on the router. Afterward, the OSPFv3 process is configured with the command router ospfv3 [process-id].
Step 2. Define the router ID. The command router-id router-id assigns a router ID to the OSPF process. The router ID is a 32-bit value that does not need to match an IPv4 address. It may be any number in IPv4 address format (for example, 0.1.2.3), as long as the value is unique within the OSPF domain.
OSPFv3 uses the same algorithm as OSPFv2 for dynamically locating the RID. If there are not any IPv4 interfaces available, the RID is set to 0.0.0.0 and does not allow adjacencies to form.
Step 3. (Optional) Initialize the address family. The address family is initialized within the routing process with the command address-family {ipv6 | ipv4} unicast. The appropriate address family is enabled automatically when OSPFv3 is enabled on an interface.
Step 4. Enable OSPFv3 on an interface. The interface command ospfv3 process-id ipv6 area area-id enables the protocol and assigns the interface to an area.
Figure 10-1 displays a simple four-router topology to demonstrate OSPFv3 configuration. Area 0 consists of R1, R2, and R3, and Area 34 contains R3 and R4. R3 is the ABR.
)
Figure 10-1 OSPFv3 Topology
Example 10-1 provides the OSPFv3 and IPv6 address configurations for R1, R2, R3, and R4. IPv6 link-local addressing has been configured so that all router interfaces reflect their local numbers (for example, R1’s interfaces are set to FE80::1) in addition to traditional IPv6 addressing. The link-local addressing is statically configured to assist with any diagnostic output in this chapter. The OSPFv3 configuration has been highlighted in this example.
Example 10-1 IPv6 Addressing and OSPFv3 Configuration
OSPFv3 Verification
The commands for viewing OSPFv3 settings and statuses are similar to those used in OSPFv2; they essentially replace ip ospf with ospfv3 ipv6. Supporting OSPFv3 requires verifying the OSPFv3 interfaces, neighborship, and the routing table.
For example, to view the neighbor adjacency for OSPFv2, the command show ip ospf neighbor is executed, and for OSPFv3, the command show ospfv3 ipv6 neighbor is used. Example 10-2 shows this command executed on R3.
Example 10-2 Identifying R3’s OSPFv3 Neighbors
Example 10-3 shows R1’s GigabitEthernet0/2 OSPFv3-enabled interface status with the command show ospfv3 interface [interface-id]. Notice that address semantics have been removed compared to OSPFv2. The interface maps to the interface ID value 3 rather than an IP address value, as in OSPFv2. In addition, some helpful topology information describes the link. The local router is the DR (192.168.1.1), and the adjacent neighbor router is the BDR (192.168.2.2).
Example 10-3 Viewing the OSPFv3 Interface Configuration
A brief version of the OSPFv3 interface settings can be viewed with the command show ospfv3 interface brief. The associated process ID, area, address family (IPv4 or IPv6), interface state, and neighbor count are provided in the output.
Example 10-4 demonstrates this command being executed on the ABR, R3. Notice that some interfaces reside in Area 0, and others reside in Area 34.
Example 10-4 Viewing a Brief Version of OSPFv3 Interfaces
The OSPFv3 IPv6 routing table is viewed with the command show ipv6 route ospf. Intra-area routes are indicated with O, and inter-area routes are indicated with OI.
Example 10-5 shows this command being executed on R1. The forwarding address for the routes is the link-local address of the neighboring router.
Example 10-5 Viewing the OSPFv3 Routes in the IPv6 Routing Table
Passive Interface
OSPFv3 supports the ability to mark an interface as passive. The command is placed under the OSPFv3 process or under the specific address family. Placing the command under the global process cascades the setting to both address families. An interface is marked as being passive with the command passive-interface interface-id or globally with passive-interface default, and then the interface is marked as active with the command no passive-interface interface-id.
Example 10-6 shows how to make the LAN interface on R1 explicitly passive and how to make all interfaces passive on R4 while marking the Gi0/3 interface as active.
Example 10-6 Configuring OSPFv3 Passive Interfaces
The active/passive state of an interface is verified by examining the OSPFv3 interface status using the command show ospfv3 interface [interface-id] and searching for the Passive keyword. In Example 10-7, R1 confirms that the Gi0/3 interface is passive.
Example 10-7 Viewing an OSPFv3 Interface State
Summarization
The ability to summarize IPv6 networks is as important as summarizing routes in IPv4 (and it may even be more important, due to hardware scale limitations). Example 10-8 shows the IPv6 routing table on R4 before summarization is applied on R3.
Example 10-8 R4’s IPv6 Routing Table Before Summarization
Summarizing the Area 0 router’s loopback interfaces (2001:db8:0::1/128, 2001:db8:0::2/128, and 2001:db8:0::3/128) removes three routes from the routing table.
Summarization of internal OSPFv3 routes follows the same rules as in OSPFv2 and must occur on ABRs. In our topology, R3 summarizes the three loopback addresses into the 2001:db8:0:0::/65 network. Summarization involves the command area area-id range prefix/prefix-length, which resides under the address family in the OSPFv3 process.
Example 10-9 shows R3’s configuration for summarizing these prefixes.
Example 10-9 IPv6 Summarization
Example 10-10 shows R4’s IPv6 routing table after configuring R3 to summarize the Area 0 loopback interfaces. The summary route is highlighted in this example.
Example 10-10 R4’s IPv6 Routing Table After Summarization
Network Type
OSPFv3 supports the same OSPF network types as OSPFv2. Example 10-11 shows that R2’s Gi0/3 interface is set as a broadcast OSPF network type and is confirmed as being in a DR state.
Example 10-11 Viewing the Dynamic Configured OSPFv3 Network Type
The OSPFv3 network type is changed with the interface parameter command ospfv3 network {point-to-point | broadcast}. Example 10-12 shows the interfaces associated with the 2001:DB8:0:23::/64 network being changed to point-to-point.
Example 10-12 Changing the OSPFv3 Network Type
After the changes are typed in, the new settings are verified in Example 10-13. The network is now a point-to-point link, and the interface state shows as P2P for confirmation.