From: dusth@comcast.net
Date: Fri Nov 11 2005 - 11:47:03 GMT-3
Good idea to preven your routing loop is always prefer stub routing protocol for it native routes and it stub learn routes. The core routing protocol - transit protocol will be prefer to route for the rest of other rotues. Saying this when you redistribute stuff into transit, allow only stuff routes enter the transit routing. When you redistribute transit routing protocol into stub, deny your stub from from entering - it works for the lab but not scalable for production in case you have ton of stub routes traverse from the far side and back to your stub entry/exist point and these stub routes are filtered from here instead of you should filter these stub routes from the far side as possible so you won't waist your bandwidth for these stub to traverse back. If you stub router has higher AD - like rip - than the transit protocol - OSPF, you best practice should lower AD on your rip to lower than OSPF.
When you do the lab or practice lab, REDUCE the redistribute point as less as possible. Don't get to fancy of need multiple redistribution points for redundancy and scalability. DO not do more than what to ask for. If you have to pick for redistribution point, you might need to analyze your network topology figure out which entry point is the best for doing redistribution.
For example, you have hub and spoke is runing OSPF and both hub and one of the spoke are running EIGRP. As the rule of thumb, do the mutual redistribution on the hub. Then change OSPF external to be higher than 170 or change external EIGRP to lower than 110. Then loop will be prevented.
Dustin
-------------- Original message --------------
> Hello group,
> I thought it might be a good idea to start a list of things to look out for
> that can waste your time on the lab and some suggestions on how to avoid
> them.
> 1. Multiple points of redistribution - I spent several hours in my last
> practice lab chasing down routing loops.
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html
This archive was generated by hypermail 2.1.4 : Thu Dec 01 2005 - 09:12:06 GMT-3