From: Joseph Brunner (joe@affirmedsystems.com)
Date: Tue Sep 25 2007 - 10:59:23 ART
Perhaps there is a more 3L33t h4ck3r way to do it, but why not just set
The keepalive timers between R4&3 so low that they will always come up
first?
R3
Router bgp 300
Neighbor 4.4.4.4 timer 2 4
R4
Router bgp 300
Neighbor 3.3.3.3 timer 2 4
Also add this to R2
Router bgp 100
Bgp update-delay 300
This will set the initial update delay to 5 minutes, so at least when
Proctor Farmer John reloads your rack, R3&4 will have a chance to do their
thing first...
Any other solutions to do this better?
-Joe
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of
Sergey
Sent: Tuesday, September 25, 2007 3:37 AM
To: 'Cisco certification'
Subject: BGP sync problem
There are two routers in AS300 with enabled synchronization, both router
have bgp peering with as 100.
R1(as100) --- R3(as300) --- R4(as300)---R2 (as100)
The restriction says that R4 should prefer R3 for any external routes, so I
simply set on R4 weight = 2000 to R3.
But I've discovered one moment. If session between R4 & R3 comes first,
everything works fine. But if
session between R4 & R3 comes after R4 & R2 session than R4 thinks that
routes from R3 are
not synced (because no IGP routes in routing table) and prefer routes from
R2 in spite of that R3 has weight = 2000.
R4(config-router)#do sh ip bgp 70.70.70.0
BGP routing table entry for 70.70.70.0/24, version 6
Paths: (2 available, best #2, table Default-IP-Routing-Table)
Flag: 0x960
Advertised to update-groups:
1 2
100 700
172.16.36.6 (metric 20) from 172.16.34.3 (172.16.103.1)
Origin incomplete, metric 0, localpref 100, weight 2000, valid,
internal, not synchronized
100 700
172.16.124.1 (metric 64) from 172.16.124.5 (172.16.102.1)
Origin incomplete, localpref 100, valid, external, best
I want that R4 always select routes from R3 in spite of what session comes
first. What is the best way to accomplish it?
This archive was generated by hypermail 2.1.4 : Sat Oct 06 2007 - 12:01:15 ART