On 2008-09-01, Russell Howe <[EMAIL PROTECTED]> wrote: > How does ospfd choose the address to send from? I thought it might be > something to do with the multicast route, but that's set to be on 'lo0', > whereas my new loopback interface is lo1.
Packets are sent out of interface X with the source IP address set to the IP address assigned to interface X (see send_packet() in packet.c). Where you set an address with the nic0:11.22.33.44 format in ospfd.conf, the address you configure there is the address used. otherwise it's the first IP address on the interface (I don't like to rely on that, better to set the address explicitly). I do use a loopback /32 on lo1 with bgpd/ospfd and don't see this sort of problem. (I do have to be careful and look out for problems with BGP nexthops when I restart things though, which may be connected with this but I'm not sure...) In my case it's from a separate subnet which is not announced as a whole subnet into OSPF, just as host routes. From the information you gave we can't tell how you arranged this, but if you're sending the loopbacks from a network which is otherwise announced, maybe it's worth trying that..
