From: Justin Menga (Justin.Menga@xxxxxxxxxxxxxxxxxx)
Date: Wed Feb 07 2001 - 06:00:23 GMT-3
Consider the router (let's call her CCIE) that is transmitting routes to
peers.
If a peer is an EBGP peer, then CCIE will send each prefix to the EBGP peer
with a next hop address of it's exit interface. Hence, the router has
automatically set the next hop address to her 'self' as a rule of the
protocol. As most EBGP peerings are over directly connected interfaces, the
receiving EBGP peer can reach the next hop as it is a directly connected
address.
If a peer is an IBGP peer, then CCIE will send each prefix but will not by
default alter the next hop address. The first decision in BGP about whether
to consider a BGP route is IF THE NEXT HOP IS REACHABLE. If it is not, then
don't even consider the BGP route. So if CCIE sends a prefix to an IBGP
peer and doesn't alter the next-hop address, and that address is unreachable
by the IBGP peer, the route is never accepted. To alleviate this, you can
change the next-hop address to the sending router's exit interface
(actually, the update source IP address), and this is the functionality of
next-hop-self.
Regards,
Justin Menga CCIE #6640 MCSE+I CCSE
WAN Specialist
Computerland New Zealand
PO Box 3631, Auckland
DDI: (+64) 9 360 4864 Mobile: (+64) 25 349 599
mailto: justin.menga@computerland.co.nz
-----Original Message-----
From: Steven Weber [mailto:itweber@netzero.net]
Sent: Wednesday, 7 February 2001 5:03 p.m.
To: ccielab@groupstudy.com
Subject: next hop self...
I know this is a silly question but I am having a hard time
understanding the function of the next hop self command. I read up on it
in Halabi and I'm still having problems. Anyone care to shed some light?
Thanks,
Steve
This archive was generated by hypermail 2.1.4 : Thu Jun 13 2002 - 10:28:40 GMT-3