From: Pandora (pandoraytchan@hotmail.com)
Date: Mon Apr 28 2003 - 04:22:26 GMT-3
Can I said that RR solution is not need for IBGP fully meshed?
Btw, what redundancy are you refer? Router redundancy or link redundancy?
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of
OhioHondo
Sent: Sunday, April 27, 2003 8:14 PM
To: Pandora; ccielab@groupstudy.com
Subject: RE: Redundant Route Reflectors
RR's used for redundancy are in the same cluster.
RR's that are not used for redundancy are not in the same cluster.
RR's must be fully meshed.
RR-Clients hang off of RR's. They are not part of the full mesh.
If 2 RR's are used for redundancy, Clients of those RR's are connected to
both of them.
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com]On Behalf Of
Pandora
Sent: Sunday, April 27, 2003 5:35 AM
To: ccielab@groupstudy.com
Subject: Redundant Route Reflectors
Hi all,
What's the "correct" and "formal" configuration of redundant route
reflectors?
Suppose: Ra to Rb, Rb to Rc and Ra to Rc (fullly meshed). Does we need to
config Ra & Rb as RR? (Rc as RRC) or Just one RR and two RRC?
I also found that there're different point of view in RR Cluster in
different cisco book:
From "Cisco BGP-4 Command and Configuration Handbook" P.66
"Two RR are configured for redundancy and therefore must have the same
cluster ID"
But, from "Troubleshooting IP Routing Protocols" P.785-790, "It's
recommended that.RR should not be put in the same cluster"
Any comment of this two different point of view?
Thanks for any help.
Pandora
This archive was generated by hypermail 2.1.4 : Thu May 01 2003 - 13:36:08 GMT-3