基本OSPF
拿Area 0举例
首先配置PE3和PE4端口IPv6地址，然后配置OSPFv3进程，把进程和端口关联，并指定域即可，router-id仍然采用loopback IPv4地址。
[PE3]dis ospfv3 peer
OSPFv3 Process (1)
OSPFv3 Area (0.0.0.0)
Neighbor ID Pri State Dead Time Interface Instance ID
4.4.4.4 1 Full/DR 00:00:36 GE0/2/0 0
[PE3]display ospfv3 routing
Codes : E2 - Type 2 External, E1 - Type 1 External, IA - Inter-Area,
N - NSSA, U - Uninstalled, D - Denied by Import Policy
OSPFv3 Process (1)
Destination Metric
Next-hop
IA 2001:DB8:24::/64 2
via FE80::AE4E:91FF:FE48:6D1, GigabitEthernet0/2/0
（通过下一跳的本地IPv6地址获得，获得的是Area1的路由，所以是IA，域间的）
2001:DB8:34::/64 1
directly connected, GigabitEthernet0/2/0
虚连接
PE1不能获得Area0和Area1的路由，因为OSPF规定，每个域必须和骨干区域Area0相连，该例子中，PE1无法连到Area0，于是虚拟连接（V-link）就发明了。PE2和PE4搭鹊桥，让area0和area2相会。
PE2配置如下
ospfv3 1
router-id 2.2.2.2
area 0.0.0.1
vlink-peer 4.4.4.4
配置前PE1路由，只有直连路由
[PE1]dis ospfv3 routing
Codes : E2 - Type 2 External, E1 - Type 1 External, IA - Inter-Area,
N - NSSA, U - Uninstalled, D - Denied by Import Policy
OSPFv3 Process (1)
Destination Metric
Next-hop
2001:DB8:12::/64 1
directly connected, GigabitEthernet0/2/5
配置虚连接后，
[PE1]dis ospfv3 routing
Codes : E2 - Type 2 External, E1 - Type 1 External, IA - Inter-Area,
N - NSSA, U - Uninstalled, D - Denied by Import Policy
OSPFv3 Process (1)
Destination Metric
Next-hop
2001:DB8:12::/64 1
directly connected, GigabitEthernet0/2/5
IA 2001:DB8:24::/64 2
via FE80::AAEC:B8FF:FE4F:6903, GigabitEthernet0/2/5
IA 2001:DB8:24::1/128 1
via FE80::AAEC:B8FF:FE4F:6903, GigabitEthernet0/2/5
IA 2001:DB8:24::2/128 2
via FE80::AAEC:B8FF:FE4F:6903, GigabitEthernet0/2/5
IA 2001:DB8:34::/64 3
via FE80::AAEC:B8FF:FE4F:6903, GigabitEthernet0/2/5
Stub域
如果PE1体弱多病，这么多路由承受不起，所以可以考虑到把PE1放在Stub域保护起来。
PE配置如下：
ospfv3 1
router-id 1.1.1.1
area 0.0.0.2
stub
可以配置default-cost 10啥的
配置了stub区域后，就不会接收AS外的路由，如果想去AS外怎办？就生成一条默认路由
[PE1]dis ospfv3 routing
Codes : E2 - Type 2 External, E1 - Type 1 External, IA - Inter-Area,
N - NSSA, U - Uninstalled, D - Denied by Import Policy
OSPFv3 Process (1)
Destination Metric
Next-hop
IA ::/0 2
via FE80::AAEC:B8FF:FE4F:6903, GigabitEthernet0/2/5
2001:DB8:12::/64 1
directly connected, GigabitEthernet0/2/5
IA 2001:DB8:24::/64 2
via FE80::AAEC:B8FF:FE4F:6903, GigabitEthernet0/2/5
IA 2001:DB8:24::1/128 1
via FE80::AAEC:B8FF:FE4F:6903, GigabitEthernet0/2/5
IA 2001:DB8:24::2/128 2
via FE80::AAEC:B8FF:FE4F:6903, GigabitEthernet0/2/5
IA 2001:DB8:34::/64 3
via FE80::AAEC:B8FF:FE4F:6903, GigabitEthernet0/2/5
如果连域间路由都不想收的话，继续在PE2上配置
area 0.0.0.2
stub no-summary
PE1上看到只有默认路由和直连路由，域间路由消失。
[PE1]dis ospfv3 routing
Codes : E2 - Type 2 External, E1 - Type 1 External, IA - Inter-Area,
N - NSSA, U - Uninstalled, D - Denied by Import Policy
OSPFv3 Process (1)
Destination Metric
Next-hop
IA ::/0 2
via FE80::AAEC:B8FF:FE4F:6903, GigabitEthernet0/2/5
2001:DB8:12::/64 1
directly connected, GigabitEthernet0/2/5
DR选举
可以看到PE2是DR，为啥？PE2 Router-id大，注意下面DR是PE2的状态
[PE1]display ospfv3 peer
OSPFv3 Process (1)
OSPFv3 Area (0.0.0.2)
Neighbor ID Pri State Dead Time Interface Instance ID
2.2.2.2 1 Full/DR 00:00:31 GE0/2/5 0
在PE上配置DR优先级，让PE1成为DR，数值越低优先级越高。
interface GigabitEthernet0/2/5
ospfv3 dr-priority 100
重启进程reset ospfv3 1
OSPFv3 Process (1)
OSPFv3 Area (0.0.0.2)
Neighbor ID Pri State Dead Time Interface Instance ID
2.2.2.2 150 Full/Backup 00:00:38 GE0/2/5 0
GR
上面提到重启OSPFv3进程，这样会影响到转发，现网是有很大损失，于是发明了GR来做到重启OSPFv3进程，并不影响转发。
做个试验，现在PE4上ping PE1，PE1 reset OSPFv3进程，PE4上丢了6个包，说明重启进程改变了路由表，从而影响了转发。
--- 2001:db8:12::1 ping statistics ---
114 packet(s) transmitted
108 packet(s) received
5.26% packet loss
round-trip min/avg/max = 8/14/44 ms
配置PE1的GR功能，然后PE2配合helper-role。
reset ospfv3 1 graceful-restart
可以看到转发表依旧存在，不会因为路由表消失而消失。
IPv6 FIB Table:
Total number of Routes : 5
Destination: :: PrefixLength : 0
Nexthop : FE80::AAEC:B8FF:FE4F:6903 Flag : DGU
Interface : GigabitEthernet0/2/5 Tunnel ID : 0x0
TimeStamp : 2015-12-08 10:58
PE4上无丢包
--- 2001:db8:12::1 ping statistics ---
677 packet(s) transmitted
677 packet(s) received
0.00% packet loss
round-trip min/avg/max = 7/14/73 ms