ospf with bfd ecmp conf

[Alon Dotan]

Hey All,

Im trying to configure the folowing topology:



lns box (multiple ppp devices that connect and disconnect) -> multiple 
default gw (gw load balance)

the issue is if one of the multiple gw crash I got routing blackhole...

Im trying to configurefd to delete dead peers from the osp routing table,

this is my conf:

debug protocols all;
log "/var/log/bird.log" all;
router id 10.10.11.10;

protocol device {
         scan time 1;            # Scan interfaces every 10 seconds
         export all;
}

protocol direct {
         interface "ppp*";
         export all;
}

protocol kernel {
         scan time 1;           # Scan kernel routing table every 20 seconds
         export all;             # Default is export none
         import all;
         kernel table 100;
         learn;
         persist;
}

protocol bfd {
         interface "eth1" {
                 min rx interval 20 ms;
                 min tx interval 50 ms;
                 idle tx interval 300 ms;
         };
}

protocol ospf MyOSPF {
         ecmp yes;
         export filter { if (net ~ 10.10.11.0/24) then reject; accept;};
#        export all;
         import all;
         merge external;
         area 0 {
            interface "eth1" {
                 hello 5;
                 bfd yes;
            };
         };
}

protocol static {
         import all;
}


routs learned by ospf:

[root at lns-testing-01 centos]# ip r show table 100
default proto bird
     nexthop via 10.10.11.11 dev eth1 weight 1
     nexthop via 10.10.11.12 dev eth1 weight 1
[root at lns-testing-01 centos]#


bfd status:

[root at lns-testing-01 centos]# birdc show bfd sessions
BIRD 1.4.5 ready.
bfd1:
IP address                Interface  State      Since Interval  Timeout
10.10.11.11               eth1       Up         08:09:55 0.050    0.250
10.10.11.12               eth1       Down       08:10:41 0.300    0.000

what Im missing?

Thanks,