Hi,

I have working setup of both ospf/ospf6. Nothing fancy.

Cisco <---> [OBSD firewalls] <---> Cisco

The inet6 interfaces on OBSD are setup with -autoconf

All routers prefers to set next-hop routes via the link-local addresses,
which apart from making my life harder (cannot easily tell who is who)
it creates a minor problem with mtr/traceroute -6 -I replies.

The active OBSD firewall prefers to reply through it's link-local address and not it's global address.

13:20:30.403197 2001:648:xxxx:x::2 > 2001:648:yyyy:y::2: icmp6: echo request [hlim 1] 13:20:30.403224 fe80::92e2:baff:feb8:715d > 2001:648:xxxx:x::2: icmp6: time exceeded in-transit for 2001:648:yyyy:y::2

This is probably because of the link-local routes:
default    fe80::2a94:fff:fe4a:5a00%vlan123 UG 0 105920403 -    32 vlan123
fe80::%vlan123/64 fe80::92e2:baff:feb8:715d%vlan123 UCn 1 0 - 4 vlan123 fe80::2a94:fff:fe4a:5a00%vlan123 28:94:0f:4a:5a:00 UHLch 13 8594 - 4 vlan123 fe80::92e2:baff:feb8:715d%vlan123 90:e2:ba:b8:71:5d UHLl 0 2156 - 1 vlan123

vlan123: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
        lladdr 90:e2:ba:b8:71:5d
        description: External-10G-ipv6
        index 11 priority 0 llprio 3
        vlan: 123 parent interface: ix1
        vnetid: 123
        parent: ix1
        groups: vlan egress
        status: active
        inet6 fe80::92e2:baff:feb8:715d%vlan123 prefixlen 64 scopeid 0xb
        inet6 2001:648:yyyy:a::2 prefixlen 126

This might be normal but is there any way to change this behavior?

Thanks

G

Reply via email to