From: Scott Vermillion (scott_ccie_list@it-ag.com)
Date: Tue Mar 04 2008 - 18:06:26 ARST
Oh, I remember this mess. Sorry Brians, this was not one of Vol II's
shining moments, IMHO. John, there's a lot more wrong with the solution
than just this!! You definitely need to read your way through the IE
forums, as there are many Lab 2 Task 4.11 threads and many posts within them
(I just now had fun reading through my own frustration with this task! - I
was about to give up on the CCIE after this bitter little pill ;~) ).
One thing that I did throughout lab prep was to save my configs at the
conclusion of each major section. So I fired up SW1 and did a 'copy
flash:Lab2_4.11 run' and had a look at my RIP config. I did _not_ include
'distance 109'! And I verified the hell out of my solution to this task (as
I recall, this lab took me about a week to ultimately complete - probably
two days of that were dedicated solely to 4.11!), so I'm relatively
confident in saying it's not required. But read on...
In reviewing my own postings to the forums, I see that I too struggled with
the question of why this was required. One guy (R&S CCIE) offered what on
the surface sounded like a really good explanation, which basically had to
do with RIP being redistributed into OSPF and OSPF being redistributed into
EIGRP and EIGRP being redistributed back into OSPF and SW1 winding up
pointing towards R1 for RIP-learned routes. Like I said, sounds good on
paper, but if you implement a truly effective solution to mutual
redistribution between OSPF and EIGRP, this simply won't be an issue. This
a word-for-word quite I scribbled into my workbook:
"Solution given sucks ass. Do route tagging and filtering instead."
;~)
-----Original Message-----
From: John [mailto:jgarrison1@austin.rr.com]
Sent: Tuesday, March 04, 2008 12:29 PM
To: Scott Vermillion; 'Sadiq Yakasai'
Cc: 'Hash Aminu'; 'Carlos Alberto Trujillo Jimenez'; ccielab@groupstudy.com
Subject: Re: redistribution
Scott,
It's lab 2 obviously the redistribution part specifically ripv2 and ospf on
SW1. This is the one command I could make no sense of. So I ass|u|me that
I'm missing something. The other issue might be is that one of the BB
routers is connected to the rip domain, which presents it''s own problems on
IE racks. Since I may or may not get connectivity to any BB router, and
have no access to them.
----- Original Message -----
From: "Scott Vermillion" <scott_ccie_list@it-ag.com>
To: "'John'" <jgarrison1@austin.rr.com>; "'Sadiq Yakasai'"
<sadiqtanko@gmail.com>
Cc: "'Hash Aminu'" <hashng@gmail.com>; "'Carlos Alberto Trujillo Jimenez'"
<carlos.trujillo.jimenez@gmail.com>; <ccielab@groupstudy.com>
Sent: Tuesday, March 04, 2008 1:06 PM
Subject: RE: redistribution
> Hey John,
>
> You're doing IEWB Vol II, no? Might be a good idea at this point if you
> just tell us which lab. Maybe somebody will remember that task or will
> have
> taken some notes on it (I took extensive notes on the earlier labs, then
> really none at all towards the end). I do recall times when this sort of
> thing made sense to do. I also seem to recall one or two cases where I
> tested the configuration and something like this made zero observable
> difference. In the latter case, I chalked it up to an "artifact" of a
> previous version of the lab. They do update these over time and sometimes
> there are orphans left behind in the Solutions Guide...
>
> Just a thought.
>
> Scott
>
>
> -----Original Message-----
> From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of
> John
> Sent: Tuesday, March 04, 2008 11:48 AM
> To: Sadiq Yakasai
> Cc: Hash Aminu; Carlos Alberto Trujillo Jimenez; ccielab@groupstudy.com
> Subject: Re: redistribution
>
> Sadiq,
>
> No there is one, and only one point of redistribution. The rip routing
> information has no way through the ospf domain to get back to the
> originating router(redistribution point) t. Thats why I didn't bothr with
> a
>
> detailed description of the network.
>
> I'm only talking about one point of redistribution between the two routing
> domains. There are no other points of redistribution between the ospf and
> rip domains. The rip domain does connect to a backbone router which is
> part
>
> of the rip domain.. So there again I'm wondering why make the AD on the
> redistributed ospf routes lower than the ospf routes on that router? What
> is it about the rip domain that would require that when there are no other
> redistribution points for that network?
> ----- Original Message -----
> From: "Sadiq Yakasai" <sadiqtanko@gmail.com>
> To: "John" <jgarrison1@austin.rr.com>
> Cc: "Hash Aminu" <hashng@gmail.com>; "Carlos Alberto Trujillo Jimenez"
> <carlos.trujillo.jimenez@gmail.com>; <ccielab@groupstudy.com>
> Sent: Tuesday, March 04, 2008 10:56 AM
> Subject: Re: redistribution
>
>
>> Hey John :) i think I got a hold of the scenario here, see if its wat
>> you are trying to explain:
>>
>> From what I understand here, you seem to have a loop in the OSPF
>> domain where the rip routes are redistributed in OSPF but they go into
>> the OSPF domain and have the potential of cycling and coming back to
>> this router that is actually redistributing these routes.
>>
>> When they come back to this router (if they do), they have the
>> potential of displacing the RIP routes and thereby disrupting this
>> routers belief of where the routes actually originated from because of
>> the fact that OSPF routes (AD 110) will displace RIP routes (AD 120).
>> Now the safest practice here is to make RIP routes have an AD of 109
>> so that no matter what, this router will always prefer the RIP
>> prefixes and will always redistribute them into OSPF.
>>
>> HTH
>>
>> Sadiq
>>
>> _______________________________________________________________________
>> Subscription information may be found at:
>> http://www.groupstudy.com/list/CCIELab.html
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html
This archive was generated by hypermail 2.1.4 : Tue Apr 01 2008 - 07:53:52 ART