From: Vytautas Valancius (valas@mail.lt)
Date: Tue Nov 23 2004 - 12:33:37 GMT-3
Hi Group,
What level of redundancy is required while doing route redistribution in
CCIE LAB?
Lets say we have core OSPF routing domain. Also we have some 'non-transit'
(as it is called in Bruce's white paper on redistribution) RIP routing
domain attached to core in two points - two routers. So we need to do
redistribution in two points to achieve redundacy.
The interesting thing is that Bruce in its whitepaper suggest to filter
non-native RIP routes from going back to core. Its easy but this prevents
full redundancy! Lets say core brakes in the middle and RIP domain will not
be used as transit to connect those broken parts.
I came to this question as I was doing preparation labs. My sollutions
worked, there was full reachability in all labs. The problem was that none
of my sollutions were identical to LAB sugested sollutions :-)
So do we really need full redundancy or not? What will give us points in the
LAB? Can we define that some routing domain is non-transit? My thought was
that all routing domains with two points of attachement could be transit in
case of failure in the core.
Best Regards,
Vytautas Valancius
This archive was generated by hypermail 2.1.4 : Thu Dec 02 2004 - 06:57:49 GMT-3