From: gladston@br.ibm.com
Date: Fri Jul 22 2005 - 16:05:30 GMT-3
Hi,
Tests show that 'ip rip triggered' updates build a stable ISDN link.
Do you agree?
When main interface is up, traffic goes over serial.
When main interface goes down, traffic goes over ISDN.
If there is no traffic, BRI goes down and remain on that state until there is traffic to remote router or there is any routing update.
This is the config used:
interface Dialer50
description RIP
ip address 148.5.45.8 255.255.255.0
ip rip triggered
encapsulation ppp
dialer pool 50
dialer string 4441111111
dialer-group 2
ppp authentication chap
!
dialer-list 2 protocol ip list 122
!
access-list 122 permit ip any any
Rack2R5#sh ip rip da | b 148.5.147
148.5.226.0/24
[10] via 148.5.45.4, 00:31:53 (permanent), Dialer50
If BRI goes down because of no traffic, this entry is still on the RIB, avoiding routing changes and do not activate BRI.
I am considering that snapshot is not used anymore on the lab. I remember to read it someplace. If it is true, then 'ip rip triggered' and dialer-watched would be the options to avoid flapping on BRI. Agree?
This archive was generated by hypermail 2.1.4 : Sun Sep 04 2005 - 17:00:30 GMT-3