OSPF and strange area ID

David Rohleder davro at ics.muni.cz
Wed Mar 7 16:32:10 CET 2001


Hello,

i have got strange problem running OSPF:


   R ----- F -----


R is Bay Networks VNR
F is Debian Linux 2.2 (with kernel 2.4.2)

when i try to configure OSPF:

protocol ospf MyOSPF {
        # rfc1583compat yes;
        area 0.0.0.2 {
        #       stub no;
        #       tick 1;
                interface "eth0" {
                        cost 500;
                        dead count 4;
                        type broadcast;
                };
                interface "eth1" {
                        cost 550;
                        dead count 4;
                        type broadcast;
                };
                interface "eth2" {
                        stub yes;
                        cost 10;
                };
                interface "eth3" {
                        stub yes;
                        cost 10;
                };
        };
}

i am getting strange debug messages:

Mar  7 16:32:53 fw5 bird: MyOSPF: Bad OSPF packet received: other area 33554432
Mar  7 16:32:53 fw5 bird: MyOSPF: Discarding

and tcpdump shows:

16:29:36.338278 fw5.ics.muni.cz > OSPF-ALL.MCAST.net: OSPFv2-hello 44: rtrid 147.251.8.17 area 0.0.0.2 [|ospf] [tos 0xc0] [ttl 1]
16:29:43.932693 routics-are.brno.cesnet.cz > OSPF-ALL.MCAST.net: OSPFv2-hello 52: rtrid routics-are4.brno.cesnet.cz area 0.0.0.2 [|ospf] [tos 0xc0] [ttl 1]

i don't see any neighbour.

when i change area to 2.0.0.0 (33554432), bird doesn't drop packets and goes to
state init/other with R, but tcpdump shows, that bird sends packets
with area 2.0.0.0 which collides with R config.

It seems to me, that bird does't do translation from net byte order to
host byte order when receiving packets (I am not sure, but area
0.0.0.0 could work :-)




-- 
-------------------------------------------------------------------------
David Rohleder						davro at ics.muni.cz
Institute of Computer Science, Masaryk University
Brno, Czech Republic
-------------------------------------------------------------------------



More information about the Bird-users mailing list