BGP 'best practices' in lab

From: Desimone, Aurelio (ADesimone@refco.com)
Date: Wed Sep 11 2002 - 17:29:52 GMT-3


Because the lab is not based on real-life scenarios, I was wondering whether
we should use certain commands unless absolutely necessary. I am used to
using 'no sync' and 'no auto' in real life and 'next-hop-self' on ibgp peers
when I setup bgp because its almost always necessary in order to ensure a
stable network. But I was wondering what people's opinions were on using
these commands in the lab. Also, is it best to combine filtering in one
route-map or to use distribute/prefix/filter-lists when possible? Some
people tell me it doesn't matter just as long as you fulfill the
requirement. Other people say that you should do it in the least amount of
lines (which might be the same if you already have a route-map started).
Obviously we must accomplish the task within the guidelines which will
probably help me make the right decision. I'm just trying to get some
feedback/opinions.

Thanks
Aurelio



This archive was generated by hypermail 2.1.4 : Mon Oct 07 2002 - 07:43:49 GMT-3