From: Howard C. Berkowitz (hcb@gettcomm.com)
Date: Sat Sep 27 2003 - 23:19:19 GMT-3
At 6:27 PM -0700 9/27/03, chipn@intraworx.net wrote:
>I'm looking for opinion on lab strategy.
>
>If you encounter a mutual redistribution scenario with multiple
>redistributions points on the lab, would it be in your best interest to
>explictly control your routing information to prevent route feedback? It
>is possible to have complete connectivity in this scenario without
>filtering, but still have the potential for routing loops.
>
>I thought this would be the best place to get some expert advice on the
>subject.
>
>Thanks.
Let me answer on several levels. First, in the real world, it's
almost always possible to avoid mutual redistribution with hierarchy,
advertising your routes to a gateway(s) and accepting default from
it. If you do need to exert tight control, that often can be done
better (again in the real world) with static or MPLS traffic
engineering, policy routing, iBGP, etc.
But given that the exam loves mutual redistribution, let me cite
several Murphyisms:
Murphy's First Law: What can go wrong, will.
Murphy's Second Law: What has gone wrong will get worse.
First Networking Corollary to Murphy's Laws: Murphy was an optimist
Second Networking Corollary to Murphy's Laws: Specify and control
everything unless you are quite certain that dynamic mechanisms
designed for the specific purpose (e.g., loop prevention within a
modern IGP) are adequate.
***Get your CCIE and a FREE vacation: Shop.GroupStudy.com***
This archive was generated by hypermail 2.1.4 : Wed Oct 01 2003 - 07:24:38 GMT-3