On Sat, May 23, 2009 at 8:55 PM, Justin Credible <mista.justin.credi...@gmail.com> wrote: > On Sat, May 23, 2009 at 6:35 PM, Justin Credible > <mista.justin.credi...@gmail.com> wrote: >> >> Hi there, >> >> I am running OpenBSD 4.4 with OpenBGPD and multiple peers. >> >> For some reason the device is selecting Level3 as the default route for absolutely everything which is not statically set. >> >> On Level3 config i have set: >> >> set localpref 100 >> softreconfig in yes >> max-prefix 100 restart 300 >> >> For the others I have not set max-prefix. >> >> Also set >> >> nexthop qualify via bgp >> rde route-age evaluate >> >> and then stopped the session for Level 3 and started it again so it would seem "less stable" to the decision engine since it is a newer session, it is still the default for every single route. I even did a route flush and flushed them all, and did a refresh from another peer, at which point all routes came back, defaulting to Level3! >> >> How do i stop this from being my default route? >> >> The reason why is because it is not the best route most of the time. E.g. to some parts of the US it takes 16 hops through Level3, whereas Global Crossing will do it in 1 hop, Abovenet in 3, etc. >> >> Thanks! >> >> Ken > > If you need more examples here you go: > > # bgpctl show rib 199.185.137.3 > flags: * = Valid, > = Selected, I = via IBGP, A = Announced > origin: i = IGP, e = EGP, ? = Incomplete > flags destination gateway lpref med aspath origin > *> 199.185.136.0/23 64.x.x.x 200 1 3549 812 812 812 812 > 3602 22512 i > * 199.185.136.0/23 212.x.x.x 100 500 3356 6453 812 3602 22512 i > # route -n show | grep 199.185.136.0/23 > # route -n show | grep 199.185.136 > 199.185.136/23 212.x.x.x UG1 0 0 - 48 vlan400 > # route delete 199.185.136/23 > delete net 199.185.136/23 > # ping 199.185.137.3 > PING 199.185.137.3 (199.185.137.3): 56 data bytes > 64 bytes from 199.185.137.3: icmp_seq=0 ttl=245 time=150.000 ms > 64 bytes from 199.185.137.3: icmp_seq=1 ttl=245 time=155.865 ms > --- 199.185.137.3 ping statistics --- > 2 packets transmitted, 2 packets received, 0.0% packet loss > round-trip min/avg/max/std-dev = 150.000/152.932/155.865/2.958 ms > # route -n show | grep 199.185.136 > 199.185.136/23 212.x.x.x UG1 0 0 - 48 vlan400 > # bgpctl show rib 199.185.137.3 > flags: * = Valid, > = Selected, I = via IBGP, A = Announced > origin: i = IGP, e = EGP, ? = Incomplete > flags destination gateway lpref med aspath origin > *> 199.185.136.0/23 64.x.x.x 200 1 3549 812 812 812 812 > 3602 22512 i > * 199.185.136.0/23 212.x.x.x 100 500 3356 6453 812 3602 22512 i > > > I've even set my config to be EXTREMELY biased against Level3 but it > (the 212 address) still populates my routing tables: > > > BGP routing table entry for 199.185.136.0/23 > 3549 812 812 812 812 3602 22512 > Nexthop 64.x.x.x (via 212.x.x.x) from gblx-p1 (208.48.250.230) > Origin IGP, metric 1, localpref 200, external, valid, best > Last update: 00:26:45 ago > Communities: 3549:4356 3549:8013 3549:8023 3549:8043 3549:8073 > 3549:8090 3549:8163 3549:8173 3549:8223 3549:8233 3549:30840 > BGP routing table entry for 199.185.136.0/23 > 3356 6453 812 3602 22512 > Nexthop 212.x.x.x (via 212.x.x.x) from level3-p2 (4.69.187.4) > Origin IGP, metric 500, localpref 100, external, valid > Last update: 00:26:45 ago > > > # traceroute -n 199.185.137.3 > traceroute to 199.185.137.3 (199.185.137.3), 64 hops max, 40 byte packets > 1 212.x.x.x 0.550 ms 0.555 ms 0.448 ms > 2 4.69.136.93 0.529 ms 0.445 ms 0.575 ms > 3 4.69.136.90 11.273 ms 17.935 ms 11.317 ms > 4 4.69.139.73 11.396 ms 11.439 ms 11.317 ms > 5 4.68.63.106 16.769 ms 17.935 ms 17.939 ms > 6 195.219.195.37 11.772 ms 195.219.83.2 11.687 ms 195.219.195.89 11.562 ms > 7 195.219.243.14 12.17 ms 195.219.195.22 164.349 ms 164.471 ms > 8 195.219.144.10 83.354 ms 195.219.144.1 12.184 ms 12.62 ms > 9 195.219.144.10 83.355 ms 83.270 ms 216.6.98.1 109.634 ms > 10 216.6.98.1 109.835 ms 109.880 ms 216.6.98.30 163.602 ms > 11 216.6.98.30 163.552 ms 163.741 ms 64.86.115.38 178.523 ms > 12 64.86.115.38 178.788 ms 179.88 ms 24.153.7.137 203.204 ms > 13 24.153.7.137 180.416 ms 210.443 ms 238.549 ms > 14 24.153.4.77 177.923 ms 178.712 ms 24.153.3.38 173.844 ms > 15 24.153.3.38 173.921 ms 174.215 ms 173.595 ms > 16 204.50.251.202 196.411 ms 207.107.204.178 177.465 ms 176.209 ms > 17 207.107.204.178 177.542 ms 177.960 ms 176.719 ms > 18 199.185.230.2 177.924 ms 199.185.137.3 177.712 ms 199.185.230.2 > 176.215 ms > # route add 199.185.137.3 64.x.x.x > add host 199.185.137.3: gateway 64.x.x.x > # traceroute -n 199.185.137.3 > traceroute to 199.185.137.3 (199.185.137.3), 64 hops max, 40 byte packets > 1 64.x.x.x 10.505 ms 10.427 ms 10.316 ms > 2 64.208.169.150 98.472 ms 98.635 ms 98.513 ms > 3 69.63.248.98 97.96 ms 102.9 ms 97.141 ms > 4 66.185.80.186 138.946 ms 107.131 ms 107.136 ms > 5 24.153.4.74 149.191 ms 152.977 ms 159.354 ms > 6 24.153.3.34 146.816 ms 146.733 ms 146.861 ms > 7 204.50.251.141 146.942 ms 146.975 ms 146.860 ms > 8 207.107.204.178 149.314 ms 149.353 ms 149.371 ms > 9 199.185.230.2 149.149 ms 149.229 ms 149.231 ms > 10 199.185.137.3 149.314 ms 149.230 ms 149.483 ms >
I figured this one out. This particular problem was caused because i had set: nexthop qualify via bgp I don't know why that setting in particular set all of my routes to point at Level 3 regardless of the preferential settings against it, but how i got around it is simple. route add -mpath default gw1 route add -mpath default gw2 etc... Then change that setting to nexthop qualify via default Also make sure that the metric, localpref, etc are equal on all of the peers (unless you want one taking up all of the routing tables). then do a bgpctl reload The routing tables seem to have evened out now and become more "realistic" and unbiased. There are now more routes through GBLX than Level3 but only a few thousand, as opposed to the previous problem of no dynamic routes pointing to GBLX. Regards, Ken
