From: Chris Home (clarson52@comcast.net)
Date: Thu Jan 16 2003 - 01:25:31 GMT-3
Do whatever is best and works for you. One way is usually not the only way
or necessarily the absolute correct way to do things in the Cisco lab world
I am not looking for a gauruntee or for some written rule to it, but there
is logic. I would like to find the logic behind why it would ever be
necessary.
I am convinced there is no need to filter when restributing one protocol to
another on a single router to avoid routing a routing loop caused by the
redistribution. I have to go by that as a my own rule until someone shows me
otherwise or I have an experience that contradicts it. I can only know what
I think I know through experience and learning until there is evidence to
the contrary.
It would be an interesting challenge to come up with a scenario that would
require it.
BTW - I have asked the question about extra configuration. I was told that
you are not docked points for having extra configuration as long as the
requirements are met. I would not gauruntee it. It is just what I was told.
----- Original Message -----
From: "Jonathan V Hays" <jhays@jtan.com>
To: "'chris a'" <chris_atkins@blueyonder.co.uk>
Cc: <ccielab@groupstudy.com>
Sent: Wednesday, January 15, 2003 8:44 PM
Subject: RE: Mutual redistribution
> Chris,
>
> I'm sensing that you are looking for a rigid rule: a guarantee along the
> lines of "if there's only one redistribution point you don't need to
> worry about route filtering." I don't think anyone can give you such a
> guarantee.
>
> To restate what I said in a previous post in this thread, you are better
> off examining each situation carefully rather than always applying some
> rule (which may not be applicable in special circumstances).
>
> Ideally you should think ahead of time where the problems will be with
> mutual redistribution. But I use this crutch, since I am still learning.
> Once I have routing working properly, what I like to do *before*
> redistribution (as my lab study partner knows ;-) is to take a snapshot
> (in notepad) of every routing table. Do this BEFORE doing any
> redistribution. This way you can tell at a glance what routes belong
> where before redistribution messes everything up.
>
> After redistribute in both directions look at the resulting routing
> tables for routes that don't belong comparing them to the
> pre-redistributed tables. Some problems may be obvious administrative
> distance problems. Use trace to follow the less obvious paths.
>
> And I feel the solution of a complete set of filters every time you do
> redistribution may not be appropriate. And in some circumstances it is
> completely unnecessary. No one but Cisco knows the intimate details of
> how they grade the lab - but it may be possible to lose points for
> configuring something that's not required. Again - I'm not sure on this
> - just a rumor I've heard. When in doubt, ask the proctor.
>
> HTH,
>
> Jonathan
>
>
>
> -----Original Message-----
> From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of
> chris a
> Sent: Friday, October 25, 2002 6:41 AM
> To: Jonathan V Hays
> Cc: ccielab@groupstudy.com
> Subject: Re: Mutual redistribution
>
>
> Hi,
> I have had many thoughts about this and with only 3 weeks to go until my
> lab I am start to worry. I have found in Karl Solie's CCIE practical
> studies on page 621 he states the following.
>
> R1----------------R2------------------R3-----------------R4
>
> | ----- RIP version 1 ------------------|---EIGRP2001------|
>
>
> When performing mutual redistribution on R3 becuace there is only one
> redistribution point there is no need to perform any route filtering
> when using mutual redistribution.
>
> Any thoughts anyone please.
>
>
> Chris
>
>
>
> >
> > Hi,
> > In case I am doing a mutual redistribution at a single point between
> > any two routing protocols , do I really ever need to filter the routes
>
> > that are being redistributed for prevention of a loop.....????????
> >
> > I have never done this and have had no problems till now..... But
> > wouldnt want to discover something new during my lab.....
> >
> > But is there any case where filtering is required in for such a
> > redistribution?....what is the normal thing to do??
> >
> >
> > Regards,
> > Vijay.
> > .
> .
.
This archive was generated by hypermail 2.1.4 : Sat Feb 01 2003 - 07:33:50 GMT-3