From: ccie2be (ccie2be@nyc.rr.com)
Date: Sun Jun 06 2004 - 14:53:35 GMT-3
Charles,
I agree with you. Redist, as a topic, isn't that well documented and the
info is spread out.
I think if Cisco Press came out with a book just covering redistribution it
would sell pretty well.
In any case, to answer your question, here's what I'd do.
1) Know the particulars and issues of redist every protocol into every
other protocol.
With regard to particulars, what I'm talking about is this:
Know what gets redist by default for each protocol e.g. do
internal & external routes get redist or just internal?
Do you need to specify metrics or special keywords lilke
"subnets"?
Will a summary route created in one protocol also be redist, by
default?
Will the connected interfaces running the redistributed protocol
also be redist ( in isis, they won't)?
With regard to issues, I'm talking about:
Is there a classless to classful issue? (Without rip v1,
probably not anymore, but, be aware of it just the same)
Is there a potential route loop?
Are any routers running more than 2 routing protocols? If so,
know that if you try to igp1 redist -> igp2 -> igp3, igp1 routes won't get
into igp3. For ex., if you redist static routes into igp1 and then redist
igp1 into igp2, the static routes won't get into igp2. You have to manually
redist the static routes twice, once into igp1 and again into igp2.
2) Know the tools at your disposal you can use to address the above
issues:
Adjusted AD
Using tags
Using metrics
Using route filtering and route-maps
3) Develop a strategy for complicated redist scenario's
(I use the strategy described by Internetwork Expert)
1) Look at a picture of your network and see what potential
issues exist i.e. are 2 protocols touching at 2 or more rtr's?
2) Start at one edge of the network & do redist both ways on
one router & then verify. Have all the routes that should be redist
actually been redist? Have any routes that shouldn't have been redist
actually been redist? Can you ping everywhere you should be able to ping at
this point? If yes, go on to next router. If no, stop and correct any
problems now before moving on.
3) Repeat step 2 until finished.
Obviously, this explanation is somewhat simplistic and overly brief, but
hopefully, it will give you a bit of a roadmap to follow.
HTH, Tim
----- Original Message -----
From: "Charles T. Alexander" <charles.t.alexander@verizon.net>
To: <ccielab@groupstudy.com>
Sent: Sunday, June 06, 2004 10:39 AM
Subject: Redistribution Headaches
> The most difficult problem in doing labs I find is multiple
> redistributions on several routers. I have read all the usually books
> many times. I still run into issues. I guess I am looking for an
> article on the strategy and approach to redistribution.
>
> _______________________________________________________________________
> Please help support GroupStudy by purchasing your study materials from:
> http://shop.groupstudy.com
>
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html
This archive was generated by hypermail 2.1.4 : Sat Jul 03 2004 - 19:40:33 GMT-3