From: Jung, Jin (jin.jung@lmco.com)
Date: Tue Dec 02 2003 - 14:59:02 GMT-3
I think this topic has been cover before about 5 times or more.
But here are my 2 cents on it.
When I started to study for CCIE - about 3 years ago.
I memorized all types of redistribution methods for each routing protocols.
- between OSPF and EIGRP use tag-
- between rip and OSPF use distribution list.
So forth and so on.
And I put in every time I saw redit. Weather or not its single of mutual
redistribution.
After while, I notice that I really did not need to do this.
In some cases, this created more problems, specially with distribution list,
blocking either BGP or some other routes from back bone.
Starting last year, I do not use any redistribution method in the beginning,
unless I see real loop. Than I fix the loop.
So its more of out come base. Trial and error.
You need to practice this and find the best solution for whatever situation
you have.
There is no single solution to fix all route loops.
You need to practice and find what works for you the best.
And don't try to over config, this may cause more problems.
--- Most time, loop does not occur.
But sometime you do,
I remember this one time, I had to use distance command, distribution-list
and tag just to fix one routing loop.
You will have to know all the methods,
This will definitely help you in the lab, but most importantly "in real
production".
Just my 2 cents.
> Jin Jung CCIE#12368
> Network Infrastructure Engineer
-----Original Message-----
From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of
David Hiers
Sent: Tuesday, December 02, 2003 12:28 PM
To: ccielab@groupstudy.com
Subject: RE: IGP Redistribution scenario
I don't know how they grade.
Enough people that passed the lab have expressed sufficiently varied
methodologies in this forum that I'm given to believe that the lab is
outcome-based. Producing the desired behavior seems more important than the
specific configs used.
David
********************************************
David Hiers
CCIE, CISSP
ADP Dealer Services
2525 SW First Avenue
Portland, OR 97201
v: 503 402 3703
email: david_hiers@adp.com
********************************************
-----Original Message-----
From: JamesGEF [mailto:jamesgef@sympatico.ca]
Sent: Tuesday, December 02, 2003 8:28 AM
To: ccielab@groupstudy.com
Subject: Re: IGP Redistribution scenario
Using ACLs for redistribution is probably the safest way, but it's not very
dynamic. If new internal routes are added to the IGP, then you will have to
modify the ACLs. Now in real life, this may be too tedious.
However, in the lab exam, is it better to be prudent and use ACLs to avoid
the loops and make your network stable? What I'm asking is, do you get
graded on how you do the redistribution, or do you get graded on simply
avoiding the loops, even if it's using ACLs?
James
----- Original Message -----
From: "David Hiers" <David_Hiers@adp.com>
To: <ccielab@groupstudy.com>
Sent: Tuesday, December 02, 2003 10:49 AM
Subject: RE: IGP Redistribution scenario
> Here's what worked for me:
>
> 1. Never redistribute without a routemap* (be a control freak) 2.
> Always write the most restrictive acl possible for the routemap (be a
bigtime control freak)
> 3. Always run "debug ip routing". It should be so quiet that you
> don't
even know that you're debugging anything at all (be a vigilant, bigtime
control freak)
> 4. Master admin distance, but only mess with it if someone holds a
> gun to
your head
>
>
> *Yes, even (and especially) for static and connected routes!
>
> David
>
>
> ********************************************
> David Hiers
> CCIE, CISSP
> ADP Dealer Services
> 2525 SW First Avenue
> Portland, OR 97201
>
> v: 503 402 3703
> email: david_hiers@adp.com
> ********************************************
>
>
> -----Original Message-----
> From: rnasr [mailto:rnasr@s2n.com.au]
> Sent: Monday, December 01, 2003 7:41 AM
> To: 'Thomas Larus'; 'Eric Cables'; 'ccielab'
> Cc: ccielab@groupstudy.com
> Subject: RE: IGP Redistribution scenario
>
>
>
http://www.cisco.com/en/US/tech/tk365/tk207/technologies_tech_note09186a0080
> 09487e.shtml
>
>
>
> -----Original Message-----
> From: nobody@groupstudy.com [mailto:nobody@groupstudy.com]On Behalf Of
> Thomas Larus
> Sent: Wednesday, December 03, 2003 2:17 AM
> To: Eric Cables; ccielab
> Cc: ccielab@groupstudy.com
> Subject: Re: IGP Redistribution scenario
>
>
> That is a brilliant idea, since it serves as a useful reminder of the
> potential for admin distance issues. I don't know if there is ever an
> end to the learning when it comes to the very difficult subject of
> route redistribution.
>
> I love this list! You know there are people who are snobby about this
list,
> and join it and quickly leave it. They think they are too smart to
benefit
> from it, or think that it is bad that you can different answers to a
> question here, and that, heaven forbid, one of the answers might be
> wrong. Or they think that the list is somehow tainted by the presence
> of people
who
> don't seem to know enough at the start to be preparing for the CCIE
> Lab.
>
> They may be right in some sense about some of the relative newcomers,
> but
if
> they leave because of these newcomers, it is their loss.
>
> I could have gone a long time without ever thinking of your neat idea.
>
> Sincerely,
>
> Tom Larus, CCIE #10,014
> Author of CCIE Warm-Up: Advice and Learning Labs
> http://www.ipexpert.com/products_services/product.asp?sku=ip7777
> tlarus@ipexpert.com
>
>
> ----- Original Message -----
> From: "Eric Cables" <skatter@blackened.net>
> To: "ccielab" <ccielab@cox.net>
> Cc: <ccielab@groupstudy.com>
> Sent: Tuesday, December 02, 2003 12:17 AM
> Subject: Re: IGP Redistribution scenario
>
>
> > That's a pretty big question. IGP redistribution is probably one of
> > the harder aspects of the lab, and you can bet you'll run into
> > problems.
> >
> > One method I like to use is redistribution with a route-map, and tag
> > the routes with an easy to remember # so that it can be identified,
> > and modified later based on that tag. Since you'd be redistributing
> > with a route-map you also have the flexibility to modify metrics,
> > etc..
> >
> > For example, you can tag the the routes based the IGP's
> > administrative distance, which should be easy to remember...
> >
> > Hope this helps,
> >
> > --
> > Eric Cables
> >
> > On Mon, 1 Dec 2003, ccielab wrote:
> >
> > > Is there any tried and true method to the approach for
> > > redistribution
of
> IGP
> > > routing protocols in the lab.
> > > Such as when you have a two point, mutual redistribution problem.
> Should I
> > > start with distributr lists on
> > > both routing protocols, or adjusting administrative distance, I am
> looking at
> > > netmaster sample lab for example.
> > > THanks
> > >
> > >
This archive was generated by hypermail 2.1.4 : Sat Jan 03 2004 - 08:25:34 GMT-3