From: Darby Weaver (darbyweaver@yahoo.com)
Date: Sat Feb 16 2008 - 01:27:22 ARST
Hmmm...
1. Learn to look at a topology
- Look for Layer 1 aka Physical Loops
- Look for Layer 2 aka Bridging Loops (SPT helps
here)
- Look for Layer 3 aka Routing Loops
When I say look at a topology, I mean draw a diagram
and touch the interfaces. Somehow this seems to help.
2. Next, ask yourself if you have a "Redistribution
Condition" - do you even have the possibility of
having a routing loop? Yes/No If you do then which
protocol has a higher AD and which protocol has a
lower AD? This is crucial to understanding how you
are going to deal with the routing loop.
Remember some protocols naturally have an interior AD
and an exterior AD. One in particular does not: RIP.
Stop for a minute and think of the options.
RIP | RIP
RIP | EIGRP
RIP | OSPF
RIP | BGP
EIGRP | EIGRP
EIGRP | OSPF
EIGRP | BGP
EIGRP | RIP
OSPF | OSPF
OSPF | EIGRP
OSPF | BGP
OSPF | RIP
See anything special here? Remember AD is key. Learn
how to match all route use distribute lists,
route-maps and AD in general as tools to influence or
even mark in the case of Route-Tagging various routes.
NMC has a nice little paper on this issue and it is
worth a trip to their website to find that gem of a
pdf.
My best advice is to take each protocol variation and
play with it. Using the tools and techniques
recommended by your favorite vendor.
Better to meet the enemy on your own terms. It may
take a good afternoon or so and or a couple of days
but if you do it methodically you should have it in a
relatively short amount of time.
3. logging console
logging buffered 100000
logging on
debug ip routing
log adjacencies per protocol
Let the router be the one to tell you that you are
having issues. Your job is to be more aware of the
tools you have to be able to influence the process.
Look at this pdf and lab it up. Thanks for NMC for an
example I know I use when my memory wants to fail me
on the matter or things do not seem to make sense any
more.
http://netmasterclass.net/site/articles/A%20Scenario%20with%20Multiple%20Redistribution%20Points.pdf
--- Scott Vermillion <scott_ccie_list@it-ag.com>
wrote:
> Hey Nate,
>
> I spent a disproportionate amount of my study time
> trying to get to that
> magical place where I could stare down any
> redistribution task of any
> complexity and just see what was going to happen
> underneath the hood (bonnet
> for our international list members). Never
> happened. What did happen was
> that 'debug ip routing' became a really good friend
> of mine. Also, just
> running full speed and smacking myself into that
> wall over and over again
> helped. I'm now to the point where I feel I can
> solve any redistribution
> task of any complexity - GIVEN ENOUGH TIME to do so.
> But sometimes it's so
> subtle that you're not likely to see it until you
> get some results from your
> debug. There were a few times when I even adjusted
> the AD of a given
> protocol or something like that just to see what, if
> any, impact it might
> have. That may or may not have sent me in a new
> direction.
>
> Don't let these discourage you like I did. Keep at
> it. Realize that no
> sane network engineer would really do what some of
> these lab scenarios are
> calling for. They're just meant to expose various
> (often times unintuitive)
> behaviors that you might at some point encounter.
>
> Regards,
>
> Scott
>
>
> -----Original Message-----
> From: nobody@groupstudy.com
> [mailto:nobody@groupstudy.com] On Behalf Of
> Cielieska Nathan
> Sent: Tuesday, February 12, 2008 7:45 PM
> To: ccielab certification
> Subject: Quick and dirty way to identify issues
> regarding redistribution
>
> Guys,
>
> So i ran into a brick wall last night which is IE's
> Workbook 3 Lab 8
> (Thanks Brians!). I thought i was really starting to
> get into the
> groove and this took me down a peg.
>
> With that said i actually ran into the section near
> the end where the
> redistribution happened. I was sitting in front of
> my routers just
> absolutely dumfounded and saved the configs and quit
> the lab early.
>
> I have listened to a few VOD's, did all of the basic
> lab scenarios
> regarding redistribution and even can deal with
> redistribution in a
> typical "i have two routers and they both have EIGRP
> and OSPF talking
> and you need to redistribute on both questions", so
> i think i got the
> BASIC jist of it down.
>
> Is there a methodolgy the truly experienced use to
> see the big
> picture with extremely complex redistribution. Maybe
> like a playbook
> one uses. I have read some folks tag their
> redistributed routes,
> others will allow redistribution on one router and
> immediately
> disallow on the second.. maybe a combination of
> both.
>
> Any suggestions on getting a truly in-depth
> understanding of
> redistribution would be extremely helpful right now.
> If that means
> labbing up some silly scenarios maybe someone can
> share a template of
> what i should be labbing up?
>
> How did you really learn redistribution.
>
> Regards,
> Nate
>
>
This archive was generated by hypermail 2.1.4 : Sat Mar 01 2008 - 16:54:48 ARST