From: Jonathan Hays (nomad@gfoyle.org)
Date: Sat Jan 31 2004 - 10:53:33 GMT-3
you wrote:
>-----Original Message-----
>From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On
>Behalf Of Stuart Reabow
>Sent: Saturday, January 31, 2004 5:30 AM
>To: 'CCIE GroupStudy'
>Subject: BGP Router-ID vs Update-source
>
>
>Hi all,
>
>It appears that if you configure a bgp router-id as the
>loopback0 add, then
>there's no need to use the update-source lo0 command. The
>router-id will be
>used to source all comms. Does anybody see any problems with this?
>
>Stuart
= = =
Perhaps I am not understanding your statement, but I can't see that
configuring the BGP router-ID alone will change the source address of
BGP communications. The router-ID only indicates which router sourced
the update. But by default the router will use the interface of the
shared network with its BGP neighbor to source all communications. The
only way to change this (that I know about) is to configure the "bgp
neighbor update-source" command.
The iBGP example below has Loopback0 (139.10.6.6) configured as the BGP
router ID and advertises Loopback 6 (6.6.6.6) to R9, in the same AS. The
"sh ip bgp" output on R9 indicates that sees the 6.6.6.0 network
advertised by 139.10.69.6 (the IP address of the egress interface of R6
connected to R9). The BGP router ID for R6 (139.10.6.6) is shown in
parentheses in the output, but this is just a label.
R6#sh ip int brief | i Loopback
Loopback0 139.10.6.6 YES NVRAM up
up
Loopback6 6.6.6.6 YES manual up
up
R6#sh run | begin bgp
router bgp 100
bgp router-id 139.10.6.6
bgp default local-preference 75
bgp log-neighbor-changes
network 6.6.6.0 mask 255.255.255.0
neighbor 139.10.69.9 remote-as 100
!
---
R9_FRSw#sh ip bgp 6.6.6.0
BGP routing table entry for 6.6.6.0/24, version 0
Paths: (1 available, no best path)
Not advertised to any peer
Local
139.10.69.6 from 139.10.69.6 (139.10.6.6)
Origin IGP, metric 0, localpref 75, valid, internal, not
synchronized
R9_FRSw#
I changed the connection to eBGP (I put R6 and R9 in different AS's) but
got the same results.
Also, I ran debug for several minutes and saw no evidence that the
source address of any communication was the router-ID. Perhaps I do not
understand what you are asserting and if so, I apologize.
Please clarify,
Jonathan
This archive was generated by hypermail 2.1.4 : Mon Feb 02 2004 - 09:07:52 GMT-3